').text(selectedLabel));
html.append(closedEnded);
}
// Handle Open-Ended Question Section
var openEndedQuestion = response.survey_response.answers.find(answer => answer.type === 'open_ended');
if (openEndedQuestion) {
var openEnded = $('
');
openEnded.append($('
').text(openEndedQuestion.question.headline.value));
openEnded.append($(' ').text(openEndedQuestion.value));
html.append(openEnded);
}
return html;
}
```
# Conclusion
The new Customizable CSAT survey API is quite easy to use and allows for insights that currently aren't available via native UI features in Zendesk. I do hope that sooner rather than later the Agent Workspace will natively reflect the new survey responses, making apps like mine unnecessary.
As for the API itself, I'd love for a way to retrieve surveys for a given ticket or user more directly than going via the Audit Log api for a given ticket. There's also currently no POST or PUT endpoint so you can't create integrations that allow customers to give feedback. You have to go over the native Messaging or Email triggers.
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Using Switchboard to combine Zendesk Bots and AI Agents powered by Ultimate
URL: https://internalnote.com/using-switchboard-to-combine-zendesk-bots-and-ai-agents-powered-by-ultimate/
Last updated: 2025-04-14T07:07:34.000Z
Earlier this month I wrote a big article on the history of Zendesk Conversation channels. This article explained the multiple solutions Zendesk offered with regard to chat channels ranging from the old Zopim Chat right up to their newest AI Agents powered by Ultimate.
[A History of Zendesk chat and messaging channelsOver the years, Zendesk evolved from simple email ticketing to offering a full suite of customer care solutions. They now support multiple channels, AI agents, social integrations, and web widgets. This article will dive into how their complex platform works, helping navigate conversational setups.Internal NoteThomas Verschoren](https://internalnote.com/history-of-zendesk-messaging/)
One of the sections of the article showed the following flowchart, showing a setup where a customer has both an AI Agent running on the existing Zendesk Bot technology, and a modern bot powered by Ultimate.
Since more and more Zendesk users will be doing the upgrade to this more modern and powerful automation solution the coming months, chances are that some of them will run into this exact scenario.
When migrating bot platforms you can't just migrate all your brands and flows at once. Most customers will want to approach this with a phased process where they migrate one brand (or channel) at a time from the *Zendesk Bot* to *Ultimate.* Or, if you're coming from the likes of Ada or Certainly, you might want to temporarily run your Ada bot and Zendesk AI Agent in tandem.
This article will show you how to set this up.
# Some context
Zendesk Suite comes with the Sunshine Conversations platform build in. Sunshine Conversations, or SunCo in short is the engine that connects channels like WhatsApp or a web widget, Bots like Ultimate or Zendesk, and Agent environments like Zendesk together.
Within SunCo there's a tool called switchboard. The switchboard orchestrates the interaction between customers, agents and bots by routing an incoming conversation to the right AI Agent, and escalate to the right platform.
Each Zendesk environment, regardless of having actively used SunCo or not, has a switchboard. This switchboard links your messaging channels to the Zendesk Bots, and handles the handover from bot to agents in the Agent Workspace.
The switchboard has a list of switchboard integrations linked to it. These integrations are your bots and agent environments. The switchboard has a default integration it routes *all* conversations too, and your integrations all have a default next responder. So for example you can have the Zendesk Bot as the default integration, and it hand off conversations to the Zendesk Workspace.
Your entire SunCo setup has a list of integrations linked to it, these are your social channels, web widgets and mobile SDKs. Each integration has a default responder (one of the switchboard integrations), and also lists a next responder for when e.g. the bot escalates to an agent. Those responders can be the default ones, or ones specifically set for that integration.
All interactions with the switchboard and its integrations happens on the API level, so when you want to make change or inspect the switchboard you need to setup a Sunshine Conversation API credentials in the Admin Center.
💡
You can find more info on the `url`, and `appId` and tokens to be used in the [SunCo documentation](https://docs.smooch.io/guide/authentication-overview/?ref=internalnote.com).
# Gathering the initial context
## Switchboard
The first API call we're going to make is one to `{{url}}/v2/apps/{{appId}}/switchboards`
This returns our main switchboard used by our Zendesk instance. From this data we need to copy the `id` of the switchboard, in our case *5f8ece73c031ec000c74a680.*
```json
{
"switchboards": [
{
"id": "5f8ece73c031ec000c74a680",
"enabled": true,
"defaultSwitchboardIntegrationId": "6419d5aad477770116cfa596"
}
]
}
```
## Switchboard integrations
Once we have our switchboard, we can collect its integrations. We do this by making an API call to `{{url}}/v2/apps/{{appId}}/switchboards/{{id}}/switchboardIntegrations` with `id` being the switchboard id we captured in the previous step.
This returns an array of all the current switchboard integrations. You'll find two important ones in this list: `zd:agentWorkspace` – in other words, your agents – and `zd:answerBot` – your Zendesk Bots.
```json
{
"switchboardIntegrations": [
{
"id": "5f8ece7337f5bd000d6ae0ff",
"name": "zd-agentWorkspace",
"integrationId": "5f8ecbf2e2059d2c993eab02",
"integrationType": "zd:agentWorkspace",
"deliverStandbyEvents": false,
"nextSwitchboardIntegrationId": "6419d5aad477770116cfa596",
"messageHistoryCount": 10
},
{
"id": "6419d5aad477770116cfa596",
"name": "answerBot",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"deliverStandbyEvents": false,
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff",
"messageHistoryCount": 10
}
]
}
```
If your instance has multiple brands and bots you might wonder why you're seeing only one `zd:answerBot` in this list. Zendesk obfuscates the list of bots in your instance and takes care of routing the right channel (brand) to the right bot with internal logic not accessible over the SunCo APIs. But no need to worry, you can (un)link Zendesk bots to channels in the Admin Center and Zendesk takes care of their own native bots.
The item we want to note down here is the `integrationId` of the `zd:answerBot`, in our case *5f8ece75e2059d2c9995186b.*
## Integrations
And finally we need a list of the existing integrations in our instance. We need the IDs of these integrations so that, once we link the Ultimate bot, we can revert some of these channels back to being linked to the Zendesk Bot.
We can collect all integrations by calling `{{url}}/v2/apps/{{appId}}/integrations` . This returns a long list of channels similar to this one (I removed some elements to make it more readable)
```json
{
"integrations": [
{
"id": "61ea8723f4aa6100eb8a69e5",
"type": "web",
"displayName": "Internal Note",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
},
{
"id": "646b411abc0c536ba1a98263",
"type": "whatsapp",
"displayName": "WhatsApp",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
},
{
"id": "6706411852c663af3215a44f",
"type": "web",
"displayName": "AI Agent (Ultimate)",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
}
]
}
```
If we look closes to one of the integrations we'll note a few things:
```json
{
"id": "61ea8723f4aa6100eb8a69e5",
"type": "web",
"displayName": "Internal Note",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
}
```
The `id` is the id of the integration, and is what we use in the API to make changes to its configuration. Note down the IDs of the channels you want to keep on Zendesk, in my case it's one of the web widgets and my WhatsApp channel.
The `defaultResponder > integrationId` is where we define which Bot (or switchboard integration) answers when a customer talks to that channel. In our case this is currently always the Zendesk Bot. The `nextSwitchboardIntegrationId` defines who should pick up the conversation if the bot needs to escalate. In our case this corresponds with the Agent Workspace.
# Connecting Ultimate
Connecting Ultimate to Zendesk happens via a Sunshine Conversation integration. This is entirely handled within Ultimate via a nice UI flow, but you do need to grab a set of Sunshine Conversation API credentials from the Zendesk Admin Center first. I **highly** recommend creating a new pair of credentials for this so you can keep a clear overview of tokens used for your own API work, Ultimate, and [outbound messages](https://internalnote.com/relay-app/) via e.g. the Relay app.
When linking Ultimate to Zendesk you can choose between SunCo or SunCo groups. I prefer the latter. The setup is a bit more convoluted but it allows for linking multiple Ultimate bots to Zendesk. You can use Ultimate's routing rules to define which bot should answer which channel, similar to how you would setup Zendesk Bots and multiple channels in Zendesk itself.
Once you link Ultimate **all** your channels will answer with the Ultimate Bot you just configured. If we take a look at our previous integrations list for example, we'll now see that our integration has a new `defaultResponder`, while retaining all other settings.
```json
{
"id": "61ea8723f4aa6100eb8a69e5",
"type": "web",
"displayName": "Internal Note",
"defaultResponder": {
"id": "670646ef88ab3af5a96c49b1",
"integrationId": "670646eeec1612bbbc225995",
"integrationType": "ultimate",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
}
```
## Reverting some channels back to Zendesk
In my setup I have two brands. One is Internal Note, one is my Ultimate demo environment.
While I want to keep the web widget I created in Zendesk for Ultimate to remain linked to the Ultimate AI Agent, I want to make sure that my Web Widget and WhatsApp channel for Internal Note still answer with a Zendesk Bot. Otherwise all [demos](https://internalnote.com/tag/bots-and-messaging/) I build for this website will break.
[Learn how to build a full-featured Flow Builder Bot for Zendesk.In this article we will build a full-featured Flow Builder Bot for Zendesk. We’ll use every step type, use API calls and variables and show you how to create a bot yourself in a full length video tutorial.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-dinosaurs/)
In order to revert some integrations back to the Zendesk Bot we need to execute a final API call for each integration we want to revert.
We need to send a PATCH command to the following url: `{{url}}/v2/apps/{{appId}}/integrations/{{id}}` with `id` being the id of the integration we want to update.
The payload of this patch command contains the `id` of the `switchboardIntegration` we want to set as its `defaultResponder`, in my case the `id` of the `zd:AnswerBot` collected earlier in this article.
```json
{
"defaultResponderId": "6419d5aad477770116cfa596"
}
```
Once you've executed this command for each integration, you'll end up with a new integrations list that looks like this:
```json
{
"integrations": [
{
"id": "61ea8723f4aa6100eb8a69e5",
"type": "web",
"displayName": "Internal Note",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
},
{
"id": "646b411abc0c536ba1a98263",
"type": "whatsapp",
"displayName": "WhatsApp",
"defaultResponder": {
"id": "6419d5aad477770116cfa596",
"integrationId": "5f8ece75e2059d2c9995186b",
"integrationType": "zd:answerBot",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
},
{
"id": "6706411852c663af3215a44f",
"type": "web",
"displayName": "AI Agent (Ultimate)",
"defaultResponder": {
"id": "670646ef88ab3af5a96c49b1",
"integrationId": "670646eeec1612bbbc225995",
"integrationType": "ultimate",
"nextSwitchboardIntegrationId": "5f8ece7337f5bd000d6ae0ff"
}
}
]
}
```
# End User experience
The above is quite technical, but luckily for our end-user we end up with a nice experience with two very web widgets that respond with either a Zendesk or an Ultimate powered AI Agent.
### Filter custom ticket statuses per form.
URL: https://internalnote.com/form-based-custom-statuses/
Last updated: 2025-09-08T06:40:45.000Z
💡
This is a Spotlight article. A short article on a new feature that's not emailed to subscribers but will be part of the monthly [roundup](https://internalnote.com/tag/zendesk-roundup/). This lowers the amount of emails you get, while still get the information you want.
Last year Zendesk released [Custom Ticket Statuses](https://internalnote.com/tag/custom-ticket-status/), a way to expand the native new, open, pending and solved ticket statuses with custom statuses that better match your company's processes.
These statuses are useful to eg highlight that a ticket is on-hold while you wait for an external supplier, or pending until the customer confirms something. It allows you to add a different open status for tickets that got a reply, versus those that get an update with an internal note or side conversation.
While these custom statuses are quite useful, they do clutter up the Submit menu, especially if you have a lot of them.
Example of a submit menu with a lot of custom statuses
Not every custom status is useful in every scenario though. A support form has no use for a "waiting for quote feedback" status that a sales form might need. Or a form to request returns and refunds might need your "waiting for shipment" or "refund method requester" status, whereas your product form might need a "waiting for supplier" status.
# Ticket statuses by form
The new *Ticket statuses by form* feature in Zendesk will fix this issue. You can now select specific statuses to be displayed per form, reducing long lists of statuses to short, more specific options.
An overview of forms and custom statuses
This new options is added as an additional tab to the Ticket Statuses menu. (Although arguably, since you select a form and add statuses to it, this might as well be an item of the forms menu, similar to how you add ticket fields to forms in that menu)
When you select a specific form you can toggle all the statuses you want to appear in the menu. Note, you can **only** filter custom statuses, the native open, pending, on hold and solved statuses will always appear in the menu.
Similarly, even though existing custom status you created *before* this feature was launched will by default by enabled for all forms (you can deselect them if needed), custom statuses that are added *after* this release will be disabled by default on all forms. It would be nice if the process of adding a custom status had a step to pick your forms in the same flow, but alas, that's not possible for now.
Once you filter your menu, changes are applied immediately to the Agent Workspace.
The same ticket view now with a reduced list of statuses.
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Zendesk News Roundup for November 2024
URL: https://internalnote.com/roundup-2024-11/
Last updated: 2024-12-01T09:44:38.000Z
AI Agents. Agent Copilot. The new Zendesk Voice. Zendesk's AI Summit earlier this month turned out to be a giant release of new AI features across the entire Zendesk Suite.
If you want to read up on what's been announced, feel free to dive into my [AI Summit](https://internalnote.com/tag/ai-summit/) series before you read this month's roundup. Even though plenty of fun new things were released this month, I'm sure the really cool stuff is in the AI Summit articles.
When I started this blog three years ago I had a goal in mind to where I wanted to go with this platform. I've not yet checked of all items on my list, but earlier this week I did manage to cross of one **giant** one with a very cool share of one of my article by Zendesk CEO Tom Eggemeier on LinkedIn 🤯🤯🤯
[Tom Eggemeier on LinkedIn: AI Summit - the new Zendesk VoiceGreat piece on Zendesk Voice from Thomas Verschoren of Premium Plus - Zendesk's EMEA Partner of the Year Plus. He reinforces how essential it is to elevate and…LinkedInTom Eggemeier](https://www.linkedin.com/posts/tomeggemeier%5Fai-summit-the-new-zendesk-voice-activity-7257158138833903617-XcKm?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop)
Let's dive in the new stuff.
# 🏢 Company
### Increase in phishing attempts to Zendesk accounts
Let's start the roundup with an important notice: scammers have been actively to [phish Zendesk users](https://support.zendesk.com/hc/en-us/articles/8257723564186-Advisory-Increase-in-phishing-attempts-to-Zendesk-accounts?ref=internalnote.com) by emailing them trying to get access to their instance, pretending to be Zendesk.
🔒
Zendesk will ****never** reach out to you to change or ask for passwords in such a way.
My advice: enable 2FA on all your accounts, and hope Zendesk launches their [new two-step verification](https://support.zendesk.com/hc/en-us/articles/7955412933274-Announcing-two-step-verification-2SV?ref=internalnote.com) for unknown browser logins sooner rather than later!
If you're interesting in securing your Zendesk instance even more, take a look at the article below.
[Zendesk Security ChecklistOne of the big Trends for 2024 is the idea that security no longer is an add-on but should be seamlessly incorporated throughout the customer journey. To get started the right way, I’ve written a Zendesk Security Checklist. It’s a practical approach to improve the security of your Zendesk instance.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-security-checklist/)
### Admin Center revamp
The Admin Center got a nice update with a couple of new features on the home dashboard:
- **Zendesk updates**, a new place to discover all updates available on the platform. (Although, why go look for it when you can get them via email like this? 😇)
- **Feature usage**, a small dashboard that shows you how often you use triggers, automations and macros
This expand the previous updates to the Admin Center Home page with the availability of Storage usage, API usage and Automated resolutions status counters.
# 🎉 New Releases
## 🤖 AI Agents
### Multi Conversations
> Admins will be able to turn on multi-conversations for their account through the Admin Center and select the channels through which users can start new conversations.
> Agents will continue to see all incoming conversations linked to a user's profile, just as they do today, making it easy to view and address multiple conversations from an end user.
This is a very welcome release that's been a long time requested. The old [Sunshine Conversations](https://internalnote.com/history-of-zendesk-messaging/) widget used to support this feature, and I'm glad it's now available for all customers via the native Zendesk Widget.
If you want to get more information, take a look at my latest *Spotlight* article:
[Multiple Conversations in the Zendesk Messaging WidgetThe new multi-conversations in Zendesk Messaging allows for, well, multiple concurrent conversations with AI Agents and your team in the Zendesk Web Widget and Mobile SDK. This feature is now available to all customers.Internal NoteThomas Verschoren](https://internalnote.com/multi-conversations/)
### Meta pricing
> We have an exciting update from Meta regarding service conversations on the WhatsApp Business Platform. Starting **November 1, 2024, service conversations will be free within the 24-hour customer service window**. This is regardless of volume, instead of the current capped 1,000 free service conversations. (Thanks [Chat Inc](https://chatinc.com/?ref=internalnote.com) for letting me know)
You can find the full announcement [here](https://developers.facebook.com/docs/whatsapp/pricing/updates-to-pricing/?ref=internalnote.com), but this is a welcome change for those that use [WhatsApp notifications](https://internalnote.com/sunshine-conversation-automations/).
## 👨🏻💻 Agent Workspace
### Follower and CC conditions in Views
A few years ago I wrote a little app for [Premium Plus](https://premiumplus.io/?ref=internalnote.com) that made it possible to view Followed and CC'd tickets in views. That app became quite popular and filled a gap in Zendesk's features.
[Follower View | Premium PlusFollower View Adds a custom view which displays Followed, CC’d or Assigned tickets. Install now €0.99 / month / agent About the app Zendesk has a couple of ways to keep agents informed on whats happens to their tickets. When using CCs, you are allowing agents to copy in other…Premium Plus](https://premiumplus.io/integrations-apps/follower-view/?ref=internalnote.com)
Now Zendesk has finally *sherlocked* my app and made CC and Followers available as conditions for Views. Which means you can now create a view that contains all your followed, cc'd or assigned tickets in one overview.
Better late than never, but with the arrival of [Agent Home](https://internalnote.com/tag/agent-home/) and its native filters for these types of tickets, I don't expect lots of people to use this new view conditions.
### View categorization
Speaking of views, for those who are a fan of lots of views, you can now [categorize](https://support.zendesk.com/hc/en-us/articles/8043835835674-Announcing-the-ability-to-categorize-views?ref=internalnote.com) views similar to how you can group macros. You can leverage the `::` syntax to add a category to a Views' name, making those views appear grouped in the Agent Workspace.
For example, a view called *Topics::Active problems* will now show up as *Active Problems,* under a dropdown menu called *Topics*. It's a nice addition that fixes the issue of having a long list of views, but I do fear it will motivate companies to keep using lots and lots of views, instead of adapting a more focused Approach to Views
[My approach to Zendesk ViewsIn this article I explain my approach to Zendesk Views, and how you only need 8 views to make an efficient setup.Internal NoteThomas Verschoren](https://internalnote.com/my-approach-to-zendesk-views/)
### Form Based Ticket Statuses
Last year Zendesk introduced [Custom Statuses](https://internalnote.com/tag/custom-ticket-status/) as a way to expand the native new, open, pending, on hold and solved statuses we've always had in Zendesk.
It's a great feature that, similar to Layout Builder and Pinned Apps, allows you to make the Zendesk UI fit the way you work. With this latest release Zendesk now allows you to show specific custom statuses for specific forms, making the UI fit even better with how you work,
💡
I've got a [Spotlight](https://internalnote.com/404/) article planned on this topic for Thursday. So keep your eye on this blog if you want to learn more.
### Advanced AI updates
Following the releases at the [AI Summit](https://internalnote.com/tag/ai-summit/), Zendesk released some smaller updates to how some of the features now work.
- **Suggested Macros** now take the entire ticket conversation into account and don't only react to the initial ticket message. This means agents will get new suggestions throughout a tickets' life.
- **Similar Tickets** (currently in EAP) has an updated recommendation engine which means more related tickets should now show up.
- **Similar Tickets** now also does not require intents in Intelligent Triage to be enabled. It'll just work for every instance with Advanced AI.
### User suspensions for messaging
Agents can now [suspend](https://support.zendesk.com/hc/en-us/articles/8258324950042-Announcing-user-suspension-for-the-messaging-channel?ref=internalnote.com) users who misbehave in a Messaging conversation. Those users then show up in the Suspended view in the Customers Tab of Agent Workspace, and can be unsuspended as needed.
Do note, if a guest user that isn't authenticated clears they browser cache they can reconnect since they'll show up as another visitor on your instance.
## 🔎 Copilot Autoassist and Help Center
### Shared Media
Zendesk's work on the Media library in Zendesk Guide keeps moving forward month by month. Where last month gave us localized assets and the option to upload assets via API, we now get the ability to share assets across editors
To be honest, I didn't even know this wasn't possible before.
I do wonder when we'll get a Media library link in the Guide sidebar though. This feature is getting quite powerful and it's a pity it's hidden in the article editor this way.
### Redirect API
This API has been in EAP for a *long* time and is now finally available to everyone. Quite useful to improve your SEO score after a migration to Zendesk, or to create vanity URLS that point to specific articles.
[Redirect Rules for Zendesk GuideUse the new Redirect Rules API for Zendesk Guide to redirect customers to other pages instead of finding a 404 error for non-existing urls.Internal NoteThomas Verschoren](https://internalnote.com/redirect-rules-for-zendesk-guide/)
### Article Multiplacement
To conclude the Help Center and Copilot section of this roundup, we get the new [Article Multi placement](https://support.zendesk.com/hc/en-us/articles/7867489163930-Placing-articles-in-multiple-sections-with-article-multiplacement-EAP?ref=internalnote.com) feature that allows you to publish one support article across multiple Help Centers or section.
This is a very useful feature for scenarios where you have one article (for example: how to update your device) which applies to multiple products (phone and tablet). Or if you have an "delivery options" article that applies to both your consumer and b2b brands.
## 🧱 Open and Flexible Platform
### OpenID Connect
OpenID connect is now supported as a [single sign-on](https://support.zendesk.com/hc/en-us/articles/7957465432474?ref=internalnote.com) option for end-users.
### User Deletion experience and bulk Importer
Zendesk has been gradually moving stuff in- and out of the Admin Center and into other places of the platform to make everything feel a bit more cohesive.
Copilot procedure management for example has been moved into Zendesk Guide's Admin Center so that all Knowledge Management features are now in one place (I kinda expect Macros to be next to be honest)
Similarly, they've been moving all day-to-day end-user and organization management out of the Admin Center and into Agent Workspace under the new Customer and Organization sections. You can use the new Bulk delete users option there to remove [those suspended users](https://internalnote.com/automatically-deflect-zendesk-spam-tickets-via-triggers-and-web-hooks/) in bulk.
The old and new interface for Deleting users
On the opposite side, the ability to import users in bulk has also been updated, but this feature remains in Admin Center as part of the new bulk importer that already supported organizations and Custom Objects.
### Refresh of the Routing configuration page
Omnichannel Routing is – in my opinion– one of Zendesk most powerful and most complex additions in a long while. The new assignment rules that take skills, queues, availability and a bunch of other features into account allow for a lot of customizations and can adapt to almost any workflow or assignment approach.
But with great customizability comes great complexity and the admin panel to manage those routing options starts to look more like something straight out of Windows 7 with multiple nested options, checkboxes and drop-downs.
This month Zendesk reorganized the entire menu and made it a lot cleaner to look at, even though all the features and setup complexity remain as is. Although I do like the division between "Global Settings" and more specific grouped email and messaging settings.
I do wonder what will happen once [Zendesk Voice](https://internalnote.com/ai-summit-the-new-zendesk-voice/) gets added to the mix. Do we use queues to route voice calls? Or will this menu get even longer? Time will tell.
Image taken from the Zendesk Support article.
## 📊 Reporting and Insights
### Customizable CSAT updates
Last month Zendesk launched their new [customizable CSAT](https://internalnote.com/preview-of-the-new-customisable-csat-for-zendesk/) and with it expanded the native thumbs up/down with more options.
Initially the emails send out via this new feature contained a simple *Share Feedback* link, but since then these emails have been expanded with built-in options for the customer to click on, immediately collecting their choice when they do so.
You can now also choose if you want to link to the original ticket (which opens in the customer portal section of your help center), or not.
```html
Hello {{ticket.requester.name}}, we’d love to hear how we did on your recent request . If you’re up for it, please take a moment to answer a few questions. We really appreciate it.
{{satisfaction.survey_section}}
```
Legacy CSAT, EAP Customizable CSAT, new experience
# ⚠ Major Changes
There's a lot of small but important changes to Zendesk this month it seems.
## Automatic Resolutions
Automatic Resolutions are Zendesk's new way of charging for the work their AI Agents do. As mentioned in my [AI Summit - Omnichannel Agents](https://internalnote.com/ai-summit-ai-agents/) article, you pay for when the bot actually deflects tickets, and don't pay when your agents get involved in the conversation.
It's a fairly clean way to keep the cost of agents and AI agents in balance, and your usage can be seen via the new status widgets on the Admin Centers' homepage.
And yep, if you the reader uses my demo bot flows, I get charged for that via AR :-)
If you want more insight and maybe exclude tickets handled by an AI Agent from your reporting in Explore you can use [tags to filter those tickets out](https://support.zendesk.com/hc/en-us/articles/8035023504666-What-changes-on-my-account-when-I-migrate-to-automated-resolutions?ref=internalnote.com). These tasks used to start with `ab_` (Answer Bot), but with the move to Automated Resolutions, these will now start with `ar_`
- `ab_marked_unhelpful` changes to `ar_marked_unhelpful`
- `ab_resolved` changes to `ar_marked_helpful`
- `ab_suggest_true` changes to `ar_suggest_true`
- `ab_suggest_false` changes to `ar_suggest_false`
## Login option changes
> Zendesk Chat will no longer offer end users the option to fetch their name and email address from Google or Facebook to identify their name and email address to businesses. This change only applies to end users who are not using Google or Facebook to sign in; end users who are currently signing in using this method will still be able to do so.
Zendesk Chat is slowly fading away, one login at a time it seems. [More info here](https://support.zendesk.com/hc/en-us/articles/8157106527258?ref=internalnote.com)
> On November 18, 2024, we’re removing X (formerly Twitter) as a sign-in option for end users.
Twitter on the other hand is [very quickly](https://support.zendesk.com/hc/en-us/articles/8176344568218-Announcing-the-removal-of-X-formerly-Twitter-as-a-sign-in-option-for-end-users?ref=internalnote.com) going the way of the dodo.
> Historically, developers using ZIS could reference the access token of an OAuth connection in a flow or custom action using the path `$.connections.{oauth_connection_name}.access_token`.
> When Zendesk announced support for additional authentication methods in ZIS connections, we introduced a new way to make use of connections in ZIS actions with the `connectionName` property.
> Until now, both referencing methods were supported. Going forward, bundles can no longer be uploaded or updated if they contain an OAuth access token reference path.
Out of interest dear reader, have you used [ZIS](https://support.zendesk.com/hc/en-us/articles/8263007584026-Deprecation-of-OAuth-access-token-reference-path-in-ZIS?ref=internalnote.com) already? I'm so used to use Cloudflare Workers to automate my Zendesk instances that I haven't really dived into ZIS yet..
## Incremental Export API Change
> Until now, the Incremental Export API attempted to include data up to the moment of the request. However, this can lead to data inaccuracies, so we're changing this behavior. Now, all ticket creation and updates that occur in the minute preceding the request are excluded.
> The API will account for this in the recorded end time, so it can pick up seamlessly where it left off for subsequent requests.
Small but important update for those exporting data.
# 💡Insights
## Cards Against Complexity
Jan from [Next Matter](https://internalnote.com/sponsor-next-matter/) invited me on his Cards Against Complexity podcast to talk about the Zendesk AI Summit, Internal Note and automation in Zendesk. I'm not used to be on camera but it turned out a fun conversation!
## AI Summit Deepdive
## Getting started with Zendesk AI
Nice article by Jeff Adkins on getting started with Zendesk AI.
[Getting Started with Zendesk AIA Starter Roadmap for Admins and Support LeadersConnective Intelligence: AI, CX, and BeyondJeff Adkins](https://jeffadkins.substack.com/p/getting-started-with-zendesk-ai?r=328dv4&triedRedirect=true)
# 📝 Articles this month
[Multiple Conversations in the Zendesk Messaging WidgetThe new multi-conversations in Zendesk Messaging allows for, well, multiple concurrent conversations with AI Agents and your team in the Zendesk Web Widget and Mobile SDK. This feature is now available to all customers.Internal NoteThomas Verschoren](https://internalnote.com/multi-conversations/)
[A History of Zendesk chat and messaging channelsOver the years, Zendesk evolved from simple email ticketing to offering a full suite of customer care solutions. They now support multiple channels, AI agents, social integrations, and web widgets. This article will dive into how their complex platform works, helping navigate conversational setups.Internal NoteThomas Verschoren](https://internalnote.com/history-of-zendesk-messaging/)
[AI Summit - Internal NoteZendesk presented their latest releases at the AI Summit in October 2024\. New announcements include Omnichannel AI Agents, the new Zendesk Voice, Agent Copilot and improvements for Zendesk QA.Internal Note](https://internalnote.com/tag/ai-summit/)
[End-User access for Custom Objects and Lookup Fields (EAP Preview)Zendesk’s Custom Objects, released last year, enable you to expand platform data, like linking assets or contracts to tickets. Initially, end-users couldn’t interact with custom objects, but a recent update allows you to add Lookup Fields in forms.Internal NoteThomas Verschoren](https://internalnote.com/end-user-access-for-custom-objects/)
[Using Messaging triggers: Customer Wait Time & Messaging CSATZendesk’s new Messaging triggers enable automations similar to email tickets. You can notify customers of wait times, request CSAT feedback, or let them close a conversation. These triggers enhance the customer experience by allowing more customized interactions within Messaging.Internal NoteThomas Verschoren](https://internalnote.com/messaging-triggers-customer-wait-time-and-messaging-csat/)
# And Finally...
Normally I end my roundups with a fun bit of Zendesk Trivia I discovered.
This week, I'm going to use this section to announce a few structural changes on the blog in the next few weeks.
I've managed to write and publish an article on Zendesk every Tuesday for the last two years. Since Zendesk's release cycle has been so fast this year, I'm currently running behind on writing about major topics like Zendesk QA, Ultimate and ZIS, while having article ideas to write about well into 2025 already.
To make sure I can write with quality about Zendesk and to not overload your inboxes with emails, I'm going to introduce the concept of [*Spotlight*](https://internalnote.com/404/) articles. These are short articles that will solely be posted on this website that highlight interesting new releases as they happen.
These articles will not be emailed to you, but will be added to each months' [roundup](https://internalnote.com/tag/zendesk-roundup/) so you can dive into them if they seem of interest. The first one was published last week, and more will be published on Thursdays if something cool is launched.
Secondly, I'm changing access to the content I'm writing. I've made it a point of making all content available without a paywall for everyone who subscribes to this blog since I'm a firm believer that knowledge and insights are only valuable if they are shared.
I will still do that but I'll start publishing some articles behind the [Internal Note Plus](https://internalnote.com/#/portal/signup/63f0d0f4034c3d004d5ef75a/yearly) subscription initially, making them publicly available to everyone after about a month's delay. This way I can hopefully convince more reads to support this project by becoming a Plus member, while still making content available to everyone.
My new monthly schedule
If you *really* like what I do, can I ask you to consider subscribing to Internal Note Plus via my [monthly](https://internalnote.com/#/portal/signup/63f0d0f4034c3d004d5ef75a/monthly) or [yearly](https://internalnote.com/#/portal/signup/63f0d0f4034c3d004d5ef75a/yearly) tier?
Thanks!
### Multiple Conversations in the Zendesk Messaging Widget
URL: https://internalnote.com/multi-conversations/
Last updated: 2025-09-08T06:44:07.000Z
💡
This is a Spotlight article. A short article on a new feature that's not emailed to subscribers but will be part of the monthly [roundup](https://internalnote.com/tag/zendesk-roundup/). This lowers the amount of emails you get, while still get the information you want.
When interacting with the Zendesk Messaging widget there's always been a few frustrations that customers mention. One of them is that the Messaging widget was, by design, a single threaded conversation. Meaning as a customer you can only talk to either an AI Agent or a human agent upon escalation. While that agent is handling your question, you can't start a second one to ask another question.
This might seem like a small issue, but imagine you've got a support ticket raised for a lost package, and it takes a week or two to get it closed. In the meanwhile, each time you open the Messaging widget, you'll see your conversation with its latest updates.
During those two weeks you have another question you want to ask on a different topic. As an end-user your only available option is asking it to the agent handling your package issue, since there's no way to reach the AI Agent unless the original ticket is closed. Where an AI Agent would give you your answer in seconds, you now need to wait for an actual agent.
As that actual agent, things also get annoying. Your now open to any kind of question from that customer since your first line of defense, the AI Agent is down until you close that ticket for that customer. If the customer reaches out with questions for other departments, or for items that require different skills, you're kinda stuck. You need that package ticket assigned to you to resolve it, but need to pass the other questions the customers asks to team members.
Single threaded conversations can get annoying fast.
# Enter multi-conversations
This is where the new multi-conversations for Messaging widget (and sdk) come in. This upgrade to the Messaging experience upgrades the widget to allow for more than one conversation.
Whenever a customer is in an active conversation, they can click a *back* button in the widget. This opens the new conversation view showing them a list of all conversations they had with you.
They can then choose to start a new one, or jump to another conversation.
This works for both authentication and non-authenticated users. However, when a browser session expires or local cache is wiped, your unauthenticated users will get greeted with an empty widget again, just like with the old experience and can then start another conversation (or more than one).
For authenticated users the widget will restore **all** their conversations upon login.
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging/)
## Proactive Messages
If you greet your customers with proactive messages, a customer interacting with them will automatically be routed to a new conversation, as to not interrupt any existing flows.
[An in-depth overview of Proactive Messages for ZendeskDiscover the new Zendesk proactive messaging. Learn about the main features and advanced flows, including the ability to show proactive messages based on specific marketing campaigns. Target customers based on their locale, or offer a premier experience to VIP users.Internal NoteThomas Verschoren](https://internalnote.com/proactive-ticketing-for-messaging/)
# Activating multi-conversations
Activation of this new feature is done via the Admin Center. If enabled it's enabled for **all your brands and channels** **and can't be disabled anymore.**
But given the benefits of this feature, chances are rare you'll want to disable this.
You can however disable the ability to create new conversations. For unauthenticated users this means they'll always have one conversation, except when you send them proactive messages. For authenticated users this means they still see a list of all conversations, but can't create a new one unless the existing one is resolved.
# Agent Experience
For agents nothing really changes. Your customers will still create the same conversations which might get escalated to your team. However, the annoying "customers goes off topic since they can't raise a new ticket" issue is gone, and agents can actively ask customers to create a new ticket if they happen to go off topic.
There's one important note though. Since we've always had a single conversation per customer until resolution, and we now allow for multiple concurrent conversations you will see an increase in customer interactions. And since conversations can now be ended via the new End Session button, and customers are more inclined to start creating multiple conversations per topic, your average conversation length (and thus resolution time) will go a bit down.
One question I do have, is one about routing. If a customer starts two concurrent conversations for different topics, would they expect to be routed to the same agent? Or would they just ignore the fact that their two conversations might be escalated to different agents?
# A few caveats
- All older Mobile SDK versions do not support this feature. You'll still see the single conversation view there. So best to move to the latest version via developer.zendesk.com (If you already use the Zendesk SDK this should be a *simple* swap of the embedded SDK to the newest version)
- The [release notes](https://support.zendesk.com/hc/en-us/articles/8195486407706-Understanding-multi-conversations-for-messaging?ref=internalnote.com) also mention a lack of support for V1 SunCo APIs. This theoretically means you can't auto-post [messages](https://docs.smooch.io/rest/v1/?ref=internalnote.com#conversation) like I showed in this [article](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/). But I've rarely seen people use this over messaging...
- Social channel linking (allowing an end user to move the conversation from the Web Widget to a social channel) is only available for the first conversation started by the end user. End users can’t continue subsequent conversations on social channels.
# Conclusion
My one line conclusion: Go enable this. Today. There's zero downsides.
### Using Messaging triggers: Customer Wait Time & Messaging CSAT
URL: https://internalnote.com/messaging-triggers-customer-wait-time-and-messaging-csat/
Last updated: 2025-09-08T06:40:50.000Z
When working with Tickets in Zendesk you can use triggers to automate interactions with customers. You can send out a confirmation email when a ticket is received, you can [send out reminders](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/) for pending tickets, or ask the customer for CSAT via an automation.
Doing the same for Messaging tickets however is tricky. Not every action you're used to do for email tickets can be replicated with the regular Zendesk triggers.
Back when Zendesk offered Zendesk Chat, you had a specific section to create triggers for chat conversations. You could do things like send a welcome message, remind customers of a lack of replies and so on.
After the migration to Messaging a lot of these capabilities were lost, and it's only with the recent release of Messaging triggers that you can now recreate a lot of these flows again.
This article will give show a few flows that you can built with these triggers.
# Showing customers their Wait Time
Thanks to a recent release Zendesk now makes it possible to tell customers their [estimated waiting time](https://support.zendesk.com/hc/en-us/articles/8009787999514-Displaying-estimated-wait-time-in-messaging-conversations?ref=internalnote.com) when they want to talk to Agent. This way customers know what to expect and don't get frustrated by an unknown waiting time.
💡
This first example borrows from the Zendesk support article but I hope to highlight some additional details in these steps.
## Setting it up
To set this up you can create a new messaging trigger with the following steps:
As for the Responder Message, you can use any text as long as it contains the `@wait_time_min` and `@wait_time_max` placeholders.
> Your estimated wait time is @wait\_time\_min-@wait\_time\_max mins.
After you setup this trigger a customer will – upon escalation – immediately be informed of the estimated wait time, based on the amount of customers in the queue.
💡
During my testing I noted that this performs best if you have a good [queue or Omnichannel Routing](https://internalnote.com/an-intro-to-omnichannel-routing-in-zendesk/) setup, since the time is calculated based on tickets being routed. If you let agents cherry pick their conversations than the estimated number might be way of.
Once the queue empties and the customer can be passed to an agent, we can tell the customer it's their turn before handing it over to the agent via second trigger:
You might notice this trigger has a *Group status* and *Group* condition. If you route your conversations to more than one group, you need to create a notification per group in your instance. This way the trigger will **only** take that group (and its agents) into account for calculating the queue length, and you don't send timings to customers who do not need to wait. So if you have three customer care teams, you'll end up with three *estimated time* triggers, and three *it's your turn* triggers.
## End-User Experience
For an end-user the experience is quite nice. The bot sends them a message telling them about their waiting time, tells them it's their turn and then passes it to an agent.
## Customize with adding queue time
One issue I notice during testing is that in scenario's where the queue is short, with only one or two people in a queue, the "*It'll take 1-3 minutes*" message and the "*It's your turn!*" message can appear immediately one after the other.
To prevent this weird behavior, you can add an extra step to your conditions and check for queue length. In my setup for example I only trigger the Estimated Time alert if the queue is larger than three people.
# Asking for CSAT in Messaging
At the end of a customer interaction you can ask your customers to provide their feedback via a feedback form. With email based tickets we can use an automation to send out an email after a short delay. And now, with the messaging triggers, we can send out a similar notice to customers directly within the Messaging conversation.
This new capability integrates nicely with the new Customizable CSAT for Zendesk which launched earlier this month.
[Preview of the new Customisable CSAT EAP for ZendeskThe new Customisable CSAT EAP for Zendesk has arrived, finally allowing you to change your rating scale, choose emoji, numbers or labels, and customize your follow-up questions. This article contains an initial overview of the new feature, and shows how it works with existing API integrations.Internal NoteThomas Verschoren](https://internalnote.com/preview-of-the-new-customisable-csat-for-zendesk/)
## Setting up Messaging CSAT
Setting up Customer satisfaction for Messaging happens within the new Customer satisfaction Business rules section of the Admin Center.
You can activate the Messaging trigger, and drill down to its settings to tweak settings like the channels where we should ask for CSAT, or you can filter out specific intents, customers or tags.
## Customer Experience
For customers this integrates nicely with the existing Messaging flows as a new modal view that pops up within the Zendesk Widget. When triggered in social channels this can either popup as a similar modal if supported, or will show a button that links to a webpage to leave their feedback in a similar layout.
💡
Fun fact (for me at least): this is the first example of a first-party [Conversation Extension](https://developer.zendesk.com/documentation/zendesk-web-widget-sdks/conversation%5Fextensions/?ref=internalnote.com) I've seen in use in the Zendesk Widget.
# Allow customers to close a ticket
Earlier this month Zendesk launched the capability for Agents to end a conversation and triggering CSAT and a return to the Zendesk Bot for the customer.
A request I hear a lot is the capability for customers to do the same. Sometimes you just want to close the conversation and begin anew with a new topic. Especially in scenarios where the issue is resolved, and you're stuck waiting for an agent to close the ticket before you can talk to the bot again.
By leveraging Messaging triggers we can *hack* this capability within Zendesk.
## Setup
First off, we need to create a Messaging trigger that looks for specific keywords like *Close Conversation*. You can add one or more variants in the *ANY* sectionof the conditions if you want.
We then tag the conversation with a specific tag like `close_conversation`.
Next we need to create a trigger that looks for updated tickets in the Messaging channel that contain that tag, and set the status to solved.
This will trigger CSAT, after which the customer goes back to your bot!
💡
We can not do this with regular ticket triggers since those triggers can only "read" email contents and not messaging replies.
# Conclusion
The above gives you a taste of what's possible with Messaging triggers. And when you dive into the options you can send messaging when tickets change groups, you can send reminders when a customer waits to long to respond, or you can tag tickets with specific keywords to change priority or routing.
What are you building next?
### End-User access for Custom Objects and Lookup Fields
URL: https://internalnote.com/end-user-access-for-custom-objects/
Last updated: 2025-09-08T06:41:07.000Z
When Zendesk released their [Custom Objects](https://internalnote.com/tag/custom-objects/) last year, they made it possible to expand the existing ticket, user, organization objects in the platform with any data type you needed to enrich your ticket and agents experience.
You can use custom objects to store assets like computers or printers, you can store service contracts or purchased licenses, or you can use it to build fun demo's and store [movies](https://internalnote.com/custom-object-placeholders/) or [Pokémon](https://internalnote.com/creating-a-pokedex-with-zendesk-custom-objects/) within the platform.
Custom objects can be linked to tickets via Lookup Fields and, when linked, will show more detailed detail right next to a ticket when an agent handles it. When linking – for example – service contracts you can see its duration and services within the contract. Or when linking an employees laptop you can see its type, operating system and warranty duration.
One element that was missing was a way for end-users to interact with custom objects. Where agents could fill in lookup fields in the ticket sidebar, you couldn't make these same fields available to end-users, like you can with regular drop-down or text fields.
In part 4 of my Custom Objects series last year I showed a workaround leveraging the Zendesk API, an external worker and modifications to the Help Center. But luckily, with Zendesk's latest [Custom Objects release](https://support.zendesk.com/hc/en-us/articles/8286928528026?ref=internalnote.com), this can now be done natively!
[Zendesk Custom Objects - Part 4: End-User and FormsThis is a four-part series on Zendesk’s new Custom Objects feature. The articles cover setup, data import, using Custom Objects in forms and with agents, expanding user profiles, and displaying Custom Objects in Help Center forms.Internal NoteThomas Verschoren](https://internalnote.com/custom-objects-part-4-end-user-and-forms/)
# What's possible now
Mid summer Zendesk released the ability to use placeholders in macros that pulled data from Custom Objects. To demonstrate that capability [I created](https://internalnote.com/custom-object-placeholders/) a Custom Object to store movie details. If a visitor of my movie theater showed interested in a movie, we could use those placeholders to reply to their email with more information.
In that article I purposefully used email as an example to work around the fact that there is no elegant way to ask for that information in forms. Or at least not while filling in the ticket field that contains the movie without resorting to custom code.
Now, with this new Custom Objects for the Help Center release we can take an existing or new Lookup Field and give permissions to an end-user to view or edit that field from within a form or request page.
This way, I can ask a customer to select a movie, a trainer to select a Pokémon or an employee to select a laptop from a nice list.
You might wonder what the benefit of these kind of Lookup Fields is compared to regular drop-down fields. For one, Lookup fields can be filtered and only show those records that apply to the current user, organization or other condition.
Secondly, even though the customer still sees a list of options, your agents can see all related data to a customer's choice (warranty date, SLA type, movie genre,...) giving them more context when working on tickets. And the record editor in Agent Workspace is a lot nicer to work in than the dropdown editor in the Admin Center.
# Setup
💡
This feature requires a Help Center theme running on the latest version 4 APIs. You get this update automatically if you run an unmodified Copenhagen theme. If your theme is customized, reach out to a developer or download a recent version of a new Zendesk theme. More info [here](https://support.zendesk.com/hc/en-us/community/posts/8043398017306-Which-Guide-theme-am-I-in-and-how-to-move-to-the-latest-theme?ref=internalnote.com).
Making Custom Object data available to end-users is a three step process.
## Create a custom object
This first step is obvious, but in order to make data available to the end-user, you should create a Custom Object type in the Admin Center, and add data via the Record editor in Agent Workspace, or by bulk importing data via the Admin Center.
## Create or Update a Lookup Field
The next step is adding the Custom Object to your forms. If you previously created a Lookup Field for your agents, you can update its settings to make them viewable or editable to end-users.
If you haven't yet, create a new Lookup Field and make sure to select either the *Customers can edit* or *Customers can view* option*.* When doing so an alert will appear underneath the options to update your custom object with view permissions for end-users. Do so.
Just like with any ticket field you can give the field an end-user visible name and description.
💡
End-users are only able to **view* custom object data. They can select an object via a lookup field to link it to the ticket, but they are not able to modify the objects themselves.
## Add your lookup field to a webform
As a final step, you can update your webform to add your new Custom Objects Lookup field.
If you already created an *Agent Only* field in the past, you can skip this step.
# End-user Experience
When a customer now opens our Movie form they see a new dropdown to pick their movie. They get a dropdown of all available movies (pulled from our custom objects) and once they submit the ticket, this choice will be visible to Agents in the Agent Workspace.
# Migrate from old article code
If you previously used [my workaround](https://internalnote.com/custom-objects-part-4-end-user-and-forms/) to get Custom Objects working for end-users, you'll need to take the following steps to undo your work and enable the native solution:
1. In the admin center:
1. Enable end-user edit rights on your Lookup Field
2. Delete the intermediate field we created
3. Delete (or deactivate) the trigger that copied the value from the intermediate to the lookup field
2. Go to your Help Center theme and open the `new_requests.hbs` page. Remove the javascript block we added to detect the form and intermediate field, as well as the code that gets the data from the external worker
3. Remove the external worker and remove the API key we used to make it work.
# Filters
One final features that's pretty cool with these new Lookup Fields capabilities are Filters. When linking a Custom Object to a Lookup Field you can filter the data shown in the dropdown.
If you have, for example, a single Object type for Assets containing laptops, phones and printers, you can use the Filter *type | is | laptop* to only show laptops in your "Select a laptop" lookup field
Or, if you have a list of Pokémon in your custom objects you can link each Pokémon to a trainer by adding a Lookup Field to the Pokémon object that allows you to select a trainer.
If you then create a ticket Lookup Field "Select a Pokémon" you can use the filters to set *Trainer | is | requester* When a logged in user then uses your contact form they will only see those Pokémon that belong to him. (Or in a more realistic scenario, an employee will only see their devices)
# Limitations
As always, even though this feature is a *major* release that will unlock tons of new flows and possibilities, there a few things I'd love to see happen with Custom Objects:
- You can't make Lookup Fields that link to organizations or users accessible to end-users. I assume this is for security reasons as to not accidentally leak your entire customer base in a dropdown, but for e.g. scenario's where you want to select a vendor, partner or manager, having those kind of fields would be useful.
- Lookup fields are currently not accessible within Zendesk Bot flows in the Ask for Details step. I'd love to see some feature parity there too!
### AI Summit - the new Zendesk Voice
URL: https://internalnote.com/ai-summit-the-new-zendesk-voice/
Last updated: 2024-11-05T10:00:31.000Z
Welcome to this fourth, and final, installment in my AI Summit series, discussing all the new announcements for the Zendesk platform. If you missed the previous issues, you can find them [here](https://internalnote.com/tag/ai-summit/).
When you look at the Zendesk Suite overall there’s a plenty of of elements that have gotten a lot of attention these last few years. Messaging, Bots, Agent Workspace all got major new features and designs. These changes impact the way Zendesk works in major ways –
The move to messaging turned direct web chat into an asynchronous channel that offers better self service solutions via bots and generative replies. It expanded the traditional chat channel to both web, mobile and social channels and make sure your customers can reach you from anywhere.
Similar, the new Agent Workspace with its omnichannel routing, agent home and context panels give agents a rich environment to handle tickets with context, automations and suggested replies to make ticking handling faster than ever before.
Even Zendesk Guide received some love with a new article editor, generative replies, tone shift and simplify AI features and – soon – support for Custom Objects, cross-brand articles and many more updates.
If you know you’re Zendesk Suite, you know where this is leading. One piece of Zendesk has always felt like the one piece of Duplo in a box full of Lego. And that’s Zendesk Talk. It’s Zendesk built-in voice offering that allows you to receive calls within Zendesk, route calls via a traditional “press 1” IVR to the right group, and it gives you some basic call recording and dashboard features.
But, compared to powerful third party voice solutions like [Babelforce](https://internalnote.com/sponsor-babelforce/), Aircall and others, Zendesk Talk felt like Duplo to me. Yes you can built cool buildings with it, but if you want details, more colors and cooler constructions, you should move to Lego. The same goes for Talk. It works. But it’s crude compared to more modern offerings.
Introducing**, the new Zendesk Voice**
> The relaunch of native voice with end-to-end Al across the entire call journey
At the AI Summit Zendesk mentioned the new Zendesk Voice features spread across the AI Agent, Agent Copilot and QA sections of the talk. Combining all new features together, it’s a complete upgrade of the Talk we know, with powerful new additions that will, later this year, bring it into this AI powered age of CX.
The product announcements spread across 5 elements:
1. Native voice upgrades
2. Al agents for voice, powered by PolyAl (Q4-24)
3. Copilot for voice (Q1-25)
4. Post-call transcripts & summaries
5. Voice QA for native voice + partners
Coincidentally this aligns nicely with the existing solutions available for text based channels like Messaging and email. Each part of the call journey gets improved, and where possible empowered with AI.
A typical call journey. (Image provided by Zendesk)
Before we dive in, first some context. Even though conversational interactions and *chatbots* are all the hype these days, most existing customer interactions still happen over – so called – traditional channels like email and voice. So when you think about automating not only your frontline but also assisting agents, improving automation and deflection rates for voice calls is an important factor.
# Incoming Calls - Al agents for Voice
If we go over the announcements in the order of a normal call journey we need to start with how an incoming call is handled.
Traditionally, for voice channels, this means a prerecorded message welcome in the customer, an announcement that the call may be recorded, and than an IVR asking the customer to press a series of buttons to reach the right team.
IVRs, similar to traditional flow based chatbots, are not that great. A customer makes a guess at each step of the route, hoping to end up with the right team, while fearing they will get disconnected or reach a dead-end with an automated “resolution”.
Similar to how these flows were improved (or fixed) with new automation flows powered by intents and generated replies for messaging, the new AI agents for Voice will remove the needs for static IVRs and turn conversations into dynamic and automated flows.
When a customers calls your number they’l be greeted with an automated response that asks them why they’re calling. The AI Agent then processes the response and can then do a couple of things to help the customer or agent.
For all inquiries the system can result indexed Help Center articles, websites and other product information to generate a *spoken* response to the customer. For more complex inquiries like an order status the AI Agent can ask the customer for an order number and then request the information from your order management platform over API. Similarly, when a customer wishes to book a table or change a reservation, the AI agent can ask for the required date, group size and preferences and execute the booking directly within the call.
And for scenario’s where we do need your actual human agents to take over, we can use this AI Agent the same way we can do it for messaging conversation: we can ask the right questions up front so the agent immediately has the customer id, order number or product type at hand when the call gets transferred.
AI Agent for Voice
So far this all seems like a great solution for a problematic channel. Voice, as opposed to email and messaging is a real 1:1 channel. Where an agent might be able to handle 2-3 messaging conversations at once, and can handle dozens of email threads due to the real asynchronous behavior of those channels, a voice call is live. You’re talking to the agent now and you can’t just put them on hold to talk to another customer at the same time.
So the only way to lower waiting times for customers is to either lower the amount of calls coming in, or by reducing the duration of each call, allowing an agent to handle more customers in a shift.
> With AI-powered voice assistants, over 50% of inbound calls are resolved autonomously, significantly reducing the load on human agents.
Both scenarios can be handled by those new AI agents for Voice. By deflecting conversations with self-service solutions like generated responses or flows that get integrate with backend systems and automate processes, you can lower the amount of calls that require an escalation to an agent. And by collecting context, you can make sure the agent can dive right in, instead of asking first for the what, who, why. This additional context is stored in the new [custom entities](https://support.zendesk.com/hc/en-us/articles/6711181959194-Automatically-detecting-unique-information-in-tickets-EAP?ref=internalnote.com) for Zendesk Intelligent triage.
Combining this with a real omnichannel strategy where you also try to push customers to other channels like messaging or email, gets you to a point where your voice channel can really shine for those complex use cases where its often needed.
## PolyAI
Personally I was surprised when I heard Zendesk had built a new Voice automation solution. All their recent purchases in the last years where focusing on intent models (Cleverly.ai) or ticket automation (Ultimate.ai), and the fact that the Zendesk Talk stack is largely based on Twilio behind the scenes doesn’t make it prime for in-house AI developments on voice. Or at least, not to my limited knowledge of how the Zendesk product teams operate.
So, when they announced at the AI Summit that this new AI Agent for Voice was to be powered by [PolyAI](https://poly.ai/?ref=internalnote.com), a leading voice automation platform, it kinda makes the above logical again. Zendesk [invested](https://globalventuring.com/corporate/information-technology/zendesk-ventures-terry-evans-jr/?ref=internalnote.com) in PolyAI earlier this year, so a deeper partnership between the two companies is kinda logical. Zendesk gets a powerful new partnership that turns their voice offering to eleven, and PolyAI gets access to a big established user base that is used to the concept of AI Agents and already pays for a voice channel as part of the Zendesk Suite offering.
This does turn AI agents for Voice into another add-on though, similar to how the Advanced AI add-on adds more capabilities to Agent Workspace, and Ultimate replaces the native AI Agent solution with a more powerful solution.
The only thing that’s not really clear right now is how this AI agent for Voice will be offered. Will this be a product on Zendesk paper – *a Voice Bot-add-on* – or will this be a tight integration with PolyAI – to be billed separately – time will tell.
# Escalation - Intelligent Routing
When the AI Agent can’t resolve the incoming call itself, it’ll need to escalate the call to an agent.
With Zendesk Talk this means setting up routing to a specific group of agents within Zendesk based on the number called, or a specific IVR choice. Calls are put in a queue and the first available agent in a group (or the first one that picks up the call) handles the conversation.
But similar to how messaging and email are moving from group-assignment towards [agent-routing](https://internalnote.com/an-intro-to-omnichannel-routing-in-zendesk/) with assignment based on skills and availability, the new Zendesk Voice will do the same. While the AI Agent is handling the conversation, it’ll also log key elements like intent, sentiment and language, and this information is used to route a call to the right agent when an escalation is needed. This assignment will be based on the same skills and availability settings that are now available for other channels like messaging.
As for how this routing works, from what I was able to piece together from Zendesk documentation, basic omnichannel routing will be possible based on tags in the IVR (and thus skills).
Deeper routing based on intents and sentiment can be achieved by using Advanced AI, or you can leverage AI Agent for voice and inherit the intent from its initial conversation with the customer.
There's lots of moving pieces in this, so you can be sure, once I get access to everything, I'll do a big[ Omnichannel routing](https://internalnote.com/an-intro-to-omnichannel-routing-in-zendesk/) for Voice update!
# During the Call
## Sentiment & Intent and summarization
During the call Agents will have the same context available as they would have for all other Zendesk channels. The information gathered for routing (Sentiment, Intent, Language) is displayed in Agent Workspace, and the knowledge panel will provide relevant answers based on the content of the conversation.
Agents will also see a summary of the conversations routed through the AI Agent so they can get quickly up to speed with the customers’ needs, without needing to read through a transcript of the entire conversation.
These features are currently not as *coming soon*, so we can expect to see them later this year, or even early next year.
## Copilot for voice
During a call things might change. A customer might call asking about a product feature, and might end up asking for an exchange since their product does not have that feature and they’d rather purchased another product. Or customers inquiry about an existing reservation and end-up upgrading to VIP tickets.
Whatever the scenario, during a call the context and intent of the conversation will evolve continuously, and agents might need to dive into multiple processes.
This is where the new Agent Copilot for Voice comes in. Announced for early next year it will continuously provide agents with new suggested actions on the ticket by listening for new intents, and than offering quick replies in the Knowledge Panel which agents can use to reply to customers.
There’s no word yet on other auto-assist Agent Copilot features as we have for messaging and email, but just the fact that it surfaces relevant articles and content for agents will make handling calls a lot easier.
# Post Call - Transcription & Summarization
Once the call has concluded, Al generates a full transcript and brief summary of the conversation and adds it to the ticket. This feature was already [launched](https://internalnote.com/preview-of-the-new-generative-ai-for-voice/) last year in EAP as *Generative AI for Zendesk Voice* and nicely fits in Zendesk's new Voice lineup.
💡
Important to note is that transcriptions[ are not free](https://support.zendesk.com/hc/en-us/articles/7470764710298-Zendesk-Talk-call-transcription-and-summarization-FAQ?ref=internalnote.com). You need to get either Zendesk Advanced AI or Zendesk QA in order to enable the feature, and they will cost you an additional 0,01$ per minute transcribed.
In my preview of the Generative AI for Voice upon its release last year I noted the following:
> (…) That being said, one major feature is currently missing for me, and that is the **lack of integration with the other Zendesk AI and Intelligent triage features.**
> If you have a summary of the call, why doesn't this also feed the Summary option in the Intelligence Panel? And why can't we see the Intent or Sentiment of the call? All the data is obviously there. And if I copy the conversation transcript into a new ticket, I get the results on the right.
> But, taking into account that feature is in an EAP, I'm sure this will all be nicely integrated once the feature work wraps up and the internal data gets hooked up.
[Preview of the new Generative AI for Zendesk VoiceWhat happens if you give Zendesk AI for Voice one of IT Crowd’s most confusing customer care interactions? Read the article to see the results!Internal NoteThomas Verschoren](https://internalnote.com/preview-of-the-new-generative-ai-for-voice/)
Hindsight is 20/20 and it’s clear I was only partially right in my remarks. Adding a summary to the context panel was a bad idea. A phone call is only part of the ticket live cycle and multiple calls can be part of one ticket. So having a summary for each call as a note on the ticket is clearly the right way to go.
But it’s nice to see that the other feature requests I had, namely intent and routing are now part of the Zendesk Voice offering.
# Quality Control - Quality Assurance
From incoming call to escalation to handing the call and wrapping it up, we finally get to the last part of the call journey: Quality Control.
Announced and released in [August](https://internalnote.com/roundup-2024-08/) the year, Zendesk QA for Voice is a new part of Zendesk’s new QA offering. Zendesk QA runs an automatic review over all your tickets and scores them according to elements like do agents understand the customer, are they friendly and have the right tone of voice, are they working towards a solution,..
Similarly it can spotlight potential risks like churn, bad csat, tickets that require repeated escalations,…
With the new Voice QA Zendesk expands this capability from traditional text-based channels to voice and uses conversation transcripts to audit your tickets.
# Wrap up
So, that’s the new Zendesk Voice. A complete rework of Zendesk Talk featuring AI agents to automate incoming calls, integration with omnichannel routing to make sure the right agent picks up the phone, insights via summaries and intent detection, Copilot to assist agents during the call, and Voice QA that uses transcriptions and summaries to provide actionable insights after the call wraps up.
> An AI Agent. An Agent Copilot. Zendesk QA.
> An AI Agent.
> Copilot
> Zendesk QA.
> Are you getting it? This is not a single product. These are three separate products. And they’re calling it Zendesk Voice.
All kidding aside, getting access to all the new features does require some work.
- You need **Zendesk Suite** to get access to a phone number and the overall voice capabilities in Zendesk.
- You need to buy **Zendesk Advanced AI** to get access to Agent Copilot with its quick replies, sentiment and intent analysis, and transcription and summary features.
- **Zendesk QA** brings you the new Voice QA features
- And finally, you need to get **PolyAI** to get the new AI Agent for Voice.
Long gone are the days of just buying Zendesk Suite and getting everything, and it feels we’re moving back to the days where Zendesk was a set of separate SKUs you could mix and match. The sole difference is that were previously you bought Zendesk to mix and match the channels you needed (talk, voice, help center, support,…), you now buy different add-ons depending on the kind of AI powered assist you need.
Do you want to deflect inquiries across channels and offer automated resolutions? You need to invest in AI Agents.
Do you want to assist agents and make their work more efficient? Look into Advance AI to empowers Agent Workspace. Want to improve the way your team works and get actionable insights? Ask for Zendesk QA.
I do like this new Zendesk Voice. Zendesk Talk was a bit long in the tooth and often overlooked when I work with Zendesk with customers reaching out to third party providers to handle their phone calls. Now more of these customers might be able to handle their conversations without requiring a third party system, although I hope the incoming calls and escalation flows will also become available to those third party providers via the Zendesk Talk partner edition APIs.
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### AI Summit - Agent Copilot
URL: https://internalnote.com/ai-summit-agent-copilot/
Last updated: 2024-11-05T10:00:38.000Z
Welcome to day three of my AI Summit coverage. Earlier this week I published an [overview](https://internalnote.com/zendesk-ai-summit-announcements/) of all announcements, and a [deep-dive](https://internalnote.com/ai-summit-ai-agents/) in the new Omnichannel AI Agent. Today, we're focussing on Agent Copilot.
At Relate last April Zendesk announced Agent Copilot, their new agent assist tool in early access. Now, at the AI Summit, this product becomes generally available to all Zendesk customers who bought the Advanced AI add-on.
Agent Copilot is positioned as a proactive AI assistant that works alongside agents and helps them resolve tickets faster and more efficient. Aside from helping agents doing their work, Agent Copilot can also assist with onboarding new agents. By proactively showing agents next steps to take, it lowers training time for new agents by removing to need to look up and know all procedures while handling tickets.
There's four big elements to Agent Copilot:
1. Auto-assist mode and suggested replies
2. Procedures that tap into knowledge and other platforms
3. Insights via intent, sentiment, related tickets and summarization
4. Write assist with tone shift and expand capabilities
Let’s dive into them one by one, starting with the biggest two auto assist and suggested replies.
# Auto assist and suggested Replies
Earlier this year I published a preview of the Agent Copilot after it became available as an EAP. You can read the entire overview in the linked article below, but in essence what Agent Copilot does is taking over the comment field in an Agents’ conversation.
[Preview of the new Agent Copilot for ZendeskA few months after its announcement at Relate, I finally had to chance to explore the new AI Co-pilot for agents in Zendesk. This article will explore what the Co-pilot can do, how it works and some initial impressions.Internal NoteThomas Verschoren](https://internalnote.com/preview-copilot/)
The comment field gets replaced with the new auto assist mode, where Agent Copilot can suggest replies and next-best actions for your agents. This could be a request for more information, information pulled from Help Center articles or macros, a confirmation we send to the customer or – and this is where it gets cool – a suggested action in external tools
> Agent Copilot doesn’t wait for agents to prompt it; it proactively assists with suggested responses and next-best actions, ensuring faster resolutions.
In its most basic form Agent Copilot is visible as suggested replies that the agent can *tab-complete*. These suggested replies are based on your existing macros, tickets and articles and allow agents to validate the message, optionally tweak it a bit, and submit it to your customers. It learn over time, and the changes you make to it feed the macro suggestions and content cues in the Help Center.
Where Agent Copilot really shines is when you start implementing procedures and enable the new auto-assist mode. This new mode will replace the comment field with a new auto-assist mode that offers next-best actions to your agents.
💡
This mode used to be called Agent Copilot. With this new release Agent Copilot is the overall name for all AI features in Agent Workspace, and the feature itself is now called auto assist
You can, for example, define a refund order procedure that requires an order number, product and reason for the return, before the refund is executed in Shopify.
Once created, Agent Copilot will read your customers’ replies and look for missing elements like the return reason, and automatically assist with offering prewritten replies for your agents to send to the customer.
If all required elements are collected, Agent Copilot can validate that information in Shopify, and if ok, the auto-assist mode will show a “Refund product” button to your agent that processes that refund directly in Shopify.
## Procedures
As mentioned above, Agent Copilot really becomes powerful once we start creating Procedures.
Procedures are a written down process of the steps an agent should take to handle a specific ticket. Examples are:
- **Subscription Refund** \- In order to refund an accidental renewal of a subscription the agent should ask for the product, email used for the purchase and check if the request arrives it’s within 7 days of renewal in the subscription management system. Once this is all in order, we can refund and let the customer know it can take 3-5 days to show up on their credit card.
- **Exchange a Product** \- In order to exchange a product we need to know the order, product and reason, we need to check if the information is correct, and ask if the customer needs a size smaller or larger.
- **Change Booking** \- for a customer to be eligible for a change in their booked trip they should provide the booking reference number, they need to have bought a flexi ticket, and the journey has to be further than 48h. If all is ok, we can rebook after asking what the new dates are.
As you can see, these are all processes an agent should know in order to handle customer tickets, and involve checking (and asking for) information, looking into external systems, and making changes in external systems.
When customer care teams onboard a new agent, teaching them all the processes often takes way longer than showing them how the ticketing tool itself actually works. With Agent Copilot, you don’t need to learn them all in advance anymore. The system will show you the required steps and responses, allowing even a new agent to be efficient from day one. And for people who do this job day in and day out, the promise of having a tool that handles a lot of the groundwork for you seems great.
With Agent Copilots’ procedures most of the above becomes a thing of the past. Procedures in Agent Copilot are a written down version of those processes. Once we write down e.g. our Subscription Refund procedure, Agent Copilot will recognize customer request about that intent (by leveraging Zendesk AI’s intent models) and will suggest replies an agent can send to the customer asking for the required information.
Once all information is available we can leverage Agent Copilots’ new actions feature to execute those actions via APIs directly in the external systems.
In our example, once we’ve collected the information we can have an action that checks for the renewal date for that subscription and then prompt the agent with a reply that confirms we can refund, or apologizes to the customer and lets them know that they are outside of the seven day limit.
An agent can then click the *Process Refund* button, which will handle all the refund action in the subscription system via API. Or, in case of a negative outcome, they can either submit the prewritten message, or overwrite and **do** execute the refund, since their manager approved a one time exception.
💡
Agent Copilot actions and procedures have **just* been released, and I haven’t gotten the chance to really deep-dive into them, but once I do get access to all the new goodies, I’ll surely do a deep dive article on the topic!
One cool detail is the fact that the procedures live now in Guide, and not in Admin Center. This makes Guide even more a knowledge manager than before with Help Center, articles, procedures and content cues with article management all being small pieces of the bigger knowledge puzzle. I’d kinda expect macros to be moved into Guide soon too, no?
# Insights via intent, sentiment, related tickets and summarization
The next main feature of Agent Copilot is actually a rebranding of existing Advanced AI features.
When Advanced AI was originally released last year, Zendesk added an Intelligence Panel to the Agent Workspace. This new context panel showed information like intent, sentiment, language and summarization to agents. Overtime it expanded to also include macro suggestions, similar tickets, merge suggestions as additional context relevant for the agents.
With these features an agent not only knows what kind of ticket they’re handling, but also gets insight in what the customer thinks and feels, allowing them to jump into the conversation with the right attitude and questions.
While I really like the context panel as an idea, it was a really overloaded piece of UI that was quite dense to parse. And, admittedly, some of the pieces of AI-powered data shown were also traditional already available in other places of Agent Workspace.
Take macro suggestions for example. At the bottom of the ticket view there’s a dropdown that shows all macros, while the suggestion macros section of the intelligence panel also shows some those macros.
With the release of Agent Copilot Zendesk also updated the Agent Workspace to refresh the way all the Advanced AI features are shown in the interface.
- **Intents** are now shown as a subtitle under the ticket subject
- **Summaries** are now a collapsable section at the top of the conversation
- **Macro suggestions** are embedded in the macro dropdown at the bottom of the screen
- **Merge Suggestions** get their own sidebar icon and section
- **Similar tickets** are added to the search dropdown at the top of the screen
Other insights like language and sentiment are now only visible as ticket fields in the leftmost sidebar. It’s a pity, I’d rather have a smily or grumpy face next to the ticket subject or intent. But I’m sure Zendesk will come up with something.
Similar tickets’ new location felt strange to me at first. It seemed hidden in a search dropdown, until I realized the most logical way to look for a similar ticket has always been searching for it, so it actually kinda is in the right spot.
Overall, I dig the new Agent Workspace updates. It feels more cohesive and all the AI richness is nicely integrated in the workspace, instead of being pushed into one sidebar panel. It also allows agents to keep another more relevant context panel or application open now, while not loosing the context provided by Zendesk AI.
# Write assist with tone shift and expand capabilities
To conclude the Agent Copilot section of the announcements Zendesk also pulled the existing writing tools that are part of Advanced AI under the Agent Copilot brand:
- **Tone shift** – which makes a reply more friendly or stern
- **Expand** – which turns a short reply into a longer reply
Even though they seem like two minor bullet points, they’re actually kinda awesome.
Personally, I use them in two ways.
- In one scenario I use the expand feature to turn a Frankenstein response into a cohesive reply. My replies often combine a paragraph copied from a Help Center article, combined with a sentence or two from a side conversation and a few short remarks based on the conversation. One click of the *Expand* feature turns these different elements into one smooth response. I control the information I type or copy/paste, AI takes care of the grammar and text.
- A second scenario is a way to make macros more personal. We all know those social media replies from Telecom providers with dozens of identical replies apologizing for power outages. The tone shift feature makes each of those replies unique by subtly shifting the macro wording per response send.
# Wrap Up
So, that’s Agent Copilot. A major rework of Agent Workspace with deeper integrated context, suggested replies and auto-assist powered by procedures available to your agents.
The feature is locked behind the Advanced AI add-on though, and not available as part of the default Suite. But for customers who want to reduce first-contact resolution (FCR), increase the overall tickets handled per agent or reduce workload (and by extension prevent burnout), it’s a powerful additional to the Zendesk product.
🧑✈️
If you have access to Advanced AI you can enable Agent Copilot in your instance ****today** by going to the **admin center > workspace > agent copilot*.
If you already used Agent Copilot via the EAP and wish to migrate your Agent Workspace to the new design, you can request so via [this form](https://docs.google.com/forms/d/e/1FAIpQLSd%5FG23gkL1exQ-jV5Ea2wc%5FAMf5xNQQKITf-8R98vnVriXRuQ/viewform?ref=internalnote.com).
Agent Copilot is but one side of the coin though. Where Agent Copilot makes handling the tickets that do get created more efficient, it’s counterpart, AI Agent, is there to automate ticket handling before tickets reach your agents, offering self service solutions to customers themselves.
There's also a bit of an overlap on both products. Allowing a customer to handle a product exchange or rebook a journey via an AI Agent, removes the need for an Agent (and Agent Copilot) to become involved, but similarly, if such a request does reach an agent, Agent Copilot can also handle that request for the customer together with the agent.
For now, if you want to both offer self service to customers, and assist agents with the tickets that *do* get created for the same use case, you need to do your work twice. You need to build an automation flow for the AI Agent, and write down a procedure for your agents, while integrating both with your backend systems.
I kinda hope that next year those two systems will grow towards each other from a management perspective, where we can turn procedures into bot flows. But for now, let’s be happy with what we have!
As for the impact, Agent Copilot has only been in EAP for half a year now, and just went GA, but at the event Zendesk did showcase one customer that saw a threefold increase in tickets handled per agent at peak time, going from 40 to 120 tickets per shift, while retaining the same, or higher, CSAT. Not bad!
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### AI Summit - the new Omnichannel AI Agents
URL: https://internalnote.com/ai-summit-ai-agents/
Last updated: 2024-11-05T09:50:04.000Z
Yesterday Zendesk streamed their AI Summit, announcing a series of improvements to the platform. The focus of the event was positioning the platform as a “complete CX solution for the AI era” and the company announced updates for CX leaders, customers and agents.
[An overview of the Zendesk AI Summit announcementsZendesk presented their latest releases at their AI Summit this week. New announcements include Omnichannel AI Agents, the new Zendesk Voice, Agent Copilot and improvements for Zendesk QA. Read my deep dive article to learn more!Internal NoteThomas Verschoren](https://internalnote.com/zendesk-ai-summit-announcements/)
For customers, one of the biggest announcements at the AI Summit has to be Zendesk’s rollout of new AI Agents features, and with it, the expansion of AI Agents across more channels.
Before I dive into the specifics, first some context.
Earlier this year Zendesk bought Ultimate and with it gained a powerful new ticket automation platform. When I wrote my [overview](https://internalnote.com/zendesk-acquires-ultimate/) of what Zendesk had bought I wrapped up with the following conclusion:
> So, where do we go from here. My ideal outcome would be the following:
> \- The Zendesk Flow Builder incorporates the more complex features from Ultimate with regard to linking to different branches, ad-hoc parameters and predefined integrations as reusable blocks. This becomes available across all Zendesk Suite customers
> \- The deeper reporting on bot behavior, conversation logs and resolution types become available for all Zendesk users
> \- Ultimate intents are mapped against Zendesk intents and appear in the knowledge panel
> \- In a year or two all Zendesk Suite users get access to the pre-trained Zendesk intents and AI features, and all Advanced AI customers, regardless of Suite type, get access to the custom models.
>
> This means that, for now, while Zendesk is busy integrating the (awesome!) Ultimate team into their company, customers buying Zendesk AI still get the prebuilt models, and Ultimate is an upsell to custom models and better training/intent mapping.
> Once the bot builder and reporting and intents are ported over to Zendesk, they can redraw their offering with Zendesk AI available across the board for all Suite customers (which will probably increase price) and reposition Advanced AI as custom models, better reporting and intent training.
That conclusion was right on many points, but wrong on some key assumptions; I made the mistake of thinking Zendesk was going to expand their Advanced AI offering with more powerful bot capabilities, whereas in reality Zendesk went the opposite way.
Since this summer, every Zendesk customer now gets a powerful AI Agent, powered by Zendesk out of the box. It offers generative replies based on Help Center content, a bot builder that offers flow based answers, and leverages prebuilt intent models for specific industries to show customers the right response. It’s included in Zendesk Suite, and (starting this fall) you pay for it via Automated Resolution pricing, or in other words, you pay for the times it does it job, and you don’t pay when the AI Agent passes the conversation to an Agent.
The Advanced AI add-on now focuses on the Agent Workspace and Admin side of Zendesk, offering powerful Copilot and insight capabilities, but more on that in tomorrow's article.
So, where does this leave Ultimate?
Well, for customers that need more than Zendesk’s default offering, and especially those interested in end-to-end email automation, custom AI models, additional language support, robust analytics and advanced integrations, Zendesk now offers *AI agents, powered by Ultimate*. A more powerful add-on that gives you more automation capabilities, replacing the native Zendesk conversational bots and auto replies via email.
In the documentation released after the AI Summit, Zendesk not only confirmed the new positioning of Ultimate and Zendesk Bots, it also announced a series of new capabilities for its platform.
The focus of the AI Agents section of the presentation came down to four new releases, each focussing on how Zendesk AI can help your customers:
1. Al Agents for email
2. Al Agents for voice, powered by PolyAl
3. Zero-training
4. More controls & governance
Looking at these releases it’s clear Zendesk is expanding its AI Agent offering into a true Omnichannel solution. Conversational bots for messaging, email automation for – well – email and webform based communication, and new this year: AI Agents that handle your voice channels. All of them build to handle customer queries instantly and autonomously, providing answers, automating (parts of) the process and providing context to agents if tickets are escalated.
☎️
I’ll handle the AI Agents for Voice in a later article this week
# Al agents for email
A few years ago Zendesk introduced Answer Bot. A basic form of email automation that parsed a customers email and replied back with three helpful Help Center links that might help the customer. Customers could solve their ticket if the proposed article worked, or reply with more context. Those tickets where than picked up by agents who can handle the rest of the replies.
Answer Bot was a solution of a time before LLM and AI models and it worked, but didn’t really provide a big impact, or at least not for the customers I implemented it for.
At last years’ [AI event](https://internalnote.com/zendesk-ai-drop-keynote/) Zendesk announced their new auto replies, or the capability of writing custom emails that would fire if specific intents were detected. It solved one side of the equation, namely you could more accurately reply with the right content, but it didn’t solve for the actual content of your responses. You had to manually write them, they weren’t personalized to the content of a customers’ email and, to be honest, I never implemented it for any customers.
Now with the new AI Agent powered by Ultimate, we *finally* have a good solution to automate email handling with the (re)introduction of a few new powerful capabilities that were part of the Ultimate platform already, and a few new ones that are really, really cool.
Looking at the screenshot above you can see a UI that has elements of Ultimate in it, but has that simplicity that Zendesk is known for. The new AI Agent builder is positioned as a fast and easy way to get started with your AI automation business setting up your bots' name, knowledge source, channel and then configuring some important settings for behavior, persona and response methods.
🤖
I will call the AI Agent powered by Ultimate just Ultimate in most of this article.
# Generative Email responses
The first big feature that’s now available to those who bought the Ultimate product is generative replies in email. Ultimate pulls information from your Help Center, website, CSV files or other imported content, and uses LLMs to generate a custom reply for your customer based on their email and your data sources.
The customer gets a custom written email with the information they requested, and can either reply with more information, or archive the email, upon which Zendesk can automatically solve the ticket after a few days. Since the AI Agent acts as a regular agent, [your automations](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/) will still apply.
It’s pretty powerful to setup since it doesn’t only automate your first email replies, it can also let the agent know it failed finding an answer, upon which we can add an internal note to the ticket. Since Ultimate can also update tickets with tags or other ticket data, we can leverage this information for when an agent *is* needed by using it as conditions in our Omnichannel Routing queue.
# Conversational Flows
Zendesk has this concept if a [Road to Automation](https://internalnote.com/road-to-automation/) where we can take very specific steps to go from tickets that are all handled by agents, to 60%, or even 80% of tickets being automated by AI Agents.
When looking at AI Agents in a messaging context, we implement this by moving gradually from offering only fully generated replies based on knowledge content, to more custom build flows that ask questions, shows options and uses API integrations to guide the customer towards a solution.
Doing this in a chatbot context is *easy*, in the sense that we are in a live conversation with the customer, can show buttons, carousels and options on screen and give and get immediate feedback from customers.
Until now, doing this via email was not possible and both Zendesk and Ultimate only offer so called first-reply automations, where we could only handle the first reply of a customer.
This already allowed for some pretty nice use cases, for example, we can receive an email and check for an order number. If the order number is there, we can call our backend systems over API, look up the order status, and reply to the customer with a (custom generated) email that contains their actual status. Or, in scenarios where the email didn’t contain any order numbers, we can reply with instructions on where to find it.
But what if the customer finds the order number and replies back? Until now, those emails had to be handled by your agents since Ultimate only processes the initial ticket being created and ignores the rest of the conversation.
But with the release of the new conversational flows for email, we can handle email conversations exactly the same way as we do chatbot conversations. If a customer replies back with an order number, we can process that response, call our API and let them know its actual status.
And imagine a scenario where we tell the customer that “hey, your package was delivered”, and actually they haven’t received anything. Here too we can, upon reply from the customer, immediately jump into our “Lost package use case”. All without resorting to agents.
This last feature is currently in EAP though, so when I do get full access to it, I’ll surely do a write up on the feature!
# Zero-training
As you can see from the features described above, the capabilities of the Ultimate platform are a lot more powerful than Zendesk’s existing conversational and email bots. They can pull from more sources, they can reply more personal and can deeper integrate with external platforms.
There is another feature though that makes it worthwhile to look into Ultimate. All AI automation and intent detection that comes natively with Zendesk Suite is based on, so called, prebuilt AI Models which focus on specific industries like Retail, Finance or Education. This offers a great out of the box experience since you can get started with Zendesk AI without any setup or training, but creates issues when your customers start asking questions about topics Zendesk AI doesn’t know about.
There are flows to request new intents from Zendesk, but ideally you would be able to add your own intents to the platform. Ultimate always had that capability.
If I want to add a “How to escape from an Imperial Star Destroyer” to my Rebels’ bot, I can create an intent, add around fifty training phrases with different variants of the question and train my AI model. A few moments later, my bot understands the intent, and I can start replying with information about the topic pulled from my knowledge base, or I can create a custom flow with options and integrations.
It’s a great feature that allows for perfectly fitting your AI Agents into the way your company works and thinks.
The only downside of this approach is that we still need to train our AI model. We need to add training phrases, we need to make sure that the system doesn’t get confused by overlapping intents (changing and canceling an order for example) and adding a new intent means typing or generating dozens of training phrases that might, or might not match the way your customer thinks.
With the rise of OpenAI and bigger LLMs, we’re used to AI platforms that just *get* what I’m talking about. Similarly to how generative replies in conversations and emails pull information from a Help Center and write down a custom reply, it would be nice if I can just tell Ultimate I want to add a use case for “Customer want to get a refund”, and the system would *just know* what I mean and can immediately map all customer queries about this topic to that use case.
And, with the release of Zero Shot Bots (or zero training as it’s called now) Zendesk did exactly that. We can add a new use case, give it a name and a clear description and.. voila. My AI Agents now knows about a new use cases and can assign conversations about that topic to it, ready for generative responses or custom flows.
It might seem like a small difference, we’re just omitting the step of adding training phrases right? But in reality, it makes a huge difference on implementation time. If you now think “I should add a flow for those times “a customer asks about our shipping times”, you’re basically halfway there.
By just writing down that use case, your AI Agent can detect that concept, and what’s left is either adding Help Center articles with information on the topic, or building out a custom flow to handle the request with more nuance.
# More controls & governance
Wrapping up this section of the AI summit presentation was a section about controlling your AI Agent.
First off you can set a predefined or custom tone of voice for your AI agent that matches your brand. You can define where data and response are sourced from, and control which terminology the bot should, or shouldn't use.
All of these feels like an expanded version of what's already available in both Zendesk and Ultimate, but I do appreciate the fact that AI for Zendesk is not only about features and improving workflows, but that data governance, privacy and security are part of the platform they're designing from day one (or year twelve to be honest).
# Thoughts
As you might have read in between the lines, I’m a big fan of the Ultimate platform, and I’m glad to see Zendesk is clearing up the way they are offering AI Agents to customers by making one an included part of the platform, and making the other even more powerful and positioning it clearly as the *advanced* AI Agent.
I do see a future where the existing Zendesk Bot disappears and gets fully replaced by Ultimate, similar to how Zopim Chat is almost entirely replaced by Messaging now.
But for now Ultimate feels to be a few years ahead of the existing Zendesk Bot, and to be honest most of the competition, when it comes to features and capabilities, so offering it as an add-on on top of the already powerful Zendesk Suite seems a logical choice. Similar to how Advanced AI is there for those we need an assist for their agents, AI Agents powered by Ultimate are there for those who want to make their self-service and ticket automation more powerful. Not everyone needs it, but those who do will be glad it exists.
One small tidbit to wrap up the article: how awesome was this slide showing how terrible the setup experience at other platforms is? When I gave the code in this slide to ChatGPT and asked which platform this might be code from, it gave me this:
> 🤖 The code snippets you shared align well with the structure of how Salesforce Commerce Cloud handles these processes.
Quite the subtle dig at the competition, and reminds me of [this](https://www.reddit.com/r/MacOS/comments/18bacds/this%5Fis%5Fstill%5Fthe%5Fdefault%5Fpc%5Ficon%5Fin%5Fmacos/?rdt=53236&ref=internalnote.com) all time great macOS icon.
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### An overview of the Zendesk AI Summit announcements
URL: https://internalnote.com/zendesk-ai-summit-announcements/
Last updated: 2025-07-06T17:51:46.000Z
Today Zendesk broadcast their AI Summit, an event focusing on innovations in Zendesk and Zendesk AI across its entire suite of products.
The actual event was held last week in New York as an on site event for customers, and today’s broadcast streamed the product release section of the event for everyone to watch.
Host of the presentation was Sarah Al-Hussaini, former co-founder of Ultimate and current VP of product at Zendesk.
The event started with a quick intro video highlighting all aspect of AI across Zendesk, and quickly moved into a conversation between writer Adam Grant and Stephanie Mehta, of Inc Magazine, focusing around how AI influenced are world.
The conversation was interesting with a few quotable topics like the question who will have the biggest impact: *those that use AI, or those that build AI*, while also focussing on the fact that there is a big part of the world growing up with conversations first who will quickly become used to having AI available. It’s an interesting discussion, and worth a listen. (Zendesk will have a digital broadcast available on demand not their website soon)
Interesting as the topic might be, I tuned into the event for product updates, so I was glad to see Sarah pass the stage to Zendesk’s CTO Adrian McDermott and Lisa Kant, SVP Product to guide us through the latest innovations in Zendesk.
They announced a lot of new features in this dense half hour presentation at a very rapid pace. While the presentation itself was scarce on details, Zendesk also uploaded tons of documentation to their website, so combining those with the latest EAP releases will give you a pretty good grasp on what’s been announced.
I will split this AI Summit overview into four articles.
Today will set the scene with an overview of how Zendesk positions itself on the market, followed by a high level overview of all the announcements. Tomorrow, and throughout this week I’ll publish three more articles each focusing on a key aspect of the presentation.
1. 🤖 [Deep dive in the new Omnichannel AI Agents](https://internalnote.com/ai-summit-ai-agents/)
2. 🧑✈️ [Overview of Agent Copilot](https://internalnote.com/ai-summit-agent-copilot/)
3. ☎️ [The new Zendesk Voice](https://internalnote.com/ai-summit-the-new-zendesk-voice/)
💡
(And.. hey Zendesk if you’re reading this:
Doing these write ups remotely after a digital presentation is quite an endeavor. Please invite me next time? 😇)
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
At Relate last year Zendesk, and again at this AI Summit positioned itself as the *Complete CX solution for the AI Era*, a suite of solutions that impacts all parts of your customer experience by offering four key features:
1. **AI-powered customer experiences**, leveraging AI to automate customer interactions via AI agents across all channels
2. **Workflow automation**, by enabling routing based on intents, and providing insights via summaries and intent detection
3. **AI-driven agent assistance**, by leveraging processes, knowledge, existing tickets and macros.
4. **Insights and analytics**, by auditing all your tickets and surfacing trends and outliers across all user interactions
Over the last year Zendesk already released tons of new features that broaden the impact of those four key elements.
The existing AI agent included in suite and powered by Zendesk got generative replies, more complex bot builder flow with parameters, and integrated deeper in Agent Workspace with a revamped conversation view for agents.
Advanced AI, the engine that powers both workflows and agent automation saw a general release of its knowledge base integrations, macro suggestions, suggested replies and tone shift features, while growing its AI models to tons of new industries including education, travel and employee service use cases. And Zendesk QA, Zendesk’s most recent acquisition got Bot QA and new Spotlight and AutoQA features over the past few months.
But even with those new releases, what was missing for me was a cohesive story for customers. Lots of the announcements of Relate, especially those around Copilot, were still in EAP and it felt that the friction between Ultimate and Zendesk Bot (or the confusion about what would happen) and the additional of AI bits and pieces across both Zendesk as a product and Zendesk as a license made the product line-up confusing.
This AI Summit presentation not only fixes a lot of the above, but also makes true on promises of earlier presentations and releases with a big line-up of product focussing on three main paths: your customers, your agents and your leaders.
Path one focusses on CX leaders and shows you how you can get **insights** in your data by leveraging Intelligent Triage and its Omnichannel Routing combined with deep insights via AI powered reporting. Traditionally a lot of the data in your company is locked in conversations and it’s costly and time consuming to read all conversations and extract valuable data from it. With AI we can automate this process and surface trends fast, automatic and at scale.
The second path is all about your customers and your customer experience. It’s about how Zendesk’s AI agents can help you **automate** up to 80% of your customer interactions while retaining, or improving your customer satisfaction ratings.
And finally, the third path focuses on tools that can **assist** your agents with resolving issues faster by leveraging Agent Copilot, reducing your average handling time by up to 20% and improve agent training on onboarding at the same time.
Or to put it differently, I asked Jon Aniano, SVP Product at Zendesk, how he would define the scope of this AI Summit:
> Zendesk’s goal for AI Summit is simple: to deliver the outcomes and benefits of the AI revolution to all CX and ES leaders - that means better & easier automation across all channels, transforming the role and quality of life of agents with copilot, and bringing automatic insights to operations leaders instantly.
Today’s article will give a high-level overview of the announcements, and as mentioned earlier, the next three days I’ll deep dive on each topic separately in more detail.
So, lets get started with the announcements:
Let’s dive in with the first star of the day: Zendesk’s AI agent and the new AI agent builder.
Powered by Ultimate, this new release will empower both conversations over web and social, and will also enable ticket automation over email channels.
The AI agent powered by Ultimate already offered robust omnichannel automation before this event, but this new release will add a couple of new capabilities making it easier to setting up a new bot, cutting down development and implementation time.
First off is the ability to use knowledge sources like your Help Center, website or blog to generate replies to customer emails or conversations. Similar to how conversational bots work, a customers’ inquiry can now be answered with a fully custom reply over email too. Also new is the capability to have multiple responses in a row and handle an entire email thread right within your AI agent, yet again similar to how chatbots work.
You can for example ask for an order number to assist a customer inquiring about the status of their order and explain the customer where they can find it. Once you have that order number, you can fetch the order status from your backend systems via API, and generate a nice reply back to your customers telling them the order should arrive in the next few days, followed by a tracking number.
This new capability unlocks the same kind of automation we’re used to for chat conversations, but now over email.
> AI agents boast an impressive 80% understood rate, resolving customer inquiries more efficiently.
Next was the announcement of the new Zero training option for AI agents. This feature does away with the concept of training a bots’ AI model by entering training data or training phrases in order to get your AI agent to understand new concepts like “How to replace a hyperdrive”.
With the new Zero Training you *just* need to enter a name for your use case and a short description, and your AI agent will immediately grasp the concept. You can then choose to have it answer based on your knowledge sources, or you can build a custom conversation flow to handle more complex scenarios.
From my experience with this feature during its early access program this cuts both implementation time and keeping your AI agent up to date with changes in your company in half.
And to wrap up the (text based) AI agent section, we got an update on expanded rule sets to maintain control on generative replies and check tone, terminology and other responses by your AI agent.
Your browser does not support the video tag.
Aside from text-based channels like messaging or email, Zendesk also has a phone channel as part of the Zendesk Suite. And at todays’ event Zendesk Talk got replaced with the **new Zendesk voice**.
The new capabilities of Zendesk Voice were interwoven across all parts of this presentation, and we got our first mention in the AI agent section of the presentation with the announcement of AI agents for Voice, powered by PolyAI
[PolyAI](https://poly.ai/?ref=internalnote.com) is a platform that leverages generative AI to create a so-called *Voice Bot*. It understands the customers’ questions, maps them to an intent and can then reply to the customer with generated replies based on knowledge sources, guide them through complex answer flows, or escalate to an actual agent, while asking the customer for the right information while doing do.
> With AI-powered voice assistants, over 50% of inbound calls are resolved autonomously, significantly reducing the load on human agents.
Notable here is that this is a partnership. Zendesk hasn’t bought PolyAI and hasn’t build this technology in house either. The two companies worked together to ensure an – at least as presented here on stage – seamless integration, helping you automate your most expensive support channels.
With the addition of voice to the AI Agent family, Zendesk now has a true Omnichannel Al agent solution, offering automation across conversations, email and voice, completing the automating customer experiences part of their promise at the start of this presentation.
The second big part of the AI Summit was all about the Agent Copilot.
Zendesk announced this new capability at [Relate](https://internalnote.com/zendesk-relate-day-1/) in April and has since been busy building out the product and feature.
They did a surprising switch here though. At Relate they announced Copilot as a mode in the comment field assisting agents with automating their replies and presenting next best actions. It lived alongside other Advanced AI features like the intelligence panel with intents, sentiment and summaries, and agents got additional context with other tabs in that panel that presented relevant tickets and macros.
It seems that in the last six months they decided that Agent Copilot as a brand was too valuable to be *just a text field.*
So today, they announced the release of Agent Copilot for all Advanced AI Customers. The *new* Agent Copilot assist agents in their ticket handling and contains a bunch of features deeply integrated in an upgraded Agent Workspace.
> Agent Copilot doesn’t wait for agents to prompt it; it proactively assists with suggested responses and next-best actions, ensuring faster resolutions.
The main release is the new auto-assist mode. This mode takes over the comment field and presents the agents with prewritten replies, and action buttons to automate part of the ticket handling. It’s powered by procedures you define for your agents, and pulls in information from your Help Center. And when you need to execute tasks in other tools like your CRM or order management system you can leverage the new Actions in Zendesk to setup reusable integrations. And all this without code and with just human readable text.
The new AI powered Agent Workspace also integrates other (pre-existing) elements like suggested macros, tone shift and other writing tools, summaries and intent prediction across the existing interface, and does away with the Intelligence Panel. This has the benefit that the AI elements feel like a more natural part of an agents workflow, and leaves room for more useful context panels like the [Customer Context Panel](https://internalnote.com/essentials-card/), Knowledge Panel or a [custom object record](https://internalnote.com/essentials-card-for-custom-objects/) to be always in view.
And just like for AI Agent, Zendesk also announced the addition of a Copilot mode for Voice. This new Copilot will be able to listen alongside a call your agents have with a customer, and will continuously prompt for next best actions and suggested replies to your agents by mapping intents and pulling information from your Help Center and procedures. It’s announced at the event with an availability of early next year. So I can only assume we’ll hear more of this at Relate 2025 next year!
And to end up where the conversation started: let’s focus on insights and data.
This is the third big use case for Zendesk AI, namely providing your team leads and company overall insights. Insights which can then be used to take action.
One of the big problems with data is that there’s too much of it, and that unlocking the important information within it has always been time consuming since you basically need to read every email, or listen to every call to get the important pieces.
With the release of Zendesk AI and Intelligent triage it’s now possible to automate this process and have Zendesk AI process all your interactions at scale, surfacing important data for you automatically, and before you know you need it.
A bigger new element for Intelligent Triage is the addition of Entity Detection. This allows you to define custom elements in Intelligent Triage which the system will search for in your tickets. Key examples here are order numbers, your products, hotel names or other parts of the conversations you can use to route tickets to the right team, or give to your agents as context for further support.
Custom entities can be detected and optionally written down to custom fields, which makes it useful for reporting purposes too.
> So where intents are there to know why the customers are contacting you, entities are there to know what they’re talking about.
All these insights become available in new Intelligent Triage (and Copilot) explore dashboards highlighting trends, making to possible to use this data to improve ticket routing, update documentation, or change processes to better meet the needs of your customers.
Measuring why customers are contacting you is one side of the coin, but it’s also important to know how you’re handling those inquiries. Know how efficient and solution-driven your answers are, know the sentiment of your customers and how tuned your agents (AI or human) are to these sentiments in their reactions and tone of voice are just a few of the elements that Zendesk QA and its autoQA feature measures, highlighting key elements like churn risks, dropped customer sentiment or other area’s you’d like to explore.
# Conclusion
To summarize the event, let’s quote Adrian McDermott at the conclusion of the presentation:
1. Leverage AI to get insights they never got before. Smarter more strategic decisions thanks to AI analysis.
2. Preparing for a world where 80% of the experiences are automated, and manage costs with AI agents and automated workflows
3. Empowering teams of human agents who take on increasingly complex cases while guaranteeing quality with Agent copilot.
*(slight paraphrased)*
In truth, the range of new announcement goes way deeper than both this one liner and this article.
Over the next few days I’ll publish three follow-up articles focussing on AI Agent, Agent Copilot and the new Zendesk Voice. Each article will focus on the new release in detail, and will wrap up with some insight on how these features will impact your Zendesk setup.
As for the event overall, I liked it. Zendesk Voice was a very nice surprise. Talk was beginning to feel a bit neglected and old, and this new AI-powered refresh will pull this product forward in a major way. As for how it actually works and if all promised improvements will actually have the impact that is promised remains to be seen, but with I’ve seen from PolyAI it’s bound to be good.
From a product line-up overall, Zendesk is getting more complex again. Where Zendesk Suite offered an all-in solution where you bought Zendesk and decided how many feature you’d need, we’re now back to buying Zendesk Suite and then choosing between a range of add-ons.
But if we put it all together into one flow, it does become easy to decide what you need.
Every customer starts with some version of Zendesk Suite. As they grow as a company, their support requirements will also grow with them and they’ll run into the issue of too many tickets and too much work for their customer care team.
That issue can be resolved in two ways:
1. You can lower the amount of escalated tickets by improving your ticket deflection and self service solution
2. You can improve the efficiency of your agents by providing them more context and automating more of their work.
Not every company needs both solutions, and not every company is a fit for both solutions. If you have a lot of recurring questions or processes that are repeatable, investing in a good AI Agent is your best bet. You can deflect tickets and give customers a better experiences since bots are faster than humans for those type of questions.
If you’ve got a high level of complex tickets or want to go for a white glove treatment for your customers – especially in the luxury sector – then going for an Agent Copilot is the better approach. Agents stay in control of those complex use cases that require a human touch, but the Copilot gives them all the context they need, and automates actions in other systems on behalf of the agent.
Once you’ve deployed either or both of these solutions you will see a decrease in average handling time and an increase in fully automated tickets. But that trend will flatten after a while as reality and changes catch up with your efforts. You’ll need to dive into reporting to find knowledge gaps, detect quality issues and find places to improve both your AI and human agents, as well as your processes. That’s where Zendesk QA comes in.
I’ve used the above approach with plenty of customers, and I’m sure it’s bound to help you deciding what’s the right product combination for you.
Tomorrow I’ll send out a second email with insights on the new Omnichannel AI agent release, followed by two more over the course of the week. Stay tuned!
#### Sources used:
- Zendesk AI Summit broadcast
- Zendesk Website
- Zendesk Newsroom articles
- Zendesk Help Center
- Conversations with product managers at Zendesk
- EAP access
- ****No data or screenshots were taken from internal resources or partner documentation.**
### A History of Zendesk chat and messaging channels
URL: https://internalnote.com/history-of-zendesk-messaging/
Last updated: 2025-03-07T13:25:24.000Z
Over the years Zendesk has expanded their platform from a simple email ticketing based tool to a full Suite of Customer Care and Employee Service solutions across multiple channels offering self service, agent workspace, AI Agents, social integrations and web widgets powered by the Zendesk API, Sunshine Conversations and other API services.
When talking to customers and colleagues it often gets confusing quickly when we want to find out on which combination of channels and services they run, and what's the right combination of those to get to a full setup. And this is especially true when we're moving customers from the classic chat solutions towards modern messaging environments integrated with the Zendesk Bot or Ultimate.
This article will serve as both a useful overview of legacy and existing solutions in there Zendesk platform and a reference for myself to explain this chaos to customers when it comes to explain setting up a good conversational channel for your Zendesk environment.
Zendesk currently has a few technologies in place to handle conversational experiences, and you might have seen or heard some of these terms: Zopim, Chat, Sunshine Conversations, Smooch, Zendesk Bot, Answer Bot, Messaging, Ultimate, switchboard, Web widget, classic widget, sunshine conversations web channel.
That's a lot of brand names and technologies, and it's easy to get lost.
# Introduction
So, what will this article contain? In it's most basic form we can explain enabling a conversational channel to Zendesk as follows:
A customer opens a web widget and asks their question. That question ends up in Zendesk and an agent replies.
A better version of this is:
A customer opens a web widget and asks their question. That question gets answered by a bot, and might get escalated to an agent. This version allows for self service and ticket deflection, lowering agent workload while improving the customer experience.
The above flows are just that, a high level overview of what happens when a customer contacts a company over a conversational channel.
But how does it actually work?
## Terminology
### Widget
A widget is a window where the customer can type there question in, see responses from a bot or agent, and read the conversation. It shows up on the website as a button that opens up into a modal pane.
### Channel
A channel is a unique way via which the customer contacts your team. It's a combination of a platform and brand. Eg. The Web Widget for brand A, or the Facebook Page for brand B, or the Instagram Feed of brand C.
Those channels can be integrated into Zendesk and get a unique `integration_id`which is used in the underlying platform and API calls.
### Business Systems
Tools like Zendesk and its Agent Workspace, Slack, Gmail inbox are all business systems that can receive customer inquires.
### AI Agents/Chatbots
AI Agents or chatbots are systems that automate responding to customers by retrieving information from your knowledge base, showing flows and asking for input for the customer. They receive input from a channel, and output to either the customer or switch the conversation to the next responder, e.g a business system.
### Conversational Platform
A Conversational platform is a set of APIs and business logic that orchestrates the interaction between customers and channels on one side, chatbots, and business systems and agents on the other sides.
Most work with a set of integration APIs or a configuration page to add the different channels, and have some kind of logic, often called a switchboard, to route specific channels to specific bots or business systems.
# Zendesk Chat
To start, let's with the basics and rewind our clock \~ 10 years.
Every Zendesk customer that wanted to use a conversational channel could enable Zendesk Chat as a channel. It enables customers to interact with agents via chat by talking to them in a widget.
Zendesk Chat used to be a separate company called Zopim that Zendesk bought in 2014\. Zopim offered a Live Chat widget and allowed agents to talk to customers in a separate chat interface that lived adjacent to the Agent Workspace.
Over the years Zendesk introduced their own widget (now called Classic Widget) that offered customers to either submit a webform or chat with agents. Here too, messages were routed over the Zopim platform to agents.
## Third Party Bots
Zendesk Chat, or more specifically Chat Enterprise, also offered a rich API which allowed for the addition of chatbots in the flow. Chatbot vendors could connect to Zopim over the API and both the Classic widget or the Chat window would then first offer a chatbot to customers, before the conversation got escalated to agents.
## Answer Bot
Back in 2018 Zendesk made its first venture into chatbots by released its Answer Bot. Powered by keyword search it would ask the customer how it could assist, and it would return three help center articles that might help. If the customer told Answer Bot those didn't help – or were bad suggestions – things got escalated to an Agent.
Answer Bot ran on top of the Zendesk Chat platform similar to how third party bots would run, but was configured entirely within the Admin center of Zendesk.
After the release of Answer Bot customers could now choose out of four options:
- Chat widget without a bot, offering only live chat
- Chat widget with answer bot
- Legacy widget without a bot, offering both chat and forms
- Legacy widget with answer bot
## Zendesk Chat phases
When looking at the Zendesk documentation, or when talking to Zendesk people, you might have heard the terms "phase 3" or "phase 4".
The Zopim integration in Zendesk went over a few iterations. Its first versions were Chat in a completely separate environment with separate logins not linked to Zendesk at all. This evolved into a separate Zendesk chat environment linked to your Zendesk account which was also reachable via a popup/overlay in Agent Workspace.
Currently, all customers that are still on Zendesk Chat use this latest, Phase 4 iteration of Zendesk Chat
# Smooch
In the late 2010s and early 2020s social messaging platforms like Facebook, WhatsApp and Twitter were on the rise and customers started connecting to companies over those channels more and more to give feedback, ask for support or well, complain.
Traditional customer care platforms like Zendesk used to be focused one webform and email based support channels and had only basic integrations with the most basic of social channels like Facebook posts or Twitter mentions.
But parallel to Zendesk's efforts a whole series of startups were beginning to focus on those social channels and started building toolkits to easily integrate social channels and CRM or customer care platforms. One of those platforms – Smooch – focussed on an easy to use API that made it possible to connect dozens of channels to bots and agent platforms. They promised to handle the complexity of routing conversations, maintaining status and, more important, abstracting away the unique elements of each platform by offering generic APIs for creating conversations, sending messages and switching conversations from bots to agents and back.
Zendesk recognized the need for expanding support for social channels in their platform and, in 2018, bought Smooch as an *easy* way to integrate their platform with dozens of social channels at once.
They bought Smooch, rebranded it as Sunshine Conversations and a new API platform was born!
💡
Right before the Smooch acquisition Zendesk released their first version of Custom Objects, User identities and Events and launches it under the Sunshine umbrella as a way to promote the extensibility of their platform. Naming their new acquisition Sunshine Conversations was a logical at that time, but now just feels confusing to me.
# Social Messaging
Integrating a platform like Smooch isn't done overnight.
In its first iteration Zendesk quickly patched and integrated WhatsApp, Facebook Messaging and Twitter DMs into their platform by building a Marketplace app called Social Messaging. Underneath this marketplace app, they routed all connected channels to Smooch and made sure that each channel would end up at the right customers' instance, and assigned to the right brand.
Smooch also offered a way more powerful Bot integration than the Zendesk Chat APIs did. Smooch supported bots across both web widget and social channels, and could more easily handle rich interactions like carousels, buttons and input fields.
By enabling the Smooch's [web channel](https://docs.smooch.io/guide/web-messenger/?ref=internalnote.com) customers could now offer a web widget on their widget that shows a rich chatbot, and escalate to Zendesk agents if needed.
# Sunshine Conversations
Even though customers got access to WhatsApp and other social channels via the Social Messaging app, Zendesk' wasn't going to stop promoting their newest acquisition as a product. Sunshine Conversations (or SunCo) as it was now called is a separate SKU customers can buy that offer some more powerful capabilities.
One of those is access to **switchboard**, an API that allows you to setup routing across different channels and platforms. Similar to the telephony switchboard of the 1900s this API allows you to route a specific channel (e.g. a specific Facebook page) to a specific chatbot. This allows for a lot more customization and enables a nice interaction between multiple channels, bots and agent platforms.
Zendesk still sells SunCo as up to this day for those that need higher API limits, more complex switchboard integrations and (some) third party channels. Most of the APIs that the paid-for SunCO gave access too are now freely available for all Zendesk Suite customers. (more on this below).
# Zendesk Messaging
So far in our little story we've seen Zendesk's early days of conversations with Zopim and Zendesk Chat. We saw a platform shift to SunCo with the Smooch acquisition and now we're reaching modern times with the release of Zendesk Messaging in 2021.
In essence Zendesk Messaging means a migration from conversations powered by Zendesk Chat/Zopim to conversations running on the Messaging/Sunco platform.
For most customers the move to messaging meant a move away from chats in a separate environment to conversations that arrive directly in Agent Workspace and look and feel like regular tickets.
After making the move to Messaging all social channels are moved from their classic setup or social messaging setup to Messaging. Which means all conversational channels like the web widget, Facebook, Instagram, the Zendesk SDK,.. now all appear under one big Messaging channels section in the Agent Workspace.
When Zendesk launched Messaging they also launched a new web widget. This new web widget was based on the design of the SunCo web widget and offered a rich conversational interface for customers to talk to, modernizing and replacing the old UI of the – now called – legacy classic Zendesk widget.
This new Zendesk widget only supports conversations and does away with the Help Center search, web forms and talk integrations the classic widget offered.
With Messaging, the flow of a customer conversation now looks like this:
## Zendesk Bot
When I started this article with "Zendesk naming gets confusing fast" I think this paragraph will proof that nicely.
Shortly after the release of Zendesk Messaging they introduced us to Answer Bot. Not the same answer bot as we knew from Zendesk Chat, but a new version that allows you to build flows via Flow Builder. These flows can guide the customer to specific Help Center articles, show button, carousels,...
[Learn how to build a full-featured Flow Builder Bot for Zendesk.In this article we will build a full-featured Flow Builder Bot for Zendesk. We’ll use every step type, use API calls and variables and show you how to create a bot yourself in a full length video tutorial.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-dinosaurs/)
The new Answer Bot got better over time and in 2023 it got renamed to Zendesk Bot with the introduction of Zendesk Advanced AI. This newer version of the bot allowed for generative responses, link responses to intents and expanded the bots capabilities with more powerful API integrations and parameters.
Since the new bot offers rich flows it means it's only compatible with the new Zendesk widget, leaving the legacy Classic widget in the dust.
# Ultimate
In March 2024 Zendesk announced their acquisition of Ultimate. Ultimate is an AI powered ticket automation platform that offers more powerful AI models, better bot builder capabilities and deeper reporting when compared to Zendesks' in house bots and capabilities.
[Zendesk acquires Ultimate: an in-depth overview of what the platform will gainZendesk is planning on acquiring Ultimate.ai, a leading AI and ticket automation platform. We’ve seen plenty of headlines, but what does this actually mean? This article dives into the nitty-gritty and shows what’s now possible thanks to this platform expansion!Internal NoteThomas Verschoren](https://internalnote.com/zendesk-acquires-ultimate/)
Ultimate, being built outside of the Zendesk, runs on its own technology stack and integrates with Zendesk by leveraging the SunCo APIs. Customers who want to use Ultimate in combination with Zendesk can do so by creating an API token in the Sunshine Conversation API section of the Help Center.
What's nice is that aside from – in my opinion — being a better bot and platform — Ultimate also abstracts away a lot of the integration work other bot vendors would need to do when integrating with SunCo and Zendesk. They handle the switchboard routing, can read [authentication metadata](https://internalnote.com/jwt-messaging/) just like Zendesk's own bot can and can set tags, user information and other data when escalating to Zendesk.
# AI Agents
This brings us to the last era, the one of the Zendesk AI Agents. After Ultimate's acquisition, Zendesk was quick in rebranding the platform and incorporating it into its broader product offering.
Customers who want to offer self service and ticket deflection can choose from a couple of AI Agents Types. They come in four flavors even though they aren't marketed as such.
1. The Zendesk Bot, the existing chatbot powered by Zendesk AI models and available in the Admin Center for configuration
2. Autoreply, email based replies to customers that suggest relevant Help Center articles, powered by Zendesk Guide search.
3. Ultimate Chat bots, powered by Ultimate's Zero shot AI models
4. Ultimate Ticketing automation, offering generative replies to customer emails powered by Ultimate's AI models.
Which one you pick depends on how complex or custom your needs are, but either one of them nicely integrates with Zendesk.
# Platform overview
The Zendesk Conversational platform grew quite complex quite fast. So if we take a look back at our initial flow: we had a customer opening a web widget, interacting with a bot, and ultimately reaching an agent, or resolving the conversation:
We can now expand this flow to the actual technically infrastructure of Zendesk:
💡
Note, this is still an abstraction, but it offers enough detail you can, hopefully, understand the pieces.
## Customer that only used the Zendesk Bot
This is the easiest flow. All native Zendesk channels and social platforms are handled by Zendesk automatically, forwarded to the Zendesk bot, and if needed escalated to your agents.
Behind the scenes however, Zendesk still leverages SunCo. Each of your native channels has an `integrationId`. They have a `defaultResponder`, namely the `zd:answerBot`, with a `nextResponder` of `zd:agentWorkspace`. All of which, luckily, you can just forget!
## Customer that uses the Zendesk Bot across multiple channels and brands
When we enable multiple brands in our Zendesk instance, things work similar from a platform perspective. However, Zendesk routes each channel to the right Bot based on the way the channels and brands are linked.
One weird quirkiness of the API though is that Zendesk exposes only a single instance of `zd:answerBot` as an integration to the Switchboard. So somewhere hidden in Zendesk, unreachable to us, there's a routing table that takes any conversation from the switchboard and passes it to a bot, and makes sure the right bot handles the conversation. But we can only affect this by assigning the bot to the right brand in the Admin Center, and there's no API access to handle this.
## Customer that uses Ultimate
When we move away from the native Zendesk Bot towards Ultimate, Sunshine Conversations begins to play a more visible role.
Upon setup we need SunCo credentials to integrate Ultimate, upon which Ultimate takes over and routes **all** available channels to itself, removing the link to the Zendesk bots. You'll find Ultimate in the SunCo integrations as `ultimateaibot`
## Customer that uses multiple Ultimate Bots
When a customer has multiple Zendesk brands and wants to link those to multiple Ultimate bots, we do not need to resort to any API work. Ultimate supports multiple bots in so-called [groups](https://support.ultimate.ai/hc/en-us/articles/360020358700-Connect-your-Virtual-Agent-Sunshine-Conversations-and-Zendesk?ref=internalnote.com) which allow you to map specific criteria like `integrationID` or website URLs to specific Bots.
## Customer that uses a mix of Ultimate and Zendesk Bots
The final scenario is the most complex use case. Let's say you're migrating to Ultimate bots and you want some channels handled by Zendesk bots, and others by the Ultimate bots.
In this scenario you'll need to get access to the paid version of Sunshine Conversations and start reading the [switchboard documentation](https://docs.smooch.io/rest/?ref=internalnote.com#tag/Switchboards) since you'll need to manually configure the switchboard and link some channels to `zd:answerBot` and others to `ultimateaibot` based on their `integrationID`.
You can read more about switchboard integrations in the following article:
[Using Switchboard to combine Zendesk Bots and AI Agents powered by UltimateZendesk AI agents come in two flavors. The native Zendesk Bot, and a more powerful one powered by Ultimate. But how do you combine both technologies in one Zendesk instance? This article will explain you how!Internal NoteThomas Verschoren](https://internalnote.com/using-switchboard-to-combine-zendesk-bots-and-ai-agents-powered-by-ultimate/)
💡
Zendesk did announce that they'll soon have the capability to manage combinations of Zendesk Bot and SunCo bots from within the Admin Center. But no ETA yet.
## A Tale of two AI Agents
In February 2025 Zendesk announced and launched a simplified line-up of their AI Agents. Gone are the different options and technologies and in comes a less cluttered line-up.
AI Agents now exist as Essential and Advanced options. The AI Agent Essential replaces the existing Zendesk Bot and offers **only** generative replies from the Help Center (and soon email). It removes Flow Builder, Article links and intents and offers powerful generative bot to get you started on the [road to automation](https://internalnote.com/road-to-automation/) and is powered by the uGPT engine of Ultimate.
AI Agents Advanced is a rebranding of Ultimate and offers the full suite of chat and email bot capabilities of the platform.
Let's pour one out for the Zendesk Bot and Flow Builder.
[What’s new for Zendesk AI Agents, essentials and advancedIn this article we’ll dive into Zendesk’s new AI Agent offering, and showing you how you can grow your automation with Quick Replies for the Help Center, AI Agent Essentials and the new Advanced offering.Internal NoteThomas Verschoren](https://internalnote.com/new-ai-agents/)
# Where to go from here
So, after all this, where does this leave you, the reader?
At least I hope you get some understanding in the possibilities Zendesk offers and how the different pieces of their ecosystem interact.
On the other side, I hope that for some people it raises some red flags. If you are still using legacy Chat in combination with the Classic Widget, I'd seriously consider moving to the modern Messaging setup since that's what gets the development, documentation and support resources these days.
If you are using messaging and still use the Classic Widget, I recommend migrating to the new Messaging Widget in order to get support for AI Agents, or conversational flows that just ask for name, email e.a.
If you are somehow using the SunCo Widget, move to the Messaging Widget. It has (almost) the same feature set and actually gets support these days. The SunCo Widget has been on live support for about a year now.
And if you're using Messaging and some kind of AI Agent already, well done. You can just keep using the new features, kick back, relax and enjoy your life.
## Bonus: Zendesk Message
Even though the above is a full overview of all Zendesk conversation products up till today, I skipped over one of them. Right before Zendesk bought Smooch they had a short-lived product called Zendesk Message.
The product provided a brand new interface outside of the regular Zendesk workspace that focused on conversations and chats and did away with any of the traditional concepts like ticket fields, statuses or other powerful features that Zendesk offers.
I [found](https://www.demeterict.com/en/zendesk-stories/introducing-zendesk-message/?ref=internalnote.com) **one** screenshot online after searching for a long time.
Even though this product was short-lived it's amazing to see elements like the Interaction History and speech bubble based layouts in this layout. Elements that only recently showed up in the Agent Workspace. I kinda like this cleaner layout. Less options, less information, and a focus on work todo (left), the conversation (center) and context (right).
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Zendesk News Roundup for October 2024
URL: https://internalnote.com/roundup-2024-10/
Last updated: 2024-11-03T10:47:55.000Z
Welcome to Octobers' newsletter. When I started preparing this article two weeks ago it felt like a very light issue. That was, until Zendesk opened the floodgates late last week and releases a dozen features all at once.
Thanks Zendesk!
# 🏢 Company
## Rebranded Zendesk QA
Last year Zendesk bought Klaus as a way to bring in Quality Assurance into the Zendesk platform. It allows for automatic processing of all your tickets – both those handled by AI Agents and human alike – and gives you insight in quality or responses, customer feedback, dangers like churn and surfaces tickets that warrant attention.
This month Zendesk fully moved Klaus into the Zendesk platform by redesigning the UI to better fit with the rest of the Zendesk platform, and also made the new features available in two specific license options:
- Zendesk QA add-on: Get Zendesk QA for all your agents
- WEM Add-On: A new add-on SKU bundling Zendesk QA and WFM, it's scheduling and employee management tool.
# 🎉 New Releases
## 🤖 AI Agents
It seems to be quiet before the storm when it comes to AI Agents. Next week Zendesk has their AI Summit (with a [digital broadcast](https://virtualevents.zendeskai-summit.com/series/ai-summit/landing%5Fpage?lang=en&ref=internalnote.com) a week later) so I think we can see a lot new bigger releases then.
# 👨🏻💻 Agent Workspace
### Agent Home
Starting THIS week [Zendesk](https://www.linkedin.com/company/zendesk/?ref=internalnote.com) will rollout their new Agent Home as the new experience for all customers who use Agent Workspace.
This means the old dashboard will be replaced with a new home page that gives agent direct access to assigned tickets, recent updates and followed tickets.It’s a big shift away from traditional view based workflows, but if you want to get the most out of this new experience, I’ve got you covered! I wrote [a full article](https://internalnote.com/agent-home-tips/) with the best tips and tricks to get started.
Leading up to this big release the Agent Home also got an update with an expanded Channel filter that now includes all Zendesk channels ranging from web (forms, api, tickets created by agents), to Messaging (chat, social and web) to Child Tickets created via Side Conversations.
This is a big release!
### Updates to Omnichannel Routing
Omnichannel Routing still seems to be the way forward when it comes to having a strategy for assigning and distributing tickets to your team in Zendesk.
Up til now you could only assign tickets based on capacity. An available agent with the least assigned tickets would get the newest ticket assigned (if skills matched). With the new [round-robin](https://support.zendesk.com/hc/en-us/articles/7990049158554?ref=internalnote.com) option you can now assign based on the last time a ticket was assigned to an agent, giving you a different way to distribute tickets.
Additionally we can now make email tickets act similar to Messaging tickets and have them automatically open in a new tab for agents upon assignment. This way agents don't always need to go back to Agent Home (or a view) to check for new work.
And to conclude this months' routing updates, we now have the ability to add both new tickets that have never been assigned to an agent, as well as tickets that are reassigned to another group to queues. Until now tickets could only go into a [queue](https://internalnote.com/beyond-triggers-moving-to-queue-based-assignment-in-zendesk/) once, but now when you reassign a ticket from your Support team to the Finance team, all queues and routing rules for that group will be applied.
Pretty useful and that will fix the routing logic for many customers!
### Saved searches
The new search function with recent searches and built in filters was launches earlier this year. It offers a quick way to retrieve relevant tickets or users, and – at least for me – offers a nicer experience than the global search experience.
With this month's update you can [save your search query](https://support.zendesk.com/hc/en-us/articles/8045381365786-Announcing-the-ability-to-save-searches?ref=internalnote.com) so that you can run the same searches more easily. You can store up to 20 searches which you can give a unique name. Note, searches are per user so there is no global search filters to be shared.
### Advanced AI
The Merge Suggestions feature for the Intelligent Panel has now been made available for all Advanced AI users. This new panel will surface active tickets from the same requester about the same topic, allowing agents to quickly [merge](https://internalnote.com/merging-tickets/) them before working on the issue.
We also got a new education intent for Intelligent Triage and the Zendesk Bot, aiming to meet the specific needs of managing and updating information about education courses, examinations, admissions, scholarships, and student life.
### View Categorization
> Zendesk is pleased to announce the ability to categorize both your shared and personal views. Categorizing views creates a folder structure in Zendesk Support so that you can more easily navigate your views list.
I've got some [views on Views](https://internalnote.com/my-approach-to-zendesk-views/), but I really like this new categorization feature. The more assignment and a list of your work moves towards Agent Home, the more we can start using Views as a way to get an overview of the outstanding work overall sorted by topic, priority, customer or team. This new categorization makes it possible to collapse similar views into nested sections, making a long list of up to a lot more manageable
In addition to [categorizing views](https://support.zendesk.com/hc/en-us/articles/8043835835674-Announcing-the-ability-to-categorize-views?ref=internalnote.com), the limit for shared views has been increased from 30 to 100\. This new feature does run the risk of enticing people to create even more views than before, but as long as you make sure there is **one** spot to see all work to be done sorted by service level or priority, I think you're good!
### Estimated waiting time
One feature that got lost in the migration from Chat to Messaging was any sense of waiting time for the customer. With the recent introduction of Messaging triggers, these last few feature gaps are now being filled.
> With this change, businesses can now set expectations on wait time with their customers. An admin can set up [messaging triggers](https://support.zendesk.com/hc/en-us/articles/8015292388378?ref=internalnote.com) to share estimated wait times with customers after the messaging ticket is added to the queue. Wait time is estimated based on the ticket's position in the queue. They can also set up triggers to send a message once the messaging ticket is assigned to an agent.
Speaking of Messaging Triggers, keep an eye out for a new article on those later this month! [Subscribe today](https://internalnote.com/#/portal/signup) for free to Internal Note to get this directly in your inbox!
### Ending messaging sessions
It's now possible for agents to end a conversation with a customer right from within Agent Workspace by clicking an *end session* button.
Before this, you needed to solve (and close) the ticket via a macro or trigger to pass control back to an AI Agent, but with this new option, those *hacks* are not needed anymore.
I do feel a conflict between Messaging sessions with its *send* and *end session* button on one end, and the ticket status buttons on the other. The amount of times I've submitted a ticket instead of sending a reply is, well, a lot, and this might add additional confusion by adding yet another button to the Agent Workspace.
## 🔎 Help Center
No new releases this month for the Help Center, but we did get two fun EAPs!
- Support for Custom Objects in end-user forms (EAP to be found [here](https://support.zendesk.com/hc/en-us/community/posts/8009077248282?ref=internalnote.com)).
- Support for multi-placement of articles (EAP to be found [here](https://support.zendesk.com/hc/en-us/articles/7867489163930-Placing-articles-in-multiple-sections-with-article-multiplacement-EAP?ref=internalnote.com)).
Both are long-awaited items and once I get access to these, I surely write about them!
## 🧱 Open and Flexible Platform
### Redaction suggestions
If you use the Advanced Data Privacy and Protection add-on you can now [setup](https://support.zendesk.com/hc/en-us/articles/6669399593882-Automatically-detecting-sensitive-information-for-redaction?ref=internalnote.com) redaction suggestions that highlight PII for your agents in tickets.
### Closed Tickets editing
You can now do limit edits on Closed Tickets via the Agent Workspace or Zendesk API. You can edit the Tags, Subject, and Priority ticket fields on closed tickets making it possible to tweak how these tickets show up in, for example, your reporting after Agents already wrap up the ticket.
## 📊 Reporting and Insights
### Customizable CSAT
Earlier this year Zendesk announced the Customizable CSAT as a better way to ask customers for their opinion. The entire feedback from got a redesign, and customers can now:
- Edit the feedback form title and text
- Select a rating scale range with either 1-2, 1-3, or 1-5 scale increments, compared to the old thumbs up, thumbs down from before.
- Display rating options as numerical, emoji, or custom text
In your reporting these will still show up as Good/Bad though. Read more about this feature in the article below:
[Preview of the new Customisable CSAT EAP for ZendeskThe new Customisable CSAT EAP for Zendesk has arrived, finally allowing you to change your rating scale, choose emoji, numbers or labels, and customize your follow-up questions. This article contains an initial overview of the new feature, and shows how it works with existing API integrations.Internal NoteThomas Verschoren](https://internalnote.com/preview-of-the-new-customisable-csat-for-zendesk/)
# ⚠ Major Changes.
## 2SV Authentication
Starting end of October Zendesk will start rolling out (and enabling) a new security feature to all customers:
> Customers who have Zendesk authentication turned on will be automatically enrolled in 2SV. When customers sign in from a new device, 2SV will prompt them to enter a 6-digit code they must retrieve from the primary email address on file before they can sign into their account.
This will be annoying for customers that share credentials across employees or with third parties, but is a welcome change from a security/phishing/credential leak standpoint.
Read more [here](https://support.zendesk.com/hc/en-us/articles/7955412933274-Announcing-two-step-verification-2SV?ref=internalnote.com)
### Sunshine Conversation integration deprecations
When Zendesk bought Smooch a few years ago they rebranded it to Sunshine Conversations, added the APIs right inside the Admin Panel, and basically left the rest of the platform alone.
Apparently that's now changing with a first set of deprecations of existing platforms:
- Deprecation of Sunshine Conversations’ Slack Connect Integration on Sept 12th
- Deprecation of Sunshine Conversations’ Help Scout Integration on Jan 6th 2025
# 💡Insights
## The information on Resolution Based pricing
> Instead of charging businesses based on how often they use the software—essentially an AI chatbot—to try to resolve customer problems, Zendesk began charging them only when the chatbot completed the task without needing employees to step in.
[Read the article](https://www.theinformation.com/articles/new-ai-business-model-charging-customers-only-when-the-tech-works?ref=internalnote.com)
## Grace Hopper on Data Storage
> \[11:30\] I thought up a couple of curves. I have no numbers to put on them and research hasn't been done yet – But at least I think I can talk about the shape of them. Suppose this is dollars, and this is time --
>
> And an event occurs here. Now the value of the information about that event goes up quite sharply immediately after the event, which the further you get away from the event in time, the more the value of that information levels off. It goes up very sharply and it levels off. Now ultimately, it either gets replaced by a new
> piece of information, or we decide we don t need it online anymore and we transfer to historical files, microfilm or something like that. Of course, in industry they have to save it for the IRS, so the value curve probably look something like that: a sharp rise, a leveling off, and then an eventual transfer to some form of historical file.
> What about the cost of that information? Cost of information is very, very low at the time of the event, but the further you get away from the event in time,
> the more the cost you pile up to start, maintain it, and add any information to it. So the cost curve starts low and then it goes zooming up – **Now there is a lovely crossover point there – that is the point at which keeping that information in our online system is costing us more than it's worth to us.**
I saw this presentation last week and this specific section made me think of [Zendesk's storage pricing](https://internalnote.com/storage-limits/). Keeping all tickets is relevant to feed your reporting, to use them in the Intelligent Panel similar tickets, and as a reference for Content Cues and Macro Suggestions.
On the other hand pulling information from tickets that are 2, 4, 8 years old becomes less and less useful overtime. So each company has some kind of crossover point where the usefulness of those old tickets is not worth the cost of keeping them around.
### Cards Against Complexity: A Podcast
> The podcast where we dive deep into the world of customer experience (CX) with a twist! Whether you're a seasoned CX expert or just starting out, join us for real conversations on complex service topics, and laugh a little along the way.
The team at Next Matter [launched](https://www.nextmatter.com/blog/we-did-customer-experience-in-the-woods-with-our-own-podcast?ref=internalnote.com) a new podcast focussing on CX with a twist. The first episode features Julian from [Babelforce](https://internalnote.com/sponsor-babelforce/). Fun to see how close the Zendesk partner ecosystem works together!
[Next Matter on LinkedIn: "CX in a Hawaiian Shirt" | Julian Hertzog | Cards Against Complexity | Ep…Ep 001 of Cards Against Complexity is live! 🎙️👕 Here's a sneak peek of our first hand of cards with Julian Hertzog from babelforce | Zendesk Voice Partner…LinkedInNext Matter](https://www.linkedin.com/feed/update/urn:li:activity:7245401795240280065?ref=internalnote.com)
# 📝 Articles this month
[Automating article attachments with the Guide Media API and ZapierThis article explains Zendesk’s new Media Gallery for Help Center articles, allowing you to manage images across multiple articles. We show off the Media Library API, which enables automated uploads from external tools like Google Drive using Zapier, improving workflow efficiency for teams.Internal NoteThomas Verschoren](https://internalnote.com/automating-article-attachments-with-the-guide-media-api-and-zapier/)
[Handling duplicate ticketsThis article explores ways to merge duplicate tickets from the same user. Customers often submit multiple requests via various channels, slowing response times. Solutions include using the Customer Context Panel, Zendesk AI-based Merge Suggestions, or automating ticket merges.Internal NoteThomas Verschoren](https://internalnote.com/merging-tickets/)
[Preview of the new cards for Custom ObjectsZendesk’s new Essentials Card lets you customize user profiles next to tickets, and now, Custom Object Cards do the same. You can reorder, hide, or add fields to streamline your agents’ view. A simple yet customizable update to enhance the agent experience.Internal NoteThomas Verschoren](https://internalnote.com/essentials-card-for-custom-objects/)
# And Finally...
> All unassigned messaging conversations become inactive after 10 minutes without end user interaction. The standard omnichannel routing configuration counts only active messaging conversations towards an agent's capacity. This behavior is controlled by the messaging activity routing setting. When messaging activity routing is off, messaging tickets that become inactive while still in a queue can be assigned to any available agent without taking up their capacity.
Omnichannel Routing.. just when you thought you understand it, they [throw](https://support.zendesk.com/hc/en-us/articles/7640406656410-Workflow-Using-omnichannel-routing-queues-to-handle-active-and-inactive-messages-differently?ref=internalnote.com) you another random feature that makes it confusing again 😅😅
Thanks for reading!
## Sign up for Internal Note
A blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Preview of the new cards for Custom Objects
URL: https://internalnote.com/essentials-card-for-custom-objects/
Last updated: 2024-11-03T10:48:03.000Z
Last summer Zendesk released the new Essentials Card for user profiles. It allows you to customize the data shown for a requester next to a ticket. You can hide user fields and reorder them to fit your needs.
[Preview of the new Essentials Card in ZendeskA quick overview of the new Essentials Card for Zendesk user profiles.Internal NoteThomas Verschoren](https://internalnote.com/essentials-card/)
Similar to user profiles, Zendesk has another type of data that can be shown next to tickets and that's [Custom Objects](https://internalnote.com/tag/custom-objects/). When you link a Custom Object to a ticket via a Lookup Field, you can use the record inspector to show its data next to the ticket conversation.
Up to now you couldn't modify this view. Fields were shown in the order they were added in the Admin Panel when you configure the object, and you can't hide fields to be shown.
In the screenshot above for example fields like *Distributor* and *External ID* are useful for some flows like [Custom Object Triggers](https://internalnote.com/multi-select-support-for-users-organizations-and-custom-objects-in-zendesk-2/) but not relevant for my agents.
# Custom Object Cards
Enter Custom Object Cards. Similar to how the User Essentials Card allows you to modify what's shown, we can dive into *Admin Center > Workspaces > Cards* and select a Custom Object type from the list.
Once selected, you can use the rightmost panel to reorder your record fields by dragging them to a new position. You can use the X to remove them from your Agents' view, and similarly you can use the *Add Field* button to add a previously hidden field.
## Agent Experience
After customizing the Card, your agents will see the new layout appear the next time they open a ticket:
# Conclusion
The new Cards for Custom Objects are a nice MVP but I'd love the Record Inspector to show a richer experience than the current plain layout.
Why can't we add some color to the mix and e.g. highlight the *Action* genre in the screenshot above with a green color? Summary could be nicer text than the grey text box, and a bit more visual hierarchy would be a nice to have too.
But aside from those wishlist items — and admit it, everyone has some – I like the way the Agent Workspace gets more customizable month after month.
### Handling duplicate tickets
URL: https://internalnote.com/merging-tickets/
Last updated: 2024-11-03T10:48:11.000Z
Earlier this month at the [Zendesk Partner Days](https://www.linkedin.com/posts/zendesk-partners%5Fthanks-to-everyone-who-made-the-2024-emea-activity-7240037729349066752-kvff?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop) a conversation came up about the best way to merge tickets made by the same end-user. It's a scenario that happens a lot. A customer has an urgent question and they shoot an email, web form and social message to every support channel they can find.
Not only will this not result in a faster response, it'll actually make matters worse since they are clogging up the ticket queue of agents with multiple identical questions, delaying responses for everyone. And they introduce the risk of getting different responses from different agents handling their tickets.
Since you can't block customers from doing this, we need to resort to features in Zendesk to handle this common issue.
# Approach 1: Customer Context Panel
The easiest solution is leveraging the Customer Context Panel to find all recent tickets from a customer. You can open these tickets in new tabs and merge them with the current ticket, before you move forward handling your customers' ticket.
The downside is that it requires manual work and introduces another step for your agents, but it's a native feature that's available for all Zendesk customers so requires no additional apps or licenses.
# Approach 2: Leveraging AI
One of the key selling points for Advanced AI is that it automates manual steps done by agents. In our case, the manual step of checking for duplicate tickets can be automated by leveraging Zendesk Advanced AI.
The [*Merge suggestions*](https://support.zendesk.com/hc/en-us/articles/6885971957914-Merging-related-tickets-based-on-suggestions-EAP?ref=internalnote.com#:~:text=The%20merging%20suggestions%20feature%20identifies,Workspace%20to%20use%20this%20feature.) options in the Intelligence Panel will offer agents a list of similar active tickets from the same requester. Agents can select tickets and merge them, thus removing these tickets from the queue and having one single ticket with all context to work on.
This works for scenarios where users fire identical requests – even those phrased differently – across different channels, or for new tickets that are responses to or continuations of existing tickets. E.g. a customer emailing a picture of the problem in a separate email.
The upside is that the filtering is better than just looking at recent tickets and the system surfaces relevant tickets automatically, removing these additional steps from your agents' workflow. But as with everything in Zendesk Advanced AI, your agents still stay in control. It's the agent that merges tickets, preventing accidental wrong merges by simply automating the matter fully.
# Approach 3: An automated, but unfiltered, approach
Internal Note wouldn't be in Internal Note if an article didn't include some over the top hack of existing Zendesk features 😅. One of the scenarios we discussed at the Partner Days was a way to automatically merge all active tickets of the same requester, regardless of their content.
So, let's do exactly that.
First we want to know if a user has an active ticket. We could do a lookup via external workers and the `/api/v2/users/{user_id}/tickets/requested` endpoint, but there's a way we can do this in-product.
In our approach we will store the *ticket id* of a requesters' latest ticket in the user profile. We can then check this value to see if a user has an active ticket, and if so merge an existing ticket into our new ticket, clearing up the queue.
💡
I purposefully use the latest ticket as the remaining ticket so that if a customer manages to resolve their inquiry themselves, that's the first message an agent will see.
## Storing the latest ticket
In order to store the latest ticket, we need to create a new user field first called *Latest Ticket* with field key `latest_ticket`.
Once setup we can use a trigger to update this field whenever a ticket is created. For this we need a webhook that updates user profiles. I wrote a full tutorial on the topic in this article:
[Update a requester name via webhooks and custom fieldsUpdate a requesters’ profile based on form field input.Internal NoteThomas Verschoren](https://internalnote.com/update-a-requester-name-via-webhooks-and-custom-fields/)
But in essence you need to create a webhook that calls `api/v2/users/{{ticket.requester.id}}.json`.
🔒
Do make sure to use `admin@domain.com/token:API_TOKEN` as the authentication method, leveraging a Zendesk API token. You no longer can use passwords for API authentication as of this summer!
Once you've configured the webhook, create a trigger with the following conditions:
And the following action:
```json
{
"user":{
"user_fields": {
"latest_ticket": {{ticket.id}}
}
}
}
```
Once setup this trigger will look at each created ticket and check if the value of *Latest Ticket* is blank. If blank, it'll update the user field with the ID of our current ticket.
## Merging new Tickets
If you're wondering why we check for the presence of the *Latest Ticket* in the previous ticket instead of just updating it every time a new ticket is created – nice catch!
The reason we need to check and can't just update the value anytime a ticket is created is simple: if we always update the ID with the latest value upon creation, we always overwrite previous values and loose the chance to merge tickets.
In this final step we'll wrap up our workflows and merge new tickets into existing tickets if they're available. First we need to create a webhook we can use to merge tickets. Similarly to the previous step, this is based on an earlier article on this blog:
[Escalating a customer request to a Zendesk Help Center form for more information.This article shows you how to escalate an existing ticket to a new ticket form submission and merge the result.Internal NoteThomas Verschoren](https://internalnote.com/asking/)
Here to, we need a webhook that calls `api/v2/tickets/{{ticket.id}}/merge` with a configuration like the screenshot below:
We then create a new trigger with the following conditions:
In this scenario we run the trigger when we **do** detect a Latest Ticket value. Our trigger will then do two things:
1. Merge the older ticket into this the newer ticket
2. Update the *Latest Ticket* field with the newer Ticket ID.
💡
We can update the **Latest Ticket* field ince we already fired our merge webhook first. Even though it might not have finished, we passed the **Latest Ticket* value so we can safely update our user field now.
### Merging Tickets
```json
{
"ids":[{{ticket.requester.custom_fields.latest_ticket}}],
"source_comment": "Closing in favor of #{{ticket.id}}",
"target_comment": "Combining with #{{ticket.requester.custom_fields.latest_ticket}}"
}
```
### Update User
And identical to the first trigger:
```json
{
"user":{
"user_fields": {
"latest_ticket": {{ticket.id}}
}
}
}
```
## The Result
Once enabled, whenever a user creates a ticket, we'll check if they have a previous ticket. If they do, we will merge that ticket into their most recent active ticket each time they create a new ticket.
Since closed tickets can't be merged, we can be sure we'll only merge active tickets
Naturally, the approach above is **crude**. If a customer emails about different topics we still merge tickets. You can prevent this by leveraging more conditions in your triggers. For example, if you have Advanced AI you could only merge tickets with similar intents. Or you could set an automation that, if a ticket is older than a week, you set the *Latest Ticket* value to *blank* so that our trigger doesn't run,...
Regardless, during my conversation at the Partner Days we were explicitly talking about this broad approach, and promises be promises, here's the article for those present on is how I would tackle it in more detail.
# Approach 4: Leveraging the Zendesk Marketplace
When it comes to Zendesk you can be sure that if there is a process that can be done via triggers, API and webhooks, there's bound to be a Partner on the Marketplace that used those elements to create an app.
[Merge Tickets](https://www.zendesk.com/marketplace/apps/support/1052353/merge-tickets/?ref=internalnote.com) by Knots (disclaimer, they [sponsored](https://internalnote.com/sponsor-knots/) this blog a few months ago) is such an app. It offers a nice UI to setup different conditions and it will then will automatically merge tickets that match those rules.
You can merge tickets with the same requester, category, subject or other custom field, add an internal note with information about the different merged tickets and even choose to turn comments from other tickets into internal comments.
The latter is useful if you combine information from different users. For example: a ticket from a customer complaining about a delayed delivery gets merged with the feedback from the delivery partner. You don't want either part to see each others' comments.
The app from Knots is only one of [many](https://www.zendesk.com/marketplace/apps/?query=merge&ref=internalnote.com) on the Marketplace that offer similar capabilities, all in different forms and functions. I kinda like how this one integrates with the rest of their suite of ticket automation apps though.
# Conclusion
The problem of duplicate tickets in support is an annoying issue. It inflates ticket numbers, duplicates agents work and might make appear certain topics bigger than they are in reporting if multiple copies of a ticket are raised by the same customer.
Self Service, good documentation and documented processes are a way to prevent duplicate tickets. Similarly, enabling an [internal SLA on first reply time](https://internalnote.com/sla-policies/) will motivate faster reply times, hopefully preventing customers' from duplicating requests due to lack or slowness of feedback.
But if those all fail, the native features in Zendesk – being it Customer Context for agents or AI suggestions – can downside this issue, and API integrations can assist with cleaning up tickets in bulk.
### Automating article attachments with the Guide Media API and Zapier
URL: https://internalnote.com/automating-article-attachments-with-the-guide-media-api-and-zapier/
Last updated: 2024-11-03T10:48:20.000Z
Over the last year Zendesk has been upgrading their attachments experience for Help Center articles in a major way.
When adding attachments, files or images, to articles it used to be that each attachment was linked to a single article. You could only see attachments when looking at an article, and if you wanted to add the same image to multiple articles you had them upload them again and again for each article.
Last year saw the release of the Media Gallery for the article editor. This central repository of files makes it possible to upload an image once, and reuse it other articles by selecting it from the gallery. Similarly you can replace an image, and every article where that specific image lives, will now show the newer version of that screenshot or photo.
## Media Library API
This new Media library is great, but there are scenarios where going into Zendesk and uploading your attachments is not very handy. In bigger companies it's often the marketing team that creates the *perfect* screenshots that can be used by the writers team in their articles.
You could give your marketing team access to the your Help Center, but thanks to the new [Media Library API](https://developer.zendesk.com/api-reference/help%5Fcenter/help-center-api/guide%5Fmedias/?ref=internalnote.com), we can allow uploads from external tools.
## Leveraging external sources.
Let's say your Marketing team creates your assets in a tool like Figma, and then uploads them all to a central Google Drive folder everyone can reference. You could go into the drive each time you need an image and upload it into the Media Gallery. But by combining the new API and a tool like Zapier or Make, we can automate this and automatically upload any new image on Google Drive right into Zendesk.
# How does the API work?
The Media API has a few endpoints that need to be called in the right order to upload attachments into Zendesk.
1. We first create an **Upload URL** that allows us to upload an image or file to Zendesk's storage.
2. We then upload the actual **file** to that endpoint
3. We then create a Guide media object which adds the upload file to the media gallery.
## Upload URL
To create an **Upload URL** we need to let Zendesk know what `file type` and `file size` we want to upload.
We do this by doing a `POST` to `https://{{domain}}.zendesk.com/api/v2/guide/medias/upload_url` with the following payload
```json
{
"content_type": "image/png",
"file_size": 12345
}
```
This returns the following data
```json
{
"upload_url": {
"asset_upload_id": "01J73HEF2PYN50PB9ZXYA1EBPF",
"url": "https://uploaded-assets-pod17.s3.eu-west-1.amazonaws.com/17/12154371/01J73HEF2PYN50PB9ZXYA1EBPF?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIARVNLH63KCBTH4T3L%2F20240906%2Feu-west-1%2Fs3%2Faws4_request&X-Amz-Date=20240906T112754Z&X-Amz-Expires=3600&X-Amz-Signature=8b4a372bf96148be65e8a27f9574f130c72ab2b14759eab828731cfc6c113355&X-Amz-SignedHeaders=content-disposition%3Bhost%3Bx-amz-server-side-encryption&x-id=PutObject",
"headers": "{"Content-Disposition":"attachment; filename=\"01J73HEF2PYN50PB9ZXYA1EBPF.png\"","Content-Type":"image/png","X-Amz-Server-Side-Encryption":"AES256"}"
}
}
```
We'll need these 3 elements in our next steps.
## Uploading our file
The previous step gave us a `url` to upload our file too. Since Zendesk uses AWS for its hosting and data storage we need to upload our file directly into one of their S3 buckets.
We do this by doing a `PUT` to the `https://uploaded-assets-pod17.s3.eu-west-1.amazonaws.com/17/12154371/01J73HEF2PYN50PB9ZXYA1EBPF?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Credential=AKIARVNLH63...` URL we retrieved in the previous step while adding our file as a `--data` element in our request.
Do note that we should include the `headers` that were provided in that first step too. Done correctly we get a status 200 back.
```
Content-Disposition:attachment; filename="01J73HEF2PYN50PB9ZXYA1EBPF.png"
Content-Type:image/png
X-Amz-Server-Side-Encryption:AES256
```
## Creating the Media Object
And finally we need to add the Media Object to our Media Library.
This is done by a third and final `POST` API call to `https://{{domain}}.zendesk.com/api/v2/guide/media` with the following payload:
```json
{
"asset_upload_id": "01J73HEF2PYN50PB9ZXYA1EBPF",
"filename": "screeenshot2.png"
}
```
Note that the `asset_upload_id` is again taken from the first step. The `filename` can be the same as our file we uploaded, or we can give a custom name here. This name will appear in the media gallery.
# Let's use Zapier
By using a tool like Zapier we can leverage this new API to automate the flow I talked about earlier in this article.
Here's the Upload folder our Marketing team uses to add assets for our Help Center.
Now let's take this folder into Zapier and automate uploading these images into the Zendesk Media Gallery.
## Step 1: Getting new images
The first step we need in our flow is the *New File in Folder* action. Select the folder you want to monitor and make sure to exclude the deleted files.
## Step 2: Request Upload URL
In the second step we use a *Webhooks by Zapier* step as a `POST` type to get an *Upload URL*.
For the URL, use `https://{{domain}}.zendesk.com/api/v2/guide/medias/upload_url`
We can use the variables from the first step to populate the data flow:
- `content_type`: *1\. Mime Type* (e.g. image/png)
- `file_size`: *1\. File Size* (e.g.123456)
For the authentication part we can use `admin@domain.com/token` then a `|` and then a `zendesk_api_token`. E.g. `admin@domain.com/token|abc123def456`
## Step 3: Uploading the file
Here things become a bit tricky since there's a lot of variables to set. We once again use a *Webhooks by Zapier* steps but choose `PUT` as the type.
- For the URL we choose the *2\. Upload Url Url.*
- For the File, select *1\. File (exist but not shown)*
For the headers:
- `Content-Disposition`: attachment; filename="*2\. Upload Url As set Upload Id*.png"
- `Content-Type`: image/png
- `X-Amz-Server-Side-Encryption`: AES256
## Step 4: Add to the Media Library
And at last, the final step.
We can use a final *Webhooks by Zapier* step to `POST` to the Add to Media Library endpoint.
For the URL use `https://{{domain}}.zendesk.com/api/v2/guide/medias`
For the data part:
- `asset_upload_id` use *2.Upload Url Asset Upload Id.*
- `filename` use either a custom name or *1\. Original Filename*
For the authentication part we can use `admin@domain.com/token` then a `|` and then a `zendesk_api_token`. E.g. `admin@domain.com/token|abc123def456`
# Putting it all together
Our marketing team just added a new movie poster to the Google Drive folder.
The Zapier flow picks up the newly added file and starts its run:
1. It detects the file
2. It creates an upload url for the file by using its file type and size
3. It uploads the actual file
4. And adds it to the media library
# Conclusion
It's nice to see how Zendesk is expanding on their Knowledge Base Content Management. They made changes to the editor allowing for more types of text, they restructured the way we add articles to sections and manage permissions and added a few powerful API tools to rewrite, simplify or tone shift your articles.
This Media Library is another example on how the Help Center is moving more towards the likes of Wordpress or other strong CRM systems.
Are there things missing? Sure! I'd love a way to get access to the Media Library from within the main Admin Panel, maybe as a new button in the left navbar.
Tagging images, seeing in what articles an asset is used, or uploading multiple locales of the same images in an asset bundle are just a few of the things I'd love to see.
But aside from those, this new API allows for some cool workflows as the one I showed above.
### Zendesk Roundup for September 2024
URL: https://internalnote.com/roundup-2024-09/
Last updated: 2024-11-08T07:37:08.000Z
A friend at an Australian Zendesk Partner reminded me this week that not everyone is having a summer holiday right now, and that for them it's actually middle of winter as we speak.
So in an effort following [trends](https://internalnote.com/zendesk-cxtrends-2024/) and personalizing the content of this blog to the reader:
🔘 I hope you're having a wonderful winter
🔘 I hope you're having an amazing summer
(check whatever applies).
And now, this months' updates!
# 🏢 Company
## AI Summit
As you'll note further down this overview, August has been a quiet month when it comes to updates to AI Agents and Zendesk Advanced AI. Many people being on holiday might be one reason, but the upcoming AI Summit is possible a bigger indicator.
Last years' AI Keynote [gave us](https://internalnote.com/zendesk-ai-drop-keynote/) tons of amazing releases for Zendesk with generative replies in bots, bot persona, similar tickets, AI for voice being just a few of last years' headlines.
For this years' keynote I've got a few items on my wishlist.
First off I really hope we will see a **general release of** [**Copilot**](https://internalnote.com/preview-copilot/) for Advanced AI users. The feature has been in beta for over a year now, and it feels the right time to release it for customers. Chances are big most other EAPs will see a general release too: merge suggestions, suggested tickets and [quick replies](https://internalnote.com/preview-of-the-new/).
Secondly I hope we'll see **improvements on Zendesk AI** **models**. Specifically I hope to see the option to train or add custom intents to your instance without needing to wait for Zendesk to update their entire model. [Ultimate](https://internalnote.com/zendesk-acquires-ultimate/), Zendesk's newest acquisition has this capability out of the box, so bringing this capability to (Suite Enterprise) customers someday is a given.
Thirdly I'd love Zendesk to **clean up their product naming**. AI Agents, Advanced AI, AI Agent Copilot, they're not memorable or clear names and don't feel part of one suite of products.
AI Agent encompassing both Ultimate and the Zendesk Bot is just plain confusing, and I hate – hate – the fact that by calling your bot an AI Agent, we're forced to call people handling tickets *human* agents. Anything that requires you to add human in front of a word either means you want to make it more friendly than it is or hide something bad.
If it were me I'd call it Zendesk AI Bot and Zendesk AI Workspace. The former being all the end-user deflection and automation, the latter anything that assist agents. Or Zendesk AI Agent and Zendesk AI Copilot, where the former is there to take over humans and deflect tickets, and the later works together with people to make their job easier.
The keynote is live in New York on October 8th, with a digital version available one week later, you can join via the link below!
[AI Summitvirtualevents.zendeskai-summit.com](https://virtualevents.zendeskai-summit.com/series/ai-summit/landing%5Fpage?lang=en&ref=internalnote.com)
## Outcome based pricing
Zendesk has always been a seat based product where you buy a certain amount of licenses to give a specific set of agents full access to the product. With each license you get a certain amount of [storage](https://internalnote.com/storage-limits/) and a shared pool of light agents and API calls.
Outbound WhatsApp messages over Sunshine Conversations and your Zendesk Talk usage were traditionally the only usage based elements on your Zendesk bill.
In a classic scenario this model works well. If you get 10.000 tickets a week you might need 5 agents to handle that workload. If your company grows and you get 50.000 tickets a week you might need 25 agents to handle that workload since without automation, deflection and self service, that rise is linear. More tickets means more work means more agents and thus more agent seats bought.
However once ticket deflection comes into play, and especially a modern AI powered strategy, the above scenario isn't applicable anymore. Those 10.000 tickets might be 5000 complex tickets and 5000 questions that can be fully handled by the bot with Help Center articles, automations or other strategies.
[A hybrid approach to AI Agents powered by Zendesk and UltimateIn a previous article I explained the Road to Automation and how it can help with automating more your support interactions. This article takes a real scenario and keeps improving the customer experience by leveraging a hybrid approach combining flows, generative replies and API integrations.Internal NoteThomas Verschoren](https://internalnote.com/hybrid-approach/)
I wrote a full article on a good AI powered strategy.
When your ticket load rises from 10k to 50k tickets, that might mean your complex agent tickets go up a bit, but chances are bigger that the majority of that increase in tickets are coming from those automateble, frequently repeated questions. Instead of increasing your team from 5 to 25 agents, you might only need to add a few more people to handle that new workload, while your bots take the brunt of the work.
### Automated resolutions
[Zendesk on LinkedIn: 💸 Zendesk is introducing Outcome-Based Pricing for AI agents:…💸 Zendesk is introducing Outcome-Based Pricing for AI agents: https://zdsk.co/4e08QXP 🤝 Our customers will only incur costs for successful resolutions…LinkedInZendesk](https://www.linkedin.com/posts/zendesk%5Fzendesk-is-introducing-outcome-based-pricing-activity-7234968093863636993--NyO?utm%5Fsource=share&utm%5Fmedium=member%5Fios)
This is where automated resolution come in. Instead of increasing the license cost of your agents to account for the new bot capabilities and AI calculations and work Zendesk does, they decided to implement an outcome-based pricing strategy.
In its most simple form it means that any customer's issue that's resolved by AI Agents without help from human agents is counted and will be billed in top of your Zendesk license costs.
This encompasses:
- A conversation bot
- An autoreply for intelligent triage
- An article recommendation on email, web form, or Web Widget (Classic)
You can look at this change in two ways. In a negative, critical way this can be seen as a surcharge on top of an already big software expense. On the other hand, since a good self service strategy lowers your ticket load and makes Agents' lives easier, leveraging bots and automatic resolutions will prevent a linear increase of agents and tickets, and thus Zendesk licenses as your company grows.
[Learn about automated resolutions](https://support.zendesk.com/hc/en-us/articles/6931689272090?ref=internalnote.com)
## Sign up for Internal Note
A blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 🎉 New Releases
## 🤖 AI Agents
### Messaging tag triggers
It's now possible to use Messaging triggers to add tags to conversations based on conditions. These tags can then be used for triggers, automations or adding conversations to [queues](https://internalnote.com/beyond-triggers-moving-to-queue-based-assignment-in-zendesk/).
## 👨🏻💻 Agent Workspace
### Rich messages in the Agent Workspace
> The Agent Workspace now supports the display of [form messages](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-step-types?ref=internalnote.com#topic%5Fil3%5Fpmj%5Ftvb) and [compound messages](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-step-types?ref=internalnote.com#topic%5Fmnf%5Fgwc%5Fk4b) options in the ticket interface. These are elements that customers interact with during their support journey, giving agents a more complete understanding of the customer experience.
Last month [saw](https://internalnote.com/roundup-2024-08/) the introduction of a new conversation view for Agent Workspace with rich rendering of Bot elements like buttons and carousels. This month we get another nice addition to the conversation view with a new way that forms are shown in the conversation.
### New supported channels for intelligent triage
Intelligent Triage, the engine that powers intent, language and sentiment tagging in Zendesk, now supports a myriad of new channels ranging from classic Facebook and Twitter posts to Sunshine Conversation channels like Apple Business chat and Messagebird. You can enable those new channels within the Admin Center, and newly added tickets should be triaged moving forward.
### Enhancements to skills in omnichannel routing
> Until now, omnichannel routing has treated all skills equally. Any skill on a ticket was treated as required until a [skills timeout](https://support.zendesk.com/hc/en-us/articles/4828787357210?ref=internalnote.com#topic%5Fag5%5Ftlr%5Fwbc) occurred. After that point, none of the skills remained in consideration when routing a ticket to an agent.
>
> Now, skills added by ticket triggers can be configured as either **required** or **optional**. Required skills are skills an agent absolutely must have, while optional skills are nice-to-have and subject to the skills timeout in your omnichannel routing configuration. When a skills timeout occurs and a ticket has both required and optional skills, only the optional skills are dropped from consideration. The required skills are preserved as routing criteria for the ticket, so omnichannel routing will continue to wait for an agent with all of the required skills to become available.
This is a welcome change to [Omnichannel routing](https://internalnote.com/an-intro-to-omnichannel-routing-in-zendesk/) in Zendesk. While we still want tickets to go to the best available agent, often when an agent that doesn't know everything about a product isn't available, it would be nice to be able to route – as a fallback – to another available agent in the system. By updating your existing skills as optional or required, you can now choose which one should be used upon initial routing, and which ones can be dropped if no matching agent is found.
## 🔎 Help Center
### Localized inline attachments
Inline article attachments on the Help Center [are now localized](https://support.zendesk.com/hc/en-us/articles/7761141483546-Help-Center-API-Localized-inline-attachments?ref=internalnote.com), linking them to a specific translation of an article rather than the original article itself.
The same image embedded in translations of an article will now create unique `article_attachment` records, each with its own unique ID and link.
```json
{
"article_attachments": [
{
"id": 21059268278418,
"article_id": 21059201642898,
"locale": "en-us",
"relative_path": "/hc/article_attachments/21059268278418",
},
{
"id": 21059234861202,
"article_id": 21059201642898,
"locale": "nl",
"relative_path": "/hc/article_attachments/21059234861202",
}
]
}
```
This makes it easier to update one image via API in a specific translation, without affecting the same images in other places.
### Media API
In extension to the above attachment changes, Zendesk has also been hard at work updating the overall experience managing media in Zendesk Guide. The new Media Library offers a central place to manage attachments for a while now, and this month saw the release of the new [Media API](https://support.zendesk.com/hc/en-us/articles/7821491779354-Announcing-Public-API-for-Media?ref=internalnote.com) allowing you to upload attachments from external resources directly into Zendesk.
This way your marketing team can create assets for your documentation and upload them, and your content managers can view these attachments in the Media Gallery and insert them into articles easily.
## 🧱 Open and Flexible Platform
### Sandbox replication
Users of Premier Sandboxes can now replicate support email addresses from their production environment to the sandbox when they create a new one.
- Internal support addresses (for example, [*help@yourbusiness.zendesk.com*](mailto:help@yourbusiness.zendesk.com)) are automatically replicated with the sandbox domain as the domain part.
- External support addresses (for example, [*help@yourbusiness.com*](mailto:help@yourbusiness.com)) are replicated but converted as internal addresses of the following format: [*help-at-yourbusiness-com@sandboxsubdomain.zendesk.com*](mailto:help-at-yourbusiness-com@sandboxsubdomain.zendesk.com).
### Solved ticket reassignment options
> Previously, when an agent was deleted, downgraded, or removed from a group, their solved tickets were automatically reassigned to the admin removing them or the longest active team member in the group.
>
> Now, admins have [the option to turn on solved ticket reassignment](https://support.zendesk.com/hc/en-us/articles/7807544385306-Announcing-solved-ticket-reassignment-options?ref=internalnote.com) in Admin Center. Solved ticket reassignment allows you to configure different behavior for how a departing agent’s solved tickets are reassigned.
Just one of those little tweaks that will make reporting correctly on data in a way that fits your company a lot easier.
### Additional condition for deletion schedules
Deletion schedules now allow for additional conditions on top of the *last updated* condition: Brand, Form, Group, Organization, Requester, Tags or Custom Fields.
This allows for a more granular approach allowing you to, for example, delete all finance tickets after 7 years, but delete support tickets after 4 years, and remove the ones categorized as *other* after a year. Perfect to keep your instance within its [allotted storage limits](https://internalnote.com/storage-limits/).
# ⚠ Major Changes.
A few important API and authentication deprecations coming up the next few months:
### Deprecation of the Zopim Chat REST API domain
> Zendesk is updating the domain for its Zopim Chat REST API. Beginning October 29, 2024, the current Zopim Chat REST API endpoint at [www.zopim.com/api/v2](http://www.zopim.com/api/v2?ref=internalnote.com) will be replaced with {subdomain}.[zendesk.com/api/v2/chat](http://zendesk.com/api/v2/chat?ref=internalnote.com).
[View the article](https://support.zendesk.com/hc/en-us/articles/7827476398362-Deprecating-the-Zopim-Chat-REST-API-domain?ref=internalnote.com)
### Deprecation of the Events Connector for Amazon EventBridge
> Beginning August 12, 2024, you won't be able to set up a new Zendesk events connection to Amazon EventBridge. Furthermore, on the deprecation and removal date, all Zendesk event connections will stop sending events to your Amazon EventBridge instance.
[View the article](https://support.zendesk.com/hc/en-us/articles/7755441158426-Announcing-the-deprecation-of-the-Events-Connector-for-Amazon-EventBridge?ref=internalnote.com)
### **End-of-life of legacy OAuth 2.0 grant type**
> In alignment with OAuth 2.0 best practices, Zendesk will stop accepting Implicit and Password grants for access tokens starting February 17, 2025\. Customers are advised to switch to Authorization Code Flow or API tokens as soon as possible due to the insecurity of the older grant types.
[View the article](https://support.zendesk.com/hc/en-us/articles/7868169712794-Announcing-the-end-of-life-of-legacy-OAuth-2-0-grant-types?ref=internalnote.com)
# 💡Insights
## Interview with Matthias Goehler
[Zendesk: How AI Chatbots can Transform Customer ServiceWe speak with Matthias Goehler, Zendesk’s EMEA CTO, about how AI-powered virtual assistants are providing fast, personalised service at scaleBizclik Media LtdMarcus Law](https://aimagazine.com/articles/zendesk-how-ai-chatbots-can-transform-customer-service?ref=internalnote.com)
Interesting interview on AI Agents with Matthias Koehler, Zendesk's European CTO.
> We should always start by thinking about the customer we’re serving, their requirements and demands,” he says. “When we reach out to a customer service team, we probably have a question, problem, or complaint – and we want an answer or help. Fundamentally, we want it fast, and that has never changed. We want the answer on the first call; we hate repeating ourselves or going through endless processes with handovers.
## Deflecting Zendesk Spam
Dominic does with video what I try to do with text: explaining Zendesk concept in a simple way without shying away from complex topics.
I really like his recent video on Zendesk spam, and really appreciate the shoutout to my [original article](https://internalnote.com/automatically-deflect-zendesk-spam-tickets-via-triggers-and-web-hooks/) showing the concept. Go watch his videos and subscribe to his channel!
# 📝 Articles this month
[Sending out automated pending reminders via Zendesk MessagingThis article will show you how you can send out automated reminders over Messaging to your customers when your agents are awaiting a response.Internal NoteThomas Verschoren](https://internalnote.com/bump-bump-solve-for-messaging/)
[No more no-replyHave the guts to remove your no-reply. It blocks your customers and creates a terrible experience. Replace it with self service, automations and routing.Internal NoteThomas Verschoren](https://internalnote.com/no-noreply/)
[Expanded multi-select field support for users, organizations and Custom objects in ZendeskMulti-select fields are now available for users, organizations and custom objects in Zendesk. This makes setting up user profiles and triggers way easier, while giving also a better overview of data when looking at those objects in the Agent WorkspaceInternal NoteThomas Verschoren](https://internalnote.com/multi-select-support-for-users-organizations-and-custom-objects-in-zendesk-2/)
[Zendesk Roundup for August 2024Subscribe to a hand-picked round-up of the best Zendesk links every month. Curated by Thomas Verschoren and published every month. Free.Internal NoteThomas Verschoren](https://internalnote.com/roundup-2024-08/)
# And Finally...
Good to know:
[Can I map the sender email domain of the automated notifications from Guide?](https://support.zendesk.com/hc/en-us/articles/7581210819226-Can-I-map-the-sender-email-domain-of-the-automated-notifications-from-Guide?ref=internalnote.com)
> The emails that notify customers of Guide or Gather updates aren't customizable. They use a standard format and are always sent using the address noreply@{subdomain}.zendesk.com.
## Sign up for Internal Note
A blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Sending out automated pending reminders via Zendesk Messaging
URL: https://internalnote.com/bump-bump-solve-for-messaging/
Last updated: 2025-09-08T06:41:02.000Z
Last year I wrote an article about sending messages to customers over Messaging (WhatsApp, Web Widget,...) via the Sunshine Conversations API. This flow not only allows you to [bypass the 24 hour rule for WhatsApp](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/) but also makes it possible to send messages to customers via triggers and automations.
[➕ Sending Automated Messages via WhatsApp with Sunshine Conversations and ZendeskLearn how to leverage the Sunshine Conversations API included in Zendesk Suite to automatically send out messages to your customers.Internal NoteThomas Verschoren](https://internalnote.com/sunshine-conversation-automations/#messages-via-webhooks)
A while back I had a reader reach out who wanted to convert my *Pending* automations for custom statuses ([link](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/)) into a flow compatible for messaging.
So, this is exactly what this article will be about. But first, some context.
# What are *Pending* automations
When an agent replies to a customer ticket, they can use the *Pending* status when submitting the ticket when they need a response from the customer before they can move forward.
This could be a confirmation that the issue is resolved, a request for more information, or just a reply for some clarification on the issue.
Putting the ticket on *Pending* has a few benefits:
- It's removed from the [Open Tickets](https://internalnote.com/my-approach-to-zendesk-views/) view.
- Your [SLAs](https://internalnote.com/sla-policies/) are put on pause.
- You can [automate](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/) these tickets and send out an email reminder to the customer that you're still waiting for their response.
Those reminders make sure your agents get the responses they need, and if no response is given, we can either *reopen* the ticket so the agent can do one final attempt to reach the customer, or an automation or the agent can manually or silently *solve* the ticket to remove it from our active tickets altogether.
This works great for email based tickets like those submitted via email, web forms or API. But when it comes to Messaging tickets via the Web Widget or WhatsApp we run into a few issues.
For most customers we do not have an email address on file, so our automation, which can fire on a messaging ticket, will not work and will not send out an email.
Messaging also has different rules. A Web Widget conversation times out after [72 hours](https://support.zendesk.com/hc/en-us/articles/7463958274586-Why-do-messaging-conversations-contain-old-and-previous-messages?ref=internalnote.com) and WhatsApp is even stricter with a 24 hour limit.
So to resolve this we can build our automations a bit different, and leverage the Sunshine Conversations API and a Cloudflare Worker to set this up.
# Setup
## Worker
As explained in the [original article](https://internalnote.com/sunshine-conversation-automations/#messages-via-webhooks), the Sunshine Conversations API requires some identifiers that are not natively accessible in Zendesk triggers and automations. We need a `conversation id` and a `user identity` . We can derive both from the `requester_id` of our ticket requester, but even so, we can't connect to the Sunshine Conversations API directly from our triggers.
For this reason we need to deploy a Cloudflare Worker (or similar cloud script on make.com, zapier, Google Cloud or similar). You can find the worker in the GitHub repository below, or you can deploy it to Cloudflare **for free** via the Deploy button below.
[GitHub - verschoren/outbound\_messagingContribute to verschoren/outbound\_messaging development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/outbound%5Fmessaging?ref=internalnote.com)
[](https://deploy.workers.cloudflare.com/?url=https://github.com/verschoren/outbound%5Fmessaging&ref=internalnote.com)
## Webhook
Secondly we need a webhook that calls our worker (or other cloud function). Assuming you deployed the Cloudflare Worker, your webhook will be hosted on `https://outbound-messages.yourdomain.workers.dev`
Example of a configured webhook
## Automation
Assuming we want to send out a reminder to our customer after eight hours, we can setup an automation with the following conditions:
💡
Note: I made use of a custom ticket status in this automation. **Pending - Awaiting Reply*. Once the automation runs it will change the custom status to another status so the automation runs only once!
Once the conditions are setup, we can configure our actions:
The payload.
```json
{
"requester":"{{ticket.requester.id}}",
"message":"We marked this conversation as pending and are awaiting your answer"
}
```
# The Result
A customer contacts our customer care team and the agent needs more information
The customer does not respond timely, so our automation fires, and the customer replies
Our agent gets a reply from the customer
Or via WhatsApp:
# Conclusion
I really hope that someday the Zendesk triggers and automations can access the Sunshine Conversations API directly without resorting to an external script to grab the right variables.
Or, since we're hoping, a way have a *Send message* action in our triggers that can directly send a message to Messaging channels would be even nicer.
Either way, for now the above flow will allow you to send out automated reminders over Messaging to your customers!
### No more no-reply
URL: https://internalnote.com/no-noreply/
Last updated: 2024-11-03T10:49:05.000Z
A few weeks ago I went to an event from Cloudflare in Amsterdam where they showcased their newest product features combined with a few presentations from their team and customers. After the event I got an email from them asking me to fill in a questionnaire about the event. That questionnaire contained an error which prevented me from completing it.
So I went back to the email, clicked the reply button to let them know something was wrong. Only to be stopped midway my first sentence since the sender of the feedback email was ``.
> When I phone you night and day
> I get no reply, no reply
> I keep writing those letters and send them away
> To get no reply, no reply
Similarly a few years ago I had a customer whose head of CX showed us a 99% first time resolution (FTR) metric for their team. Which is an astounding number (the industry average lies [around 70%](https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.zendesk.com/blog/first-contact-resolution-friend-foe-frenemy/%23:~:text%3DA%2520first%2520contact%2520resolution%2520rate,strive%2520to%2520hit%2520that%2520number.&ved=2ahUKEwiMo5jKkNGGAxVJR6QEHRqiD6YQFnoECBMQAw&usg=AOvVaw0ms335XulmW1DMdEXXz75u)). When we dove into the numbers and reporting we discovered two things: they replied to customers from a noreply@company.com email, and that inbox was not forwarded or read by anyone in the company. When we reached out to the IT team to check the logs for that mailbox we quickly discovered the inbox contained thousands of unread emails and the average FTR was around 45%.
A quick search through my own inbox for the last month gave me \~200 noreply emails ranging from marketing emails, order confirmations, invoices and a range of other business and online sales related emails from a myriad of companies.
[](https://www.threads.net/@tomcashman1/post/C8ZE%5Fg6SQ7c/?ref=internalnote.com)
Source: Threads
And this makes me wonder..
> Why would you send email to a customer from a noreply email address?
## Assumption: we don't expect customers to reply to this
One motivation I often hear is: "this is a transactional email, we don't expect customers to reply to this".
In the case of an order confirmation, payment receipt, renewal alert or other transactions you often see the email coming from a noreply address. But what happens if there is a mistake in the email? Imagine the customer booked for the wrong date. Or the customer doesn't want to renew their service. Or they get an order receipt and see it has the wrong ship-to address.
A noreply email smells like arrogance to me. It separates the company from a customer and makes the customer think the company isn't interested in your feedback or comments.
Why would you want to prevent the customer from replying to the email to fix the issue? By using a noreply email you force the customer to go search for your support page, contact email or other channel and create a brand new ticket, thereby preventing your agents from having all the context they need to solve the issue.
The original email contains all the information and context required to solve the ticket. So instead of sending out these emails from noreply@, send it from an email address managed by the team that can resolve complains and questions about the email. If customers have questions they'll reach out to you anyhow.
💡
Sending out transaction email from a noreply means that agents will lose context when a customer reaches out about that transaction.
## Assumption: We only sent out the email. It's not our job to handle the replies.
*Or on other words: the other department send out these emails, why should be get the replies and customer feedback for those emails?*
For me, any company who emails with a noreply email is a company whose departments are silo'd.
If marketing sends out a "Summer Promotion" email from their noreply@company.com email, instead of emailing with a normal reply-to address, that means that any feedback from customers – I just bought X, can I use the promo, can I apply the promo to Y,...– will not have a team to handle those questions.
In parallel, if you allow replies to those emails, you get measurable feedback. You can collect customer replies to these emails, you can see what went wrong, or even amend your FAQ to make things clear for the next customer.
The irony of asking for feedback while not allowing customers to reply to that email
💡
Sending out email without reflecting where replies to the email should go to means there is no owner of this process or project.
## Assumption: if we put up a barrier the problem will go away.
Customers who have an issue with their order will reach out to support to get it fixed. Sending out the confirmation email from noreply will not make that problem go away. It will just put up a barrier between the customer who has an issue, and the team (your team) that can resolve that issue.
> There's a variation of this noreply flow: the dreadful "this mailbox is not monitored email".
The above is even worse. Sometimes you get an email from `` but when you reply you get an automated response to let you know this email address is not used. You know someone took the time to setup that deflection on purpose. The same effort could be done to forward the email to the team that handles my inquiry, but someone in the company decided that no, deflection is the best approach here.
💡
Customers will always have questions. Instead of putting up barriers you should offer clear self service options to handle inquiries without channel switching.
# A better way
I think the best approach is the one where every email you send to your customers allows your customers to reply to.
This does not mean every email should land in the inbox of your team. We can leverage features like auto reply, intelligent triage or even links to related articles right in the original email to make sure customers can try to resolve their issue themselves first. Any reply that can't get resolved should land in the inbox of the *right* team.
Take a look at this example from, coincidentally, my favorite Belgian beer.
- The reply-to is an email that goes to customer care
- The email has all the references like order number and shipping number right in the email body
- There's a link to my account to get status update or make changes.
Also, Duvel, if you're reading this: I'd love to setup Zendesk for your team! 😇
So what happens if I have an issue? I'd go to my account first and try to modify the order. If that doesn't work I can just *reply* to the email and get support from an agent. You can't get it easier than that!
> As a company, you should have the guts to disable your noreply emails and make sure every email arrives somewhere. Self Service and ticket deflection are the best way to handle customer feedback. But trying to prevent it by using a noreply address is just bad practice.
## Sign up for Internal Note
The blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Expanded multi-select field support for users, organizations and Custom objects in Zendesk
URL: https://internalnote.com/multi-select-support-for-users-organizations-and-custom-objects-in-zendesk-2/
Last updated: 2024-11-03T10:49:26.000Z
Zendesk tickets have always been able to store additional data in ticket fields. You've got drop-down, text fields, checkboxes and myriad of other options to store order numbers, preferences, or other related information.
Similarly, you can expand user profiles and organizations in Zendesk with similar user and organization fields to store things like user types, account numbers, membership types or other relevant data to enrich your profiles.
And Custom objects allow you to create completely custom things in Zendesk to store data specific to your organization like products, contracts or orders. These records too can be enriched with custom fields to store data like colors, types or IDs.
Even though all these fields allow you to enrich Zendesk objects, not all objects are equal. Up till now for example you could only use Multi-select fields on ticket fields for example.
# Multi-select fields
There's a few big types of custom fields:
- **Input fields** that allow for *free* text input.
Text, integer, decimal, regex or textfield, date
- **Checkboxes** that are a binary on/off
- **List of options** that allow you to select from a preset list of options.
Drop-down, Multi-select or Lookup fields
Multi-select has now been made available across all object types in Zendesk. So let's dive into some use cases that are now possible.
## For Users
User fields are often used to categorize end-users.
This can be used:
- to set SLAs (e.g. VIP users get higher priority)
- in triggers (e.g. partners who often email you don't get a confirmation email upon ticket creation)
- to filter Lookup fields (e.g. the Lookup field "Supplier" only shows users with the option "supplier" selected)
Since users often can have multiple roles, adding user roles to a dropdown wasn't ideal. The only possible solution was to create a checkbox per user type to user profiles.
With the availability of the new Multi-select field for end-users, these checkbox can now be combined in one Multi-select dropdown. This has a few benefits: of you previously had checkboxes to note down user types, but also had checkboxes for other use cases like "don't send CSAT email" or "VIP" or "Membership type", it's now a lot cleared which options pertain to your user type since they're all combined in one dropdown.
💡
If your checkboxes contain a tag, an easy way to migrate to Multi-select fields can be handled by noting down the checkbox tags. You then delete the checkboxes, and create a Multi-select dropdown with those same tags for your options. The new Multi-select dropdown will automatically select the right options based on the tags on your user profiles.
Additionally, since we now use a Multi-select option, our [Essential Card](https://internalnote.com/essentials-card/) is now a lot nicer to look at too:
## For organizations
A similar scenario is now also possible for Organizations. Where we previously would have used drop-down or checkboxes to note down the type of organization we're working with, we can now use a single Multi-select dropdown to note down its type or types.
And similarly, if we want to create a trigger that acts upon the organization type, we can now find all possible types in one dropdown action in the trigger conditions. This makes it a lot easier to select one or more options via the **ALL** or **ANY** conditions and removes the need of looking through a long list of organization fields in order to select some or all available options.
## For Custom objects
In a previous article I build a `Movie` object to store information about sold movies to later use in Placeholders.
One *issue* I had when creating the movie object is that each movie could only have a single genre associated to the movie. But sometimes movies (or other objects) can have multiple options associated with it.
Superman can be both a bird and a plane. A dress can both be black and blue, and Bulbasaur is both Grass and Poison type.
With the expansion of Multi-select compatibility across all objects in Zendesk, we can now add those fields to Custom objects too.
When we create our multi-select field and update our records, we can now create Movie records that contain multiple genres, and use these options in placeholders, object triggers or just as additional context for our agents.
Dropdown on the left, Multi-select on the right.
# What's still missing
Although these new capabilities are pretty awesome, there's still a few things missing in Zendesk when it comes to multi-select fields.
## Support in the Messaging widget
When creating [custom answers](https://internalnote.com/flow-builder-ask-for-details/) for the Zendesk Bot, the *Ask for Details* step only supports regular dropdown fields, but there's no support for multi-select fields like we have in regular Zendesk forms.
Speaking of feature parity, only text fields and drop-downs are currently supported for Messaging, with not only multi-select but also checkboxes, numeric fields, regex fields and all other types being incompatible with your bot flows.
## Support for importing multi-selects
When I migrated my movies from single genres to multiple genres I tried being smart about it.
I still had my import file with my movies, so I updated that file to allow for multiple genres by comma separating the tags for those genres. Sadly, when I imported (and updated) the movies via the Data Imported I got errors for all of them. It seems importing multi-select fields is not yet possible.
## Multi-select Lookup fields
The [announcement](https://support.zendesk.com/hc/en-us/articles/7537006572058-Announcing-support-for-multi-select-custom-fields-for-users-organizations-and-custom-objects?ref=internalnote.com) of the expanded Multi-select support contained the following paragraph:
> *This enhancement provides greater flexibility and the opportunity to collect more detailed data categorization, making it easier to map your customer-relationship management data in Zendesk. (...)*
> *For example, you can now use multi-select custom fields to track information such as:*
> *\- Products a customer has purchased*
This, briefly, got my hopes up, cause I thought it meant we could link multiple custom objects to tickets, users or each other. But sadly, Lookup Fields are still a 1:1 relationship. So if you want to link multiple products to a single user, it's currently only possible if you have a dropdown with products. But you can't have a Lookup field and select multiple linked Custom Objects.
## Sign up for Internal Note
The blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
##
### Zendesk Roundup for August 2024
URL: https://internalnote.com/roundup-2024-08/
Last updated: 2024-11-03T10:49:38.000Z
> Just when I thought I was out, they pull me back in.
Summers are supposed to be quiet months. Time for travel, family, relaxing. But not so if you write about Zendesk it seems. The amount of new releases this month was staggering, with major changes in bot translations, custom object placeholders and conversation views in Agent Workspace.
This month I also had the chance to play around with Zendesk's new [AI Co-pilot for Agents](https://internalnote.com/preview-copilot/). You can read my full review in the linked article but if you want a concise conclusion this one will do:
> The efficiency gains enabled by Co-pilot are immediately visible. For every procedure defined for Co-pilot, our agent only needs to verify the proposed actions are correct, and they can submit or approve the next steps for the customer.
Next week I'll be on holiday in Italy, but no worries, I've got a few fun articles all done and scheduled to be send out to you!
Let's dive into this month's releases.
# 🏢 Company
### Quantifiable impact of Zendesk AI
Zendesk's been quite active on their own blogs this month prompting Zendesk AI.
They published [an independent review](https://www.zendesk.co.uk/blog/the-quantifiable-impact-of-zendesk-ai/?ref=internalnote.com) of Zendesk AI and how it impacts organizations. It's a short 4 page read, so worth your time, but the key takeaways of the research are nothing new but it's nice to see them summed up in one document:
- **Efficiency Gains**: Zendesk AI significantly improves efficiency by automating routine tasks and streamlining customer service operations. This leads to faster resolution times and reduced workload for support agents.
- **Customer Satisfaction**: higher CSAT by providing quicker and more accurate responses. AI-driven support ensures that customers receive timely assistance, which improves overall customer experience.
- **Cost Savings**: Zendesk AI creates a substantial cost savings for companies. Automation reduces the need for a large support staff, cutting down on labor costs while maintaining high service quality.
### How AI redefines the role of customer service agents
In this [newsroom article](https://www.zendesk.com/newsroom/articles/zendesk-head-of-ai-and-customers-weigh-in-on-how-ai-redefines-the-role-of-customer-service-agents/?ref=internalnote.com) Cristina Fonseca, Head of AI at Zendesk, explains how AI will impact Zendesk AI and evolve in the future. The article also contains some insights and quotes from other pundits of the Zendesk ecosystem and contains some nice nuggets I'll certainly use as the basis for future articles.
> “As AI continues to evolve, we anticipate a significant shift towards fully automated resolutions within the customer experience sphere. It’s a future where AI doesn’t just assist, but leads the charge in providing exceptional experiences. The latest breakthroughs in AI and large language models (LLMs) are paving the way for a fluid integration of bots and human agents. AI will influence at least ten times more interactions than in 2023, propelling the industry towards a future dominated by fully digital agents.”
# 🎉 New Releases
## 🤖 AI Agents
### Updates to the Zendesk AI Models
Zendesk works with a pre-trained AI Model per industry for all of its customers. This means that, out of the box, you get a working classifier model that assigns intents to tickets without the neat of training on your tickets. This gives you the benefit of immediate results when you enable the feature.
The downside of this approach is that if an intent is missing, you're stuck. You can't teach the system that they should categorize tickets related to a specific topic.
To solve this issue, Zendesk allows customers to request new intents. Customers can fill in a request form with an intent description and sample tickets. Once approved (or if approved..) the intent gets added **to all Zendesk customers**' setups.
This month saw the first release of about a dozen new intents across the retail, software, insurance, financial, employee experience, travel, and entertainment and gaming industries.
Additionally, next to the industry specific models available, Zendesk is also adding a general Zendesk Intent Model available. This model contains *all* industry-specific intents and customers who get this model enabled can pick and choose from the list of intents and enable the once they need.
I still hope for the availability of fully custom models for Zendesk Intelligent triage, that doesn't require Zendesk to manually accept and approve intents for all of their customers. Ultimate, Zendesk's newest AI Agent, has this feature for their Chatbot and Ticketing automation, so migrating this capability to Zendesk's intelligent triage is something I can't wait to happen.
### Custom translated messages in the Zendesk Bot
The Zendesk Bot used to have only automatically translated messages for when you enabled multiple languages in your bot. This gave the convenience of enabling a new language with just a checkbox, with the downside of not being able to control the translation. If the translation went bad and nuance or meaning went lost in translation, you couldn't really fix it.
This has now been fixed with the ability to add custom translations to your Zendesk Bot's reactions and answers. You can read all about it in last weeks' article:
[Managing custom translations in the Zendesk BotThe new custom translations for the Zendesk Bot make it possible to combine automatic translations for all your flows and answers, while making sure that specific replies or words do not lose their context in a bad translations.Internal NoteThomas Verschoren](https://internalnote.com/managing-custom-translations-in-the-zendesk-bot/)
Speaking of translations, the automatic translations have been [expanded](https://support.zendesk.com/hc/en-us/articles/7728852470170-Announcing-additional-languages-supported-with-Zendesk-AI-agents?ref=internalnote.com) with eight new languages: Chinese (Traditional), Czech, Danish, Indonesian, Norwegian, Romanian, Swedish.
### Better handling of handover to email
Continuous conversations is one of those awesome hidden features that make Zendesk Messaging a lot more efficient. Since Messaging is asynchronous by design, and it's not a live chat in the classic sense, it allows customers to interact with your AI Agent or leave a message even when agents aren't available yet. Or customers might leave mid conversation, and close their browser (or lock their phone).
With Continuous conversations Zendesk will send out your agents' last messages over email to the customer, allowing customers to continue the conversation via email, or switch back to the widget at their convenience. This feature used to auto-trigger after 10 minutes of inactivity, but will now fire once a conversations becomes inactive, which could be less than that time. It's a small change, but seems more logical than a set timeframe?
## 👨🏻💻 Agent Workspace
### Zendesk AI new uses ChatGPT-4o
Zendesk [upgraded](https://support.zendesk.com/hc/en-us/articles/7711631447450-Announcing-enhanced-generative-AI-features-in-Agent-Workspace-with-ChatGPT-4o?ref=internalnote.com) the generate AI features in Agent Workspace from ChatGPT-3.5 Turbo to the newly released ChatGPT-4o.
> This update will offer the following benefits:
> \- improved response accuracy.
> \- Reduction of language errors in expansions and tone changes, as well as accuracy issues with ticket summarization.
> \- Mitigation of errors related to max ticket length when summary is used, due to ChatGPT-4o's larger context window.
> \- Enhanced feature performance, ensuring efficient agent operations.
### Enhancements to the omnichannel routing skills timeout
If you've got Omnichannel Routing enabled in Zendesk and choose to route tickets based on skills, tickets will be assigned to available agents if their skills match.
If no available agents have that skill, the ticket is added to a queue waiting for someone to become available. Since this waiting can be indefinitely, you've got the option to enable a timeout which drops the skills requirement and assigns tickets to any available agent that matches the group, queue or other secondary requirements.
Until now, agents would only be eligible for assignment after timeout if they did something in Zendesk, like changing their status or updating a ticket. This has now been changed and agents are added to the list of potential assignees all the time if their availability and capacity allows for it. This means more agents will be available faster, leading in, hopefully, faster assignment and a faster FTR.
[An introduction to Omnichannel Routing in ZendeskThis article will give you an overview of Zendesk’s Omnichannel Routing, Agent Availability and the brand new Queues features.Internal NoteThomas Verschoren](https://internalnote.com/an-intro-to-omnichannel-routing-in-zendesk/)
### New dashboard for Queues
For customers using [queues](https://internalnote.com/beyond-triggers-moving-to-queue-based-assignment-in-zendesk/) in Zendesk, there's now a new Dashboard in Explore that gives live insights in the amount of tickets and average time in queue for tickets. It's a basic dashboard that only shows a few datapoints, not a full dashboard that allows you to dive into the queues themselves and inspect tickets or act upon them.
[Announcing new prebuilt live Explore dashboard for omnichannel routing queuesAnnounced on Rollout on July 8, 2023 July 8, 2023 The Explore product team is excited to announce a new addition to our collection of prebuilt dashboards: Omnichannel routing queues - live mon…Zendesk help](https://support.zendesk.com/hc/en-us/articles/7569028703258-Announcing-new-prebuilt-live-Explore-dashboard-for-omnichannel-routing-queues?ref=internalnote.com)
### Generative AI for Voice available
I [wrote about this feature](https://internalnote.com/preview-of-the-new-generative-ai-for-voice/) a while back, but Zendesk made their Generative AI for Voice available to all Advanced AI customers this month. The feature enables three new features:
- **Call transcription**: Automatically convert call recordings into text and save it to the ticket conversation log for added context after a call ends.
- **Call summarization**: A concise, AI-generated summary of the call transcript can be automatically added to the ticket conversation log after the call has ended.
- **Voice QA:** For customers that have Zendesk QA (Klaus) this release enables the tool to generate reports based on phone conversations similar to how it already works for email or messaging.
### Custom layouts and Essentials cards in Professional plans
Custom Layouts and Essentials cards used to be available for Enterprise users only, but these features are now also available to Suite Professional users.
[Preview of the new Essentials Card in ZendeskA quick overview of the new Essentials Card for Zendesk user profiles.Internal NoteThomas Verschoren](https://internalnote.com/essentials-card/)
### New conversation view
This month saw the rollout of the new conversation view for all customers. This new layout shows a more chat-inspired layout to customer conversations collapsing multiple short replies into a single bubble, and having an overall rounded softer look.
Additionally, this new layout also supports [carousel](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-step-types?ref=internalnote.com#topic%5Fil3%5Fpmj%5Ftvb) and [quick reply](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-step-types?ref=internalnote.com#topic%5Fmnf%5Fgwc%5Fk4b) messaging options, rendering those flow builder elements in a similar fashion as they would appear in the web widget. This feature is enabled by default, and requires no admin actions.
## **🔎 Help Center**
### Help Center Themes version 4
Every Zendesk customer gets the Copenhagen help center theme as part of their Guide setup. Most customers quickly move towards a [custom theme](https://www.zendesk.com/marketplace/themes/?ref=internalnote.com) from the Zendesk Marketplace, or even forgo these options all together and build their own with a [fully custom template](https://developer.zendesk.com/documentation/marketplace/building-a-marketplace-theme/build-and-test-your-theme/?ref=internalnote.com).
💡
I'm working on an article on this topic.
Whichever option you choose, Zendesk now [made a version 4](https://support.zendesk.com/hc/en-us/articles/7538459259162-Announcing-Help-Center-accessibility-update-and-templating-API-v4?ref=internalnote.com) available with some major changes:
- **Accessibility:** more components of the themes are now powered by the same components as Zendesk Agent Workspace and are build in their Garden design system, based on react. This gives better accessibility support (think voice over, alt text, navigation, labels, colors) in this themes that support it.
- **Redesigned forms:** The request forms are now styled in the Zendesk Garden style and are client side rendered, making a lot more customization options than before available.
- **JSON serialization helper** that makes it easier to use server-side rendered data from the templating api in the frontend with JavaScript.
- **Security**: by default the credit card field in the request form will only accept the last four digits of the credit card to ensure PCI compliance of the request form when using the credit card field.
Before (left) and after (right)
### New article settings layout
Continuing the work on the Guide editor these last few months, July saw a consolidation of all article settings in a new sidebar with two sections. One contains all edit permissions, the other contains placement and visibly of the articles.
I do like this renewed focus on the article editor. These new changes allow you to collapse all settings into the sidebar and leaving only the article editor available. this makes for a cleaner and more focused writing experience.
It's still not as easy as e.g. Ghost, which I use for my blog, with its `/actions` but compared to the editor of old, this new Guide experience is nice!
## 🧱 Open and Flexible Platform
### Multi-select custom fields for users, organizations, and custom objects
Previously, the multi-select custom field type was only supported for tickets. With this new update the multi-select field can be used for custom fields on users, organizations, and custom objects.
### Placeholders for custom objects
Speaking of custom objects, placeholder support has been expanded to support Custom Objects and fields stored in custom objects.
You can read all about it here:
[Custom Object Placeholders and Object Triggers in ZendeskThis article explores the new capabilities for Custom Objects with the new support for object triggers and dynamic placeholders.Internal NoteThomas Verschoren](https://internalnote.com/custom-object-placeholders/)
### Filtered Search for Custom Objects
In my [Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/) series I *complained* about the lack of full search capabilities for Custom Objects. The standard search returns any object that contains your search query in any of its objects. So searching for `Scott` in a movie data base would not only result in *Scott Pilgrim vs. The World*, but would also return *Gladiator* and *Top Gun*, since these are directed by Ridley *Scott* and Tony *Scott* respectively, and search also parses the text of any custom field.
The new [filtered search endpoint](https://developer.zendesk.com/api-reference/custom-data/custom-objects/custom%5Fobject%5Frecords/?ref=internalnote.com#filtered-search-of-custom-object-records) allows for custom queries that narrow the search to a specific field and object.
For example:
```JSON
//POST https://{{domain}}.zendesk.com/api/v2/custom_objects/pokemon/records/search
{
"filter": {
"custom_object_fields.type": { "$eq": "fire" }
}
}
```
returns an array of Pokémon whose type is fire:
```JSON
{
"custom_object_records": [
{
"id": "01HD3EAPK9FDA3RM3RPCEXF8B8",
"name": "Quilava",
"external_id": "156"
},
{
"id": "01HD0W1D6RJCQPSKGSXJ4HRME7",
"name": "Vulpix",
"external_id": "37"
},
// .... //
],
"meta": {
"has_more": false,
"after_cursor": null,
"before_cursor": null
},
"links": {
"prev": null,
"next": null
},
"count": 20
}
```
### New agent permissions to manage other team members
A new option [has been added](https://support.zendesk.com/hc/en-us/articles/7536595614746-Announcing-new-agent-permissions-to-manage-other-team-members?ref=internalnote.com) to the custom roles options:
> On Enterprise and Enterprise Plus plans, agents in custom roles can now be granted permission to view and manage other team members with the following options: *Not Allowed*, *View only*, and *Create, assign roles, edit, and delete*.
>
> These new options enable agents to assign roles to other agents but not themselves. Agents with these permissions also can't manage assignment to admin roles. These permissions are separate from permissions for creating and managing custom roles.
>
> Additionally, agents can also be granted permission to search and view lists of end users, separately from their permission to access individual end user profiles. With this new permission, agents will be able to search for agents by name, email address, phone number, or organization.
### Enablement of all Sunshine Conversations channels in the channel name trigger
> We are enabling all channel types in the Channel Name trigger. Previously, only messaging channels were enabled, but now, we are expanding this functionality to include all Sunco channels. This means that any channel type managed through the Sunco dashboard or API, such as Twilio, will be available in the Channel Name trigger dimension.
### New authentication configuration options
You've always been able to choose between Zendesk authentication, and External authentication when setting up agent or end-user authentication. Weirdly, if you wanted to only offer social logins but didn't want end-users to login with a Zendesk username/password you couldn't do that.
This has now been updated: you can setup Zendesk authentication, social logins (X, Google, Microsoft, and Facebook) and external SSO (SAML, JWT, ..) as independent options now.
### Announcing the ability to merge organizations
This used to be available as an EAP, was then pulled due to security risks, and is now available for everyone: you can now merge one organization into another.
When organizations are merged, all users, tickets and domains are merged into one organization.
There's also a [public API](https://developer.zendesk.com/api-reference/ticketing/organizations/organizations/?ref=internalnote.com#merge-organization-with-another-organization) available for merging organizations. After a successful merge you can check the action via a new getOrganizationMerge API:
```
// GET /api/v2/organization_merges/{organization_merge_id}
{
"organization_merges": [
{
"id": "THX1138QWERTYUIOP",
"loser_id": 123,
"status": "complete",
"winner_id": 456
}
]
}
```
## 📊 Reporting and Insights
### Advanced SLA configuration settings
This is a totally unexpected but awesome upgrade to SLAs!
Previously a first reply time would only exist if there wasn't a first public comment to a customer. Now, with the new advance options we can start or ignore the first reply time SLA even when tickets are created with internal notes, or created by light agents.
Similarly, we can resolve our first time reply SLA whenever an agent replies, public or internal, or any combination thereof.
These changes are optional, and the defaults you get are the behavior [we're used to](https://internalnote.com/sla-policies/), but you can dive into the settings of any SLA and make changes we're needed:
## Sign up for Internal Note
A blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 💡Insights
### What exactly is an AI Agent
[What exactly is an AI agent? | TechCrunchRegardless of how they’re defined, the agents are for helping complete tasks in an automated way with as little human interaction as possible.TechCrunchRon Miller](https://techcrunch.com/2024/07/13/what-exactly-is-an-ai-agent/?utm%5Fsource=dlvr.it&utm%5Fmedium=threads&guccounter=1&guce%5Freferrer=aHR0cHM6Ly93d3cudGhyZWFkcy5uZXQv&guce%5Freferrer%5Fsig=AQAAAMuGAJMmWRuV5Bs171DEmvQaQGk%5F%5Fs2PBJsn4F63h3Cr5oA0NUHENrWE63GbbZ8jNyHs1jARd0kVgefC8t2%5F1dhLFuv6H87Sn32en3Cz391hi7z9ZNcyMy-thaZbSon6e3Wi8-MXoHfOdppIAqha3yuqWdks2ZEuiowvAP0Dw5og)
> At its simplest, an AI agent is best described as AI-fueled software that does a series of jobs for you that a human customer service agent, HR person or IT help desk employee might have done in the past, although it could ultimately involve any task. You ask it to do things, and it does them for you, sometimes crossing multiple systems and going well beyond simply answering questions.
>
> Seems simple enough, right? Yet it is complicated by a lack of clarity.
Thoughtful article by TechCrunch.
### Mapping the landscape of gen-AI product user experience
[Mapping the landscape of gen-AI product user experiencePosted on Friday 19 Jul 2024\. 1,601 words, 29 links. By Matt Webb.Interconnected, a blog by Matt Webb](https://interconnected.org/home/2024/07/19/ai-landscape?ref=internalnote.com)
This article has quite the deep dive on the different ways we can leverages Generative AI in products. They end-up with a grid of experiences ranging from realtime to structured or contextual, and categories the experiences in 4 types:
> Users relate to the AI in different ways:
> \- Tools. Users control AI to generate something.
> \- Copilots. The AI works alongside the user in an app in multiple ways.
> \- Agents. The AI has some autonomy over how it approaches a task.
> \- Chat. The user talks to the AI as a peer in real-time.
Comparing this to Zendesk:
- Tools: Tone Shift, Expand and Summary seem to be three features in Zendesk we can categorize as tools.
- Copilots: Yep, [they're actively working on it](https://internalnote.com/preview-copilot/) even though the scope of Zendesk's Co-pilot is bigger than how the article defines it.
- Agents: Seems to be the future of the Agent Co-pilot. Soon. But not yet.
- Chat: Hello Zendesk Bot 👋
# 📝 Articles this month
[A hybrid approach to AI Agents powered by Zendesk and UltimateIn a previous article I explained the Road to Automation and how it can help with automating more your support interactions. This article takes a real scenario and keeps improving the customer experience by leveraging a hybrid approach combining flows, generative replies and API integrations.Internal NoteThomas Verschoren](https://internalnote.com/hybrid-approach/)
[Preview of the new Agent Co-pilot for ZendeskA few months after its announcement at Relate, I finally had to chance to explore the new AI Co-pilot for agents in Zendesk. This article will explore what the Co-pilot can do, how it works and some initial impressions.Internal NoteThomas Verschoren](https://internalnote.com/preview-copilot/)
[Managing custom translations in the Zendesk BotThe new custom translations for the Zendesk Bot make it possible to combine automatic translations for all your flows and answers, while making sure that specific replies or words do not lose their context in a bad translations.Internal NoteThomas Verschoren](https://internalnote.com/managing-custom-translations-in-the-zendesk-bot/)
[Custom Object Placeholders and Object Triggers in ZendeskThis article explores the new capabilities for Custom Objects with the new support for object triggers and dynamic placeholders.Internal NoteThomas Verschoren](https://internalnote.com/custom-object-placeholders/)
# And Finally...
I found [this](https://support.zendesk.com/hc/en-us/articles/7463958274586-Why-do-messaging-conversations-contain-old-and-previous-messages?ref=internalnote.com) interesting edge case in the Zendesk documentation:
> **How come my customer messaging conversation includes ten old messages?**|
> Bots consider a conversation expired after 72 hours of inactivity and deletes metadata for the message in the back-end, but not the message itself.
>
> If there isn't a point of pass-off to an agent after 72 hours, since there are still messages within the messaging infrastructure, when the customer reaches back out, Zendesk will look for essential, missing metadata in the back-end, such as `first_message_id`. Since there is no `first_message_id`, the system grabs the last ten messages from the previous interactions of the user and appends them into the ticket to attempt to provide some context.
### Custom Object Placeholders and Object Triggers in Zendesk
URL: https://internalnote.com/custom-object-placeholders/
Last updated: 2025-09-08T06:41:28.000Z
When I wrote my [Custom Objects](https://internalnote.com/tag/custom-objects/) series earlier this year, one of the main items I missed in building out flows was the ability to use the values of custom object fields when using macros or triggers.
For example, while you could use a macro that said "Thanks for contacting us about `{{product.name}}`", we couldn't use the same to e.g. reference the purchase date or product type of the selected record.
Similarly, if we have a user contacting us, and we need to contact that users' company main contact person, we couldn't create a side conversation for `{{ticket.organization.main_contact.email}}`.
Luckily, this has now changed with the release of placeholders for Custom Objects
[Announcing placeholders for lookup relationship fieldsAnnounced on Rollout starts Rollout ends July 9, 2024 July 9, 2024 July 10, 2024 Zendesk is pleased to announce powerful new placeholders on lookup relationship fields for tickets, ticket requ…Zendesk help](https://support.zendesk.com/hc/en-us/articles/7576570617882-Announcing-placeholders-for-lookup-relationship-fields?ref=internalnote.com)
Let's put this into practice.
# Overview
Imagine we're the support department for a small movie theater. We've got a series of movies playing and store those movies in Custom Objects so we can easily provide information to our customers.
Similarly, we also leverage custom objects to store mobile ticket purchases in Custom Objects so we can quickly reference a customers' bookings.
In our demo scenario we've got a customer contacting us requesting information about a movie playing in the theater. They decide they like the movie, and purchase tickets.
We'll use a combination of macros and [custom object triggers](https://support.zendesk.com/hc/en-us/articles/6294230624410?ref=internalnote.com) to handle this scenario.
## Using placeholders in Macros
### Custom Object
As mentioned above, we've got a list of movies stored in Custom Objects. Each movie has a title (name), director, summary, genre and a list of actors.
To make the demo easy to setup, I used the new [Data Importer](https://support.zendesk.com/hc/en-us/articles/6705584080794?ref=internalnote.com) in the Admin Center to import a set of movies.
This new importer allows you to import CSV files and create and/or update organizations and custom objects in bulk in your instance. For example, the csv table below can be imported into our custom object and it would create all movies.
| name | director | genre | external\_id |
| --------------- | ----------------- | ---------------- | ------------ |
| Inception | Christopher Nolan | science\_fiction | 1 |
| The Dark Knight | Christopher Nolan | action | 2 |
| Forrest Gump | Robert Zemeckis | drama | 3 |
| Titanic | James Cameron | romance | 4 |
If I'd later import a similar file, but this time add columns for `summary` and `actors` then the importer would update the existing movies, as long as the `external_ids` match.
[moviesmovies.csv5 KBdownload-circle](https://internalnote.com/content/files/2024/07/movies.csv "Download")
### Ticket Fields
Our ticket form for this scenario has a custom field to log the Movie the customer is interested in. Since our movies live in custom objects we're using a **lookup field**, linked to the Custom Object, instead of a text field to store this value. This way we can reference the chosen movie in our flows and macros.
### The Macro
Macros in Zendesk are an easy way for agents to reply with a predefined text. They've always supported placeholders to reference a requesters' name, custom field values or an assignee's name in the macro text, making the replies more personal and contextual.
If for example you want to use a requesters' name in a macro, you can use `{{ticket.requester.first_name}}`. Or if you want to print the Order Number stored in field 123456789 you can use `{{ticket.custom_fields.custom_fields_123456789}}`.
With the new placeholders for custom objects we can do a couple of fun things in our macros. Let's assume we have the following movie stored in our objects, and we have a lookup field on our ticket form with ID 123456789:
```
name,director,genre,external_id
"Inception","Christopher Nolan","science_fiction","1"
```
- `{{ticket.ticket_fields_123456789.name}}` will return *Inception*,
which is the **name** of the custom object.
- `{{ticket.ticket_fields_123456789.external_id}}` will return *1*,
which is the **external id** of the custom object.
- `{{ticket.ticket_fields_20080718149010.custom_fields.director}}` will return *Christopher Nolan* **,**
which is the value of the custom field **director** on the custom object.
- `{{ticket.ticket_fields_20080718149010.custom_fields.genre.title}}` will return *Science Fiction*,
which is the value (not the tag) of the dropdown field **genre** on the custom object.
You can find a full overview of all placeholders in this support article:
[Zendesk Support placeholders referenceWhat’s my plan? Zendesk Support placeholders are containers for dynamically generated ticket, user, and custom data. The format is a data reference contained within double curly brackets. Since yo…Zendesk help](https://support.zendesk.com/hc/en-us/articles/4408886858138-Zendesk-Support-placeholders-reference?ref=internalnote.com)
Our macro will have the following content*:*
> Hey *{{ticket.requester.first\_name}}*,
> It seems you're interested in knowing more about **{{ticket.ticket\_fields\_20080718149010.name}}**
>
> Here's some info:
> Director: {{ticket.ticket\_fields\_20080718149010.custom\_fields.director}}
> Genre: {{ticket.ticket\_fields\_20080718149010.custom\_fields.genre.title}}
>
> **Summary**
> {{ticket.ticket\_fields\_20080718149010.custom\_fields.summary}}
>
> **Some actors you might know**
> {{ticket.ticket\_fields\_20080718149010.custom\_fields.actors}}
## Putting it all together
With the setup complete, we can test out our flow.
1. We get an email from a customer interested in the Gladiator movie.
2. The agent goes to the ticket fields and selects Gladiator in the Movie lookup field.
3. They then select the macro that returns movie information.
4. The comment field gets filled with our macro text, with all the placeholders replaced with the information stored in our custom object.
## Conclusion
As you see above, we can leverage all the available fields in our custom object in our trigger and create rich replies with a lot of contextual data.
There's however still a few scenarios that aren't possible yet. Currently our placeholders all make a single *hop*. We can return a movies' `title` or `director` as shown in the macro above.
But imagine the director field is a lookup field itself that references a user in Zendesk. That user has an email, age, Oscars won and other custom fields.
What we currently can't do **yet** is returning a director's age or Oscars won in a macro. Something like `{{movie.custom_fields.director.custom_fields.oscars}}` is not possible for now.
We **can** return these so called *second hops* for ticket requesters and organizations though. So something like `{{ticket.requester.custom_fields.manager.email}}` would return the email of a customers' manager, stored in a lookup user field on that user.
# Object Triggers
Macros are one of the scenarios where we can leverage placeholders. A second feature in Zendesk that supports them is Object Triggers.
Object triggers are a new feature in Zendesk that allows you to automate processes based on changes in custom objects.
Similar to how a Ticket trigger gets fired when a ticket gets updated or created, object triggers run when the same action happens on a custom object.
## What's possible
Why would you use this? Well, custom objects are mainly used to store associated data on top of tickets, users or organizations. If you use Zendesk to handle support tickets, and you sell Service Contracts to customers, you can leverage custom objects to store the contract data (types like gold or silver, renewal date, associated services,..) and link it via a lookup field to an organization.
Similarly, in an asset management scenario you might link newly purchased assets to users or locations.
Object triggers can be used in these scenarios to automate processes on top of your records. When a service contract is renewed, or is about to expire, you can use an object trigger to send out a notification to the customer.
Or when you've received the new laptop or license for an employee, you can send out the serial code or confirmation to the employee when you log the asset in your custom object records.
### Conditions
Object triggers can fire when certain conditions are met:
- A custom object of a specific type is created or updated
- A custom object field has (or hasn't) got a certain value
(e.g. order status is updated to shipped)
- A lookup field attached to the object is linked to a certain other record
(e.g. manager is John)
[Understanding object triggersWhat’s my plan? In addition to using custom objects related to tickets in ticket triggers, you can also define triggers that run any time an object’s record is created or updated. These are called…Zendesk help](https://support.zendesk.com/hc/en-us/articles/6294230624410-Understanding-object-triggers?ref=internalnote.com)
### Actions
Once the conditions are met we can:
- Update a value of the custom object
(e.g. set notified to true)
- Notify someone via email, webhook, or text
(e.g. ticket requester, the assignee, or a linked user via lookup fields)
## Example flow
To show how object triggers work in practice, we'll continue our previous demo scenario. This time the customer is interested in buying tickets for their movie of choice, Gladiator.
### Custom Object
Just like with the movies earlier, we first create a Custom Object to store movie tickets sold. Each ticket has a name, date, time, number of tickets and other related data associated with it.
One of the fields to note is the *Buyer* field. This is a lookup field that links to users in our Zendesk instance and is used to link ticket to their respective buyers.
### Custom Object trigger
Now that we have our object, we can leverage the new custom object triggers. These live in a separate tab under Objects and Rules > Triggers in your Admin Center.
Our goal with our trigger is to send an email to the customer when new tickets are purchased by them.
To start, we setup a condition that looks at our *Movie Ticket* object, and used the *is created* condition.
Next, we'll add an action to notify via email. Since each record has an associated buyer, we can select them as a recipient. The system will automatically use their primary email address to send out the email.
💡
Emails send out via this way are send via **noreply@subdomain.zendesk.com* and not your regular support addresses!
My email body will contain the following payload:
> Hey,
>
> Thanks for buying tickets for {{custom\_objects.movie\_ticket.custom\_fields.movie}}.
>
> The show starts at {{custom\_objects.movie\_ticket.custom\_fields.time}} on {{custom\_objects.movie\_ticket.custom\_fields.date | date: '%Y-%m-%d'}}
>
> You bought {{custom\_objects.movie\_ticket.custom\_fields.number\_of\_seats}} seats for a total of {{custom\_objects.movie\_ticket.custom\_fields.total}}$
You'll notice the format of the placeholders here are a bit different than those used for macros. Since they run on objects rather than tickets, we need to use a `custom_objects.object_type` syntax. Since our movie ticket object has an identifier of `movie_ticket` (as set when we created the type), we can call all custom fields in the object via `{{custom_objects.movie_ticket.custom_fields.field_key}}` .
- `{{custom_objects.movie_ticket.custom_fields.movie}}` returns the associated movie stored in a lookup field.
- `{{custom_objects.movie_ticket.custom_fields.total}}` returns the value of the a custom field called total.
- `{{custom_objects.movie_ticket.name}}` returns the name of the record.
⏰
Since we're using a date field, we can use some formatting syntax to turn the date returned into a nice 2024-07-04 instead of an ugly long date via `{{custom_objects.movie_ticket.custom_fields.date | date: '%Y-%m-%d'}}`
### Putting it all together
So, back to the demo.
Our customer has read the summary and information about *Gladiator* and now decides he wants tickets for the movie.
The agent goes into custom objects and registers a new set of tickets by logging the user, movie (both lookup fields) and all associated data.
They let the customer know they'll book the tickets, and once they add the record our object trigger will fire and send out a, separate, email to the customer with the contents of our object trigger email configured earlier.
## Conclusion
The object trigger example used in this article might not be *that* realistic since most purchases will be handled by a proper sales platform. But for the sake of demonstration it serves its purpose on explaining how we can leverage object triggers to react to changes to the custom object records.
Since we can use the custom objects API to add records, you might imagine scenarios were we import or update records (like service contracts) based on external platforms, and Zendesk will then take care of the rest.
💡
One nice option that this also enables is the concept of a proactive ticket. By combining an internal webhook in Zendesk and the `{{custom_object.movie_ticket.buyer.email}}` placeholder, we can create a new ticket via API in the Zendesk instance. Useful for scenarios where we want to reach out for a contract renewal or other automated alert.
# Wrap up
Custom Objects launched last year and opened up a world of new possibilities in Zendesk. Over the last year updates like these placeholders, but also better API search, bulk import and update of elements and better permission levels made the platform that more powerful.
If you want to get insights in what's possible with Custom Objects on top of the examples in this article, take a look at my original series via the link below:
[Announcing the Custom Objects series for ZendeskIntroducing is a four-part series on Zendesk’s new Custom Objects feature. These articles cover setup, data import, using Custom Objects in forms and with agents, expanding user profiles, and displaying Custom Objects in Help Center forms.Internal NoteThomas Verschoren](https://internalnote.com/announcing-the-custom-objects-series/)
The only big missing piece for me is any end-user access to Custom Objects. You can make objects available [in the bot via API calls](https://internalnote.com/hybrid-approach/) but making them available in forms on the Help Center still requires [custom code](https://internalnote.com/custom-objects-part-4-end-user-and-forms/) for now. Although from what's been announced and shown at Zendesk Relate in April, end-user available lookup fields are coming this year.
## Sign up for Internal Note
A blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Managing custom translations in the Zendesk Bot
URL: https://internalnote.com/managing-custom-translations-in-the-zendesk-bot/
Last updated: 2024-11-03T10:51:14.000Z
Zendesk always has had a deep support for languages in its platform. The Help Center supports translated variants of articles, the Agent Workspace can be translated to fit the language of the Agent, and macros, intents and triggers can be customized to detect and use to reply with the right language to end-users.
Traditionally most of the translations in Zendesk were either done manually (in the case of Help Center articles) or via Dynamic Content placeholders for forms, triggers and macros. Both these options offer you the ability to translate the content yourself with tight control over tone of voice and words used, with the downside that everything needs to be translated manually before it can be used as a translated variant.
One outlier of this has been the Zendesk Bot. Instead of offering the ability of adding your own translations, the bot launched with the convenience of automatic translations. This meant you could write your flow and articles in one language, and the bot takes care of offering that content in other, enabled, languages.
The upside of this approach is easy to measure. Built once, and make other locales available with the click of a button. The downside however is that you lose control over your translations.
Some system translations might lose context and tone of voice, and your brand or product names might get lost in translation. I've heard stories of *Windows* (the brand) get translated into literal translations of *windows* (as in doors and windows) when used in a bot. Similar, a famous retailer in Europe saw its French name being translated to *crossroads*. Not ideal.
# Custom Translations
New in Zendesk Messaging is the ability to add manual translations to your bot flows. This allows you to overwrite the machine translation for any step or response of your bot with a fixed manual translation.
[Managing languages in a conversation botWhat’s my plan? This functionality is part of AI agents. Conversation bots include language settings that allow you to specify a default language and let you configure them t…Zendesk help](https://support.zendesk.com/hc/en-us/articles/4408842754202-Managing-languages-in-a-conversation-bot?ref=internalnote.com#topic%5Fhxt%5Fhzq%5Fbpb)
Take a look at the flow below. It contains an English welcome message, and it's automated translation in Dutch and French
However, when reading those messages they don't feel right for me. With the new manual translations feature we can however add a specific reply for some languages, while keeping the rest of the bot and locales auto-translated:
When we have our French and Dutch customers talk to our bot now, we get custom translations for our welcome message, while the rest of the buttons and actions are still machine translated.
This becomes even more powerful when we dive into custom answers. These flows often contain very specific steps, and interface buttons or product names might be different in specific countries or languages. For example, in the flow below the second screenshot shows an automatically translated response of our reset password flow. The reply contains a rather clumsy *inlog-ID*translation for the English *login ID*.
This *login ID* would be better translated with the same term as the English version as we did in the last screenshot. We can dive into the flow builder and add a manual translation to our flow for the response as such:
# Impact
This new option fixes a lot of issues customers' were seeing with the Zendesk Bot, and removes the need for workarounds around the lack of manual controls for translations.
I've seen customers add a brand per language with a unique bot per brand in order to add specific translations per country. Similarly I've seen customers that add a "What language do you speak" at the start of every flow, and then create a branch per locale.
All of these hacks can now be removed and replaced with the manual translations option for those languages.
One thing I really like about this flow that it's an option and you're not required to translate the entire bot, like you have to do in Ultimate's bot. Being able to trust the machine translations, and only dive into the settings for the few options where you want to add nuance or fix a translation is way less work and easily manageable.
There's one feature you currently loose though, and that's the Generate variants option for the basic bot behavior steps like the greeting and clarification steps. When you enable manual translations, the system will no longer generate a unique variant of your welcome message. This is understandable since the entire point of the manual translation is giving you control over what's send to the customer. But in the scenarios where you want to use the translations to fix a single word (brand name, button, product), it's a shame Zendesk AI can't just take the translated text as input to create variants in that locale.
# Dynamic Content
One question this new feature prompted for me was: will this replace Dynamic Content? Dynamic Content is one of the oldest pieces of Zendesk still in use and its UI harkens back to the old Zendesk UI from \~10 years ago when Buddha's were still a thing.
Dynamic Content was never added to Messaging, doesn't show up in the Zendesk Bot, is not available to use in newer features like Content Blocks in Guide and honestly, feels like an old feature now.
Seeing this new manage translation feature makes me wonder: why don't we have this for fields and macros? Manual translations and placeholders were nice in a time when AI and LLMs weren't a thing yet, but not that Zendesk showed us what's possible for the bot, I want this everywhere!
Take a look at the layout below: we've got the ticket field editor in Admin Center, but added the translation features from the Zendesk Bot. You'd be able to auto-translate dropdown options and labels, or overwrite a specific labels' value with a custom translation. Would be nice to have no?
That being said, I understand that Zendesk Bot and Messaging are a clean slate for Zendesk to start building on. They don't need to worry about compatibility with active installations, there's no migration paths or a dozen years' of existing setups to handle. Messaging and the Zendesk Bot started three years ago as very basic solutions and gradually worked towards feature parity with the existing Chat features Zendesk already had, while taking the time to reinvent and reimagine how things like escalation, translation or "being online" work.
Similar to how Apple used iPadOS to reinvent file handling, multitasking and other productivity features, Zendesk seems to use Messaging to rebuild their feature stack on modern technology. With the big difference that where Apple still needs to proof iPadOS is a real platform to do work on, Zendesk already did that with Zendesk Bot. 🤖
### Preview of the new Agent Copilot for Zendesk
URL: https://internalnote.com/preview-copilot/
Last updated: 2024-11-05T09:50:16.000Z
🕹️
****Update 2024-09-23** \- Zendesk has expanded access to their EAP and made it available for “everyone”. More info [here](https://support.zendesk.com/hc/en-us/articles/8000884897050-Announcing-wider-availability-of-the-early-access-program-for-agent-copilot?ref=internalnote.com).
At Relate 2024 Zendesk presented their latest Zendesk AI features, turning Zendesk into a *Complete Solution for the AI era.*
They focused on the customer, the agent, and the company as a whole within their releases.
*Customers* now interact with an AI Agent first. This bot offers self service with generated responses pulled from a help center, offers flows to route the customer to a solution, or used hybrid flows powered by integrations to automate processes for the Agent.
Human *agents* are still part of the flow. Even though AI can handle a lot of inquiries, agents are still there to help customer with the complex, the personal and the, by lack of a better name, human touch when handling tickets. Zendesk’s intelligent triage with its summary and intent features, and the contextual features like the knowledge panel, macro and suggestions or similar tickets all assist in finding the information quickly to solve a ticket.
And thirdly a *company* can get insights in their efforts and team with Zendesk’s new WEM solution combining both Quality assessment, workforce management and the traditional Explore reporting.
# Omnichannel Automation
Most, if not all, of the features announced at Relate are now available within Zendesk as released products or as early access programs to subscribe too.
One announcement of the event however still remains partially under wraps, and that is their new **Copilot** feature.
💡
**Spoiler: I got the chance to test it out. Scroll down if you want to read about it now, or read on for some context first.*
When Zendesk AI launched in 2023 it came with some additions to the comment field in Agent Workspace. Agents could **expand** an existing comment and turn a short reply into a longer, more nuanced and personal reply with context like the customers’ name added to the comment.
And **Tone-shift** takes an existing comment and turns it into a more friendly and casual, or professional reply.
Earlier this year Zendesk added two new features to assist agents with replying. **Suggested Replies** offer a prewritten reply which can be accepted by agents by pressing the *tab* key. It pulls data from the knowledge base, previous tickets and macros and learns over time.
The new **Quick answer** for the Knowledge base offers prewritten replies in the context panel pulled from your Help Center articles, which can be added to the ticket with the click of a button
The new Zendesk **Copilot** takes the existing reply features in Zendesk and turns them to 11.
## Sign up for Internal Note
The blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# Ticket automation
Traditionally there's two types of tickets we see in a ticketing tool.
One kind are those tickets that can be solved by a help center article. The article nicely explains the steps to take to resolve the tickets and the actionability lies on the side of the end-user requesting the ticket. The agents' role lies in making sure the customers sees the right Help Center article and guide them through the steps.
The other kind of tickets are those where the actions lie on the side of the agent or company. Executing a refund, updating an order, looking up a status or handling exceptions are all tickets that an agent needs to handle, and the customers is just awaiting a resolution or update.
Where the first kind of tickets was originally handled by the agent dutifully replying with content pulled from your Help Center, macros or internal knowledge base. But, the more powerful AI Agents became, the more of these kind of questions are resolved by chatbots and auto-replies via email (the old Answer Bot).
What’s left for agents are those tickets that are actionable and often require a process or a lot more information.
## An Example
Let’s take a refund flow for a product for example. Even though the chatbot can explain the customer when they’re eligible for a refund, or ask for information like order numbers or return reasons, it’s the agent that needs to approve and execute the actual refund in your webshop.
While AI Agents over chat can collect all the required information, the conversation still ends up with an agent who needs to act upon it.
Similarly, when a customers emails your team with “I want a refund now!”, your agents are required to reply to the email and ask for all that information. And we can only hope the customer replies with all the data in one reply. Which always happens. Right?
When we finally have that information, our agent takes that order number, product and reason and goes to the CRM executes that refund, updates the ticket, lets the customer know all is well, and finally solves the ticket.
## We can do better
Assuming we already leverage a good AI Agent, big parts of the above flow can be optimized already. But repeat the above scenario across dozens of tickets and we can quickly discover were agents loose time, or get numbed by the same actions all the time.
1. They need to **learn** all processes in your company
2. They need **identify** the **intent** and **process** (refund)
3. They need to **read** the conversation and ticket fields to see if we have all the required information (order number, product, reason,...)
4. They need to **validate** that everything is correct
5. They need to **execute** a refund
6. They need to **inform** the customer.
This is where the Zendesk AI Copilot comes in.
# Copilot
Copilot takes the above steps and automates them for the Agent. It is there to assist the agent throughout the entire process.
1. It is **based on** your processes
2. It can **identify** the intent and link it to the right procedure
3. It can **ask for the right information** to complete the task
4. It can **validate** that everything is correct
5. It can **execute** actions to be taken
6. It can propose a reply to the agent to **inform** the customer.
Opposite to for example the GitHub Copilot we see in Visual Studio Code, or [Intercom's Copilot](https://www.intercom.com/support-for-agents/ai-copilot?ref=internalnote.com) we don't need to ask the Copilot how to do something, but Zendesk's Copilot will actually take over from the Agent and just do the right thing. The agent, similar to a real pilot is there to make sure the Copilot does the right thing, and take over if things escalate or if a more personal touch is needed.
💡
Copilot is currently in EAP. The version available for testing only speaks English, has knowledge of **retail* intents and can only integrate with Shopify. Future updates should expand languages, intents and will offer ways to integrate with **any* API platform.
## How does it work.
Setting up Copilot starts with writing down a procedure to be followed for a certain Intent. Setting it up does to require any code or flow-building. If you can explain your process in logical steps, you can set up Copilot.
If Copilot detect an intent for a ticket, it will then follow the linked procedure. It will offer the agent replies they should send to the customer in order to resolve the ticket or if actions should be taken, it'll show buttons to the agent to do so.
🤫
I got approval from Zendesk to write and share about my experiences of the Copilot EAP. Since some pieces of the interface aren't "final" yet, I can't share screenshots of the entire setup process, but the steps described are copied from my testing environment as-is.
We can define our procedure as follows:
> **Exchange a product**
> To exchange a product we need to know:
> \- the order number
> \- the product name
> \- the return reason
>
> Once we know this, we need to check in Shopify if this is a valid combination. If so we can refund the product to the customer.
> We need to let the customer know this can take 2-3 days to be processed.
You’ll immediately notice that this is just regular text. I’ve written down the process in plain English without any fancy code or markup.
And this is one of the most awesome parts of Copilot from an admin perspective. You can define a process as a company, write it in a Google Doc and give that to your Zendesk Admin or CX Team Lead. They can copy the process, paste it Copilot as a procedure and that's it. The AI model underlying Copilot reads those procedures, and trains itself to take care of this process.
In this EAP I could only test retail flows but even there, the easy of use and wins in efficiency already showed. For example, when a customer mentioned an order number, not only does Copilot pickup that number, it also verified in Shopify if the user and the order number match:
## An email flow with Copilot
Let's make this real and go through an actual flow.
Let's say a customer emails customer care because they want to exchange or refund product.
When our agent opens the ticket in the Agent Workspace they’ll notice that the comment fields has been replaced with a new Copilot mode.
The Copilot read our customers' email, and offers a reply to collect the missing information. The Agent can either accept and submit the reply
If our customer then replies with a product and a reason, the Copilot once again offers a reply. This time confirming information, and politely asking the customer for the order number.
When we’ve finally collected all required metadata, we see a new button appear in Copilot. This time the system offers to cancel the items and the agent can execute the refund with the click of a button, and reply to the customer with some additional information on what happens now.
Copilot then offers to solve the ticket.
## A chatbot flow with Agent Copilot
Let’s take this same flow, but now our customer contacts us over Messaging and an AI Agent intercepts the conversation.
💡
The AI Agent in this flow was built on [Ultimate](https://internalnote.com/hybrid-approach/) for its easy of integrating with Shopify. You can however accomplish the same with the regular Zendesk Bot.
The first steps of the flow are handled by our bot. It shows a list of available orders, and then shows the items within an order to our customer.
At the end of the flow we pass the conversation to our agents. When the ticket arrives in Agent Workspace we already get a lot of context. The bot collected the order number and product already, so this time Copilot offers a reply asking for the return reason only.
And similarly to the example earlier, Copilot wraps up by offering a refund button, and we can close our ticket.
For the customer, this interaction feels like a regular interaction with a real agent:
## Efficiency Gains
The efficiency gains enabled by Copilot are immediately visible. For every procedure defined for Copilot, our agent only needs to verify the proposed actions are correct, and they can submit or approve the next steps for the customer.
Only in scenarios where the solution is wrong, or where Copilot doesn’t know how to handle the ticket, will agents actually need to jump in and use the existing AI tools and their own skills to resolve the ticket.
Copilot doesn’t require **every** procedure to be written down for it to work. Similar to [generative replies](https://internalnote.com/quick-look-at-the-new-generative-ai-bot-for-zendesk/) for the Zendesk Bot and its custom answers, if no procedure is defined, Copilot will fall back to replies generated from macros and your knowledge base, similar to the suggested reply feature available today.
I was also surprised by how nicely it could context switch. In one of my tests I asked the Copilot for washing instructions for a jersey and I had imported Help Center articles for denim, cotton, polyester. The Copilot asked me which fabric I had and I replied with “denim”.
While I saw the reply being generated as an agent (which takes less than a second), I quickly had my demo user reply again with “cotton”.
Copilot immediately started generating a new response and gave instructions for both fabrics, prefaced with a short “in case you need both..”.
# Conclusion
For now, Copilot is in private beta and isn’t available yet for everyone to try. It only works in English, and its intents are limited to retail environments. And actions can only be executed in Shopify for now.
But ignoring these limitations for now, if we look at the feature for what it can be it’s clear that this will be a game changer for both agents and Zendesk admins.
Once this becomes broadly available agents this will unlock some great benefits from the get go. The obvious one is the fact that agent will need to write less and that team leads can be sure processes are followed.
This means agent can focus even more on the exceptions and really dive deep into the complex tickets that require their real attention.
Similar, when onboarding new people to the team it’ll become way easier to get them up to speed. Now it requires learning complex processes and taking in all the data needed and steps taken to resolve a ticket. With Copilot the system will guide the new agent through their first *refund ticket* and while reading along with the responses the new agent can learn those processes on the go.
There’s a few things I wonder about though. If you need to take care of hundreds of Refund ticket a day, accepting those replies will become a shore after a while. Luckily Zendesk already announced at Relate that the end goal of Copilot will be a way for Admins to set a threshold. And for all replies where that threshold is met, Copilot will autonomously reply, leaving only the edge cases for Agents to approve.
The relationship between Copilot and AI Agents is also tricky. Where AI Agents stop, Copilots begin. If your process requires human approval for refunds, the AI Agent can only do as much and identify the user, tag the intent and collect metadata. It’s the Copilot that then takes over, with final approval by the Agent.
I kinda hope we can someday define our procedure in Copilot and have an AI Agent feed from that as well. If our process requires an Order number and reason for a refund, there should be no reason we need to manually built that Answer flow in the bot, when Copilot can generate that process automatically.
Putting aside future wishes, I was really surprised by how solid the Copilot worked. The example flows I showed you above were built in a matter of minutes without resorting to documentation and worked immediately.
Really curious how this will evolve in the next few months and years!
## Like this kind of content?
I write about Zendesk every week. Subscribe today and get new Zendesk updates in your inbox. For free.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### A hybrid approach to AI Agents powered by Zendesk and Ultimate
URL: https://internalnote.com/hybrid-approach/
Last updated: 2024-11-05T10:02:13.000Z
In a previous article I explained the [Road to Automation](https://internalnote.com/road-to-automation/) and how it can help with automating more and more of your support interactions.
In this article I’m going to take a real scenario you might want to support, and keep improving the customer experience by leveraging a hybrid approach where I combine traditional flows, generative replies and API integrations.
Let’s imagine the following scenario: we’re a webshop selling beers to customers. We offer a wide range of beers but we notice our customers reach out for the same problem each and every day: how I pour the perfect pint?
# Using the Zendesk Bot
## The basic: showing a Help Center article
This is where most companies who want to use a chatbot start. They have an existing set of Help Center articles, and when we move to a chatbot, we can have the bot show those articles to the customer.
In our scenario, when a customers asks *How to pour a beer?*, we respond in Zendesk with a list of articles explaining the steps for a variety of beers.
It works, but has the downside that it requires some additional work from the customer. They need to look through articles to find the right beer, read the entire thing, and shift from the web widget (or social channel) to a website.
## Improvement #1: Generative Replies
Let’s do the same thing, but this time we enable generative responses in Zendesk.
The customer ones again asks: *How to pour a beer?*
Since we’ve enabled Zendesk’s generative replies, our bot now reads the article and responds with a reply for the customer.
However, we run into an issue here. Unless the customer specifies their exact beer, our bot will take a random article to reply, and possible gives instructions on how to pour a Guinness for the person who wants to enjoy their Duvel. (And trust me, the steps here are widely different!)
Within Zendesk, we can’t solve this issue elegantly within the generative reply options. We can only offer a potentially wrong answer, and then have the customer ask the same question again, this time specifying their beer.
## Improvement #2: Use a custom answer
One way we can solve this in Zendesk is by leveraging answer flows. We can create an answer that responds to phrases like “how to pour a beer”.
This answer can then show a list of available beers, and when the user selects their beer, we show the right help center article for the selected drink.
The only downside here is that we loose the nice generative replies, and have to fall back to a list of articles.
# Using Ultimate
This is where Ultimate comes in. Ultimate’s generative bot capabilities not only work to reply to direct questions, but can also be used within flows to dynamically create responses.
## The basis: using a generative reply
uGPT, as their technology is called, can be used across the entire bot to generative custom replies to your customers just like Zendesk's Bot. Similar to Zendesk Ultimate indexes your Help Center (or other sources) and generates a reply for questions asked by the customer.
So all things equal, we also have the same issue as we had in Zendesk. Unless the customer specifies their preferred beer, chances are they get the wrong instructions.
## Improvement #3: Use generative replies in bot flows
In Ultimate we can fix this issue again by creating a custom answer, or reply as it's called in Ultimate, to ask the customer first to specify their choice.
Here you see a a similar flow as we’ve build in Zendesk. We once again ask the customer to select their beer choice and then go to the next step.
The main difference between Ultimate and Zendesk is that instead of showing a list of articles, we ask uGPT to search for pouring instructions for the specific beer.
The bot will search for an article that matches their choice, and generate a custom reply with content from the matching article.
And while where pointing out differences and benefits between the two platforms, you’ll notice that after we offer the customer the available beers, we can link all those choices back to one flow to generate the response. There’s no need to duplicate the next steps as we have to do in Zendesk. This makes the flows cleaner and less complicated.
As you can see in the images above, our bot nicely responds with custom instructions for each beer.
However, once the customer confirms everything is well, our bot always uses the same predefined response to wrap up the conversation. Which, if our customer repeatedly needs help, will quickly feel repetitive for them.
## Improvement #4: Even more personal
So, even though our bot already handles the "How do I pour a beer" question completely, we can make the entire flow a bit nicer by also closing the conversation with a uGPT generated response:
Now, let’s talk to the bot a few times.
You’ll notice that the responses take the context of the conversation into account, and that our bot replies with very personal and unique confirmation messages.
## Improvement #5: Leveraging APIs
There’s one final optimization I’d like to make to our bot. Currently our flow only accounts for two beers: Guinness and Duvel. But our webshop has dozens of beers available, and we’d like to offer the customer a list of all the available beers to choose from in our flow.
We could manually add them all, but not only is this a lot of work, it’s also not ideal since new beers are added to the webshop every day.
A better way to solve this issue is by adding an API integration to our chatbot. This will pull in all beers via API, and render them as a nice carousel.
I’ve written about this in the past on how to do this in Zendesk, but since we’re using Ultimate this time, let's see how we can build a similar flow in this new Zendesk tool!
[Expanded support for variables in the Zendesk BotZendesk recently added variables, dynamic options and carousels to their Bot Builder. No better way to showcase these new capabilities than building a bot powered by the Avengers!Internal NoteThomas Verschoren](https://internalnote.com/expanded-bot-variables/)
For easy of use I’ve added all the available beers to [custom objects](https://internalnote.com/tag/custom-objects/) in Zendesk. Each beer has a name, a percentage, brewery and image we’re going to use in our bot flow.
First step is adding an API integration to Ultimate. The nice thing here is that all integrations live outside of bot flows, which means we only need to setup an integration once and we re-use it across multiple flows.
Next, we’re adding the API to our existing flow and turn our manually created carousel of available beers into a dynamic carousel to show the options to the end-user as made available via the API.
And now, when a customers asks *how to pour a beer*, they get the follow flow:
# **Wrap Up**
So, hopefully this article gave you some inspiration on how to improve your flows and experience in your own Zendesk bot across both Zendesk and Ultimate.
One key take away from this article is that by levering hybrid flows you get much more powerful and effective flows which will result in a higher ticket automation rate, by combining help center content, APIs, custom flows and generative responses.
The article above also demonstrates a few ways in with Ultimate is more powerful and efficient than Zendesk’s own bot in handling such flows. So if your AI Agent reaches the limits of Zendesk’s build in features, it might be worthwhile exploring their newest acquisition!
### Zendesk Roundup for July 2024
URL: https://internalnote.com/roundup-2024-07/
Last updated: 2024-11-03T10:54:11.000Z
The past month was a month full of Zendesk Showcases around the world. Showcases are mini-Relate events where Zendesk presents their newest updates to customers in key cities across the globe.
I had the chance to visit the [Paris Showcase](https://youtu.be/3AhezmN2njo?si=kv-PmYDTgm-k-FuW&ref=internalnote.com) as part of [Premium Plus](https://premiumplus.io/?utm%5Fsource=internalnote) meeting up with customers and Zendesk friends for a day of product updates, demos and conversations.
# 🎉 New Releases
## 🤖AI Agents
### Expanded conditional operators
Zendesk's bot builder has a conditional step where you can create branches in your flows based on users input. You can for example show a different flow if a user picks the blue or the red pill.
It used to be a fairly limited set of conditions where we could only filter based on *is* or *contains* operators, but now the options are expanded with a lot more variables.
## 👨🏻💻 Agent Workspace
### Agent Home
Almost a year after the initial release of [Agent Home](https://internalnote.com/agent-home-beta/), Zendesk has now [announced](https://support.zendesk.com/hc/en-us/articles/7436748335642-Announcing-rollout-of-Agent-Home-for-Agent-Workspace-customers-default-experience?ref=internalnote.com) that this new dashboard will become the default experience for all Zendesk customers. This means that starting this summer Agents who launch Zendesk will be greeted with the new overview and get a direct view on active tickets assigned to them, the tickets they follow or are cc'd on.
Parallel to this roll-out the new Agent Home also got a bunch of new features, giving your agents even more context.
- Ticket statistics are now clickable, and route you to a predefined search page where you can analyze your customer feedback
- New Satisfaction Statistics are now visible
- Tickets Assigned to My Groups are counted on Agent Home as they were on the legacy Dashboard, and are hyperlinked to search results that will let you see the tickets relevant to that count (up to 1,000 tickets)
- And to wrap it all up, Agent Home now also supports Talk tickets in its views.
Personally, ever since Agent Home arrived, I've rarely used any of my Views anymore. Looking at new tickets for me happens right from the Dashboard, and I only look at Views if I want to check my backlog or reference an older ticket.
Can't wait to get a similar dashboard in the mobile app.
[How to get the most out of Agent Home for ZendeskThis article shows you ten practical tips on how to gain the most out of Zendesk’s new Agent Home.Internal NoteThomas Verschoren](https://internalnote.com/agent-home-tips/)
### Retain original groups and assignees on Follow-up tickets
You can now choose wither reopened (follow-up) tickets get assigned back to their original assignee, or if they are created without assignee and are reassigned by [queues](https://internalnote.com/beyond-triggers-moving-to-queue-based-assignment-in-zendesk/) or triggers.
### Consolidated Macro Management
After Relate, I wrote that this years' Zendesk felt more like AI integrated into the Suite, whereas last years' release felt like Zendesk with some AI sprinkled on top.
It's nice to see that this integration is now also visible in the Admin Panel. Instead of treating AI features as separate elements in the UI, they're now moving into their logical sections. Looking to do something with Macros? Easy. All macro features, both editing and suggestions, are now in one dashboard.
*Can't wait to see more of these integrations. Messaging authentication settings being outside of the rest of authentication being one of them.*
### Group level access for Agent Status
As part of Omnichannel Routing you can leverage (custom) Agent Statuses to make agents available for all or only a select set of channels. By default you've got online, offline, transfer only and away statuses, but you can add custom statuses to, for example, only allow email or messaging for that user.
In bigger organizations that list grows quickly, and not every team might need a "Talk Only" status. With this months' update you can now assign a Custom Status to only be available to specific groups, similar to how we can assign macros and views.
💡
Note: Custom statuses created prior to June 10, 2024 are available to all groups unless an admin [edits the status](https://support.zendesk.com/hc/en-us/articles/5133588225690?ref=internalnote.com) to restrict access by group. After this date, admins must specify group accessibility for each status they create.
## 🔎Help Center
### Multiple user segments in view permissions
One of the main limitations in [segmented articles](https://internalnote.com/zendesk-guide-membersonly/) for the Help Center was that you could only assign one segment per article. This lead to scenarios where if you want to make some articles available to Organization A and Organization B, and others to only A or B, you'd end up with three segments:
1. Users belonging to organization A
2. Users belonging to organization B
3. Users belonging to organizations A or B
Doing the same for three organizations required up to six segments and keeping things organized and scalable is often impossible for bigger setups.
Starting this month this issue is resolved with the ability to add up to ten segments per article. Instead of creating a segment for users in organization A or B, we can [now add the segment](https://support.zendesk.com/hc/en-us/articles/4408824005914?ref=internalnote.com) for organization A and the one for organization B to the article. Perfect 🤩.
## 🧱 Open and Flexible Platform
### Sandbox replication
Premium Sandboxes are limited to [a subset](https://support.zendesk.com/hc/en-us/articles/4408822049818-Creating-a-premium-sandbox-with-data-replication?ref=internalnote.com#topic%5Fnxm%5Frkf%5F5yb) of Zendesk users, but for those who have access, the replication of your instance from production to sandbox can now also copy over your Marketplace apps, as long as they are free, without logins and don't require configuration during setup.
As a side-note, it's nice to see the sandbox in Zendesk slowly getting more and more capable. A full mirror of your instance with all settings (Bot Answers come to mind) is still not possible, but piece by piece we're moving towards that future.
### Data Importer enhancements
Zendesk's new Data importer can [now](https://support.zendesk.com/hc/en-us/articles/7471865612826-Announcing-enhancements-to-the-data-importer?ref=internalnote.com) do a *create and/or update* when importing new organizations or custom objects. This used to be limited to either only creating new objects, or updating existing ones, but with this new feature we can ingest an export and the system will take care of adding what's missing, and updating the existing objects.
This feature used to be available in the old organization importer, and is also available over API, so it's nice to see this being available in the new importer too.
*Up next, importing users? 🙏*
# ⚠ Major Changes
### Removing support for password API authentication
Zendesk is *finally* removing access to APIs over passwords. I say finally cause it's one of the less safe methods to authenticate. Passwords can be reused by users, and if another tool or platform where a user uses the same password is breached, anyone testing that email and password combination against your Zendesk has access to its API.
Ironically, this also means Zendesk is removing a feature that made password authentication **safer** than token authentication. With password API access the access limits are set by the user that's logged in, combined with their password. If a user logs in, they can access what they can for their role.But if you give that same user an API token, and they swap out their email with that of an admin, all of the sudden they get admin access. So here's to me hoping this is the first step towards making API tokens with granular access rules.
> From **31 July 2024**, we’ll no longer provide the option to use email and password for API call authentication for accounts that didn’t use this feature. Our focus will switch entirely to API token and OAuth as the supported authentication methods.
>
> **Next Steps:**
> If you’re not actually using this feature, you don’t need to do anything. We’ll deactivate this feature and after 30 July, the feature will be permanently removed from your account.
>
> However, if you’re still using this feature, please reactivate it before 30 July to maintain access until 31 December 2025\. You can find this under Admin Centre: Apps and Integrations > APIs > Zendesk APIs > Settings > Password Access.
### EOL of the legacy Social Messaging app
On July 31st Zendesk will kill the Social Messaging integration. This app was a precursor to Messaging and allowed you to link WhatsApp, Facebook Messenger and other social apps before they became available as part of the Messaging features.
Both are based on Sunshine Conversations, but where Messaging is deeply integrated with bots, reporting and agent status, Social Messaging was always a bit of a quick hack to get Sunshine Conversations enabled in Zendesk and was always a temporary features awaiting the, back then, future release of Messaging proper.
This summer Social Messaging will finally be removed, so if you Zendesk still shows the icon above, [you'd better migrate quick](https://support.zendesk.com/hc/en-us/articles/7413502356506-Announcing-the-in-ticket-notification-for-the-EOL-of-the-legacy-Social-Messaging-app?ref=internalnote.com).
To make it even more clear you're part of the affected environments, Zendesk will show an in-app alert for these customers on every ticket.
# 💡Insights
### Escalation criteria
Nice article on how you can define a process for escalating tickets and customer requests by the team from Birdie.
[Building an Effective Escalation Criteria Table \[+ template\]Maxime Manseau](https://www.birdie.so/toolbox/escalation-criteria-table?ref=internalnote.com)
### Deep-dive into Custom Objects
Nice overview of [custom objects](https://internalnote.com/tag/custom-objects/) with a preview of the new end-user editable lookup fields in forms.
### Recommending articles to users who are authenticated with JWT
Zendesk's Messaging [authentication](https://support.zendesk.com/hc/en-us/articles/7431328780314-My-bot-is-not-recommending-articles-to-users-who-are-authenticated-with-JWT?ref=internalnote.com) sure is the gift that keeps on giving. It surely has to be one of the more complex setups in the entire suite when it comes to grasping all the edge cases and exceptions.
Take [this article](https://support.zendesk.com/hc/en-us/articles/7431328780314?ref=internalnote.com) for example:
> End user who are authenticated with JWT can only access articles when their user profile in Zendesk is created before their conversation starts with the bot.
> For this to happen, ensure the end user exists in your Zendesk database before the user interacts with the Web Widget. This creates the external ID used to check the user profile and their access permissions. Users profiles without an external ID cannot see article recommendations, even if they are authenticated with JWT.
So if I understand this correctly:
- Unauthenticated users get recommendations by default
- Authentication users only get recommendations **if** they exist in Zendesk and have an `external_id` associated with them.
Which basically means that if you ad-hoc provision users the first time they login to your widget, those users will probably get no articles recommended. Importing your users in advance into Zendesk seems to be the best way forward here.
# 📝 Articles this month
[Quick Look at the Zendesk Relay App for outbound WhatsApp messagesThe new Relay app from Zendesk makes sending outbound messages over WhatsApp via Sunshine Conversation available to all Zendesk users without custom coding or development.Internal NoteThomas Verschoren](https://internalnote.com/relay-app/)
[Showing device Information in the customer panel in Zendesk Agent WorkspaceThe new Device information section in Agent Workspace gives you information about your customers’ context right next to a Messaging ticket.Internal NoteThomas Verschoren](https://internalnote.com/device-information/)
[Handling multiple support addresses in a support emailThis article offers a solution for scenarios where customers email multiple support addresses in your Zendesk instance and you only see a single ticket being created.Internal NoteThomas Verschoren](https://internalnote.com/multiple-recipients/)
[Zendesk Roundup for June 2024Subscribe to a hand-picked round-up of the best Zendesk links every month. Curated by Thomas Verschoren and published every month. Free.Internal NoteThomas Verschoren](https://internalnote.com/roundup-2024-06/)
# And Finally...
I came across a list of Chrome extensions compatible with Zendesk
[Quicktab for Zendesk by TymeshiftWell behaved browser tabs for Zendesk agents](https://chromewebstore.google.com/detail/quicktab-for-zendesk-by-t/hhbimbckgheipimadcknkfogegmpoibj?ref=internalnote.com)
When you click on a ticket link, it opens the ticket in an existing Zendesk tab without opening a new one.
[ZendeskTicketCopyCreates a simple button in Zendesk to copy the current ticket number.](https://chromewebstore.google.com/detail/zendeskticketcopy/ngkngoghlonilnnejjcecljbbaekapim?ref=internalnote.com)
Adds a 1-click copy button next to the ticket number.
[Zendesk Dark Mode ThemeTake care of your eyes day and night using dark theme for Zendesk tools and services.](https://chromewebstore.google.com/detail/zendesk-dark-mode-theme/ndneacmpkcghablknbeiiopilgeomhon?ref=internalnote.com)
Zendesk. But after hours.
## Sign up for Internal Note
The blog about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Quick Look at the Zendesk Relay App for outbound WhatsApp messages
URL: https://internalnote.com/relay-app/
Last updated: 2024-11-03T10:54:20.000Z
At [Relate](https://internalnote.com/zendesk-relate-event-recap/) this year Zendesk has a Zendesk Labs booth with previews of new ideas they were working on. The Zendesk Labs initiative builds and tests out concepts before they're ready to go into the main Zendesk Suite.
Examples are the [Conversational Commerce](https://www.zendesk.com/newsroom/press-releases/zendesk-conversational-commerce/?ref=internalnote.com) concept, or now the new [**Relay App**](https://www.zendesk.com/marketplace/apps/support/1040322/relay/?ref=internalnote.com)for Zendesk.
This new Marketplace app from Zendesk allows you to send out messages over WhatsApp and SMS to customers making use of templates and the Sunshine Conversations API.
It's a free app that takes care of both creating templates and sending those messages in bulk to your customers.
[](https://www.zendesk.com/marketplace/apps/support/1040322/relay/?ref=internalnote.com)
## What's possible
Similar as to how you can leverage the Sunshine Conversations API to [by-pass the 24 hour limitation](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/) for customer communication, you can also use this API to send outbound messages to your customers.
You can use this to send out marketing messages, order confirmations, delivery updates or any other type of outbound message to WhatsApp (or SMS) users.
I explained a similar flow in an earlier article, but now the same is possible with a native and much more powerful app by Zendesk itself.
[➕ Sending Automated Messages via WhatsApp with Sunshine Conversations and ZendeskLearn how to leverage the Sunshine Conversations API included in Zendesk Suite to automatically send out messages to your customers.Internal NoteThomas Verschoren](https://internalnote.com/sunshine-conversation-automations/)
# Installation
The process for installing the app is fairly well documented in the support article linked below. It does require admin access to your Zendesk instance and a working WhatsApp for Business account linked to both Zendesk and your Meta Business Manager.
[Installing and using the Relay messaging appThis article includes instructions on setting up and using the Relay messaging app. Important: You are responsible for using the WhatsApp and text features in Relay in compliance with all applicab…Zendesk help](https://support.zendesk.com/hc/en-us/articles/7051551958426-Installing-and-using-the-Relay-messaging-app?ref=internalnote.com)
## Enable custom objects
This app stores a lot of its configuration data in Custom Objects. So if you haven't yet, go to the Admin Center and enable Custom Objects first!
And if you're interested in Custom Objects, take a look at my Custom Objects introductions series!
[Zendesk Custom Objects - Part 1: IntroductionThis is a four-part series on Zendesk’s new Custom Objects feature. The articles cover setup, data import, using Custom Objects in forms and with agents, expanding user profiles, and displaying Custom Objects in Help Center forms.Internal NoteThomas Verschoren](https://internalnote.com/custom-objects-part-1-introduction/)
## Creating an Authorization Token
First step is creating the authorization key by adding a new Conversations API token.
Since the apps' installation settings require a base64 encoded combination of the Key ID and Secret key it's easiest to go to this [Sunshine Conversation token generator](https://zendesklabs.zendesk.com/hc/en-us/p/sunco-token-generator?ref=internalnote.com) to create the encoded key/secret token.
## WhatsApp Namespace ID
The documentation mentions that the WhatsApp Namespace ID is optional, but I noticed it doesn't seem to work as reliable without it, so best to go to the [Meta Business Messenger](https://business.facebook.com/settings/?ref=internalnote.com) and grab the ID from *WhatsApp Manager | Account Tools | Message Templates*. Once on that page, you can click the ⚙ Namespace button to copy the ID.
## Adding the app
Once you've collected the required tokens and IDs you're all good to setup the app.
There's a few caveats I noticed:
- The Zendesk subdomain should be your main domain, and not a subdomain of one of your brands. The API calls failed for me when I used a different domain, even though the WhatsApp channel I used was linked to another brand.
- While testing, best to set the Blackout value, which defines how often you can send the same message to a customer, to zero. This way you can test with the same customer until you're happy with the result!
Once you've setup the app you can jump to your Agent Workspace and open the Relay app. It's located in the navbar on the left.
At first launch the app will take a few moments to load your Meta/WhatsApp data and configure itself. This process took \~two minutes for me, so don't worry if it takes a while.
After wrapping up the initial configuration, you'll be greeted with a list of your existing templates (if any) and the option to create new ones or send a message.
# Creating templates
WhatsApp has strict rules on who can send messages to customers, and what those messages can contain. Before we can send out a proactive messages we first need to create a template which has to be approved by Meta.
This approval process takes only a few minutes so it's not that big of a hassle.
So, to create a new Template, click the *Create button* on the apps main screen. Give your new template a name, and select the channel it should apply to.
Next you'll be asked to select a category. Make sure to select the right one, since this will impact your template approval!
- Marketing: used to promote your product or service to customers. Useful for sending out promotions, discount codes e.a.
- Utility: used for transactional emails like order confirmations, shipment updates e.a
- Authentication: used for sending out password reset codes, 2FA codes, e.a
- Service: used in customer care scenario's to send out CSAT, [bypass the 24 hour window](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/), e.a
💡
Note that sending outbound messages over WhatsApp is ****not** free. Depending on the category you select, prices will vary. See [this article](https://business.whatsapp.com/products/platform-pricing?ref=internalnote.com) for more info.
When filling in your message template you can make use of *dynamic values*. E.g. if you want to send the customer the following:
> Your order #1138 has been shipped. Track its status via the link below
We can add `{{order_number}}` as a placeholders in the text.
```
Your order #{{order_number}} has been shipped. Track its status via the link below.
```
When we submit our message the app will offer text fields where we can add *1138* and as a value and dynamically insert it into our message.
Similarly we can add a Call to Action button that adds a tracking link, or we can add Quick Replies for the customer to easily confirm or escalate a message. Since messages are send out over the same WhatsApp channel we use to receive requests in Zendesk, any response from the customer will be picked up by the Zendesk Bot or an Agent.
Note that your URLs can also contain those placeholders, so `https://track.me/{{order_number}}` is a valid URL to enter in this field.
After filling in the template we can submit it for review with Meta. As mentioned above this approval flow often takes minutes and you can clearly follow along with its status from within the Relay app.
# Sending messages
## Selecting a template
Once you've got your templates configured you're ready to send out your messages. The first step is selecting a template and choosing a channel from the list.
If your template contains placeholders, Relay will present you with a form to input your order number, delivery date or other elements. The preview on the right nicely previews the result.
💡
One immediate limitation I noticed is the fact that even though we can select multiple users, the placeholders are the same for every message we're sending. So while placeholders are ideal for bulk updates with the same content to all customers, sending unique and personal messages per customer is not possible.
## Selecting users
After configuring the contents of our message, it's time to select the recipients. Weirdly, instead of offering a list of users or an option to import users, we're shown a text field that accepts a search query.
If you've entered a valid entry – no error messages or useful information here – the *Next* button will become active, and the *Audience* title will show the amount of recipients between (brackets).
Clicking the *Manage audience* link will show a modal view that lists all recipients, and allow you to exclude specific users from the list.
## Useful search entries
If you've never used the Zendesk search syntax, well, it's not that difficult to grasp. The article below gives a full overview of all the options, but I've listed the most relevant for Relay below.
A useful query is the syntax to list a whole list of phone numbers in one search query. It's similar to importing a list of users, with the only caveat that these users have to exist in Zendesk as end-users already.
[Searching users, groups, and organizationsWhat’s my plan? Data in the end users, team members, group, and organization objects can be searched on their respective pages. This article covers the following topics: About searching The us…Zendesk help](https://support.zendesk.com/hc/en-us/articles/4408883318554-Searching-users-groups-and-organizations?ref=internalnote.com)
- `name:"John McClane"` to send a message to a user with that name
- `tags:vip` or `tags:"vip jedi"` to select users with a specific tag
- `email:james@universalexports.co.uk` to select a user with that email. To select multiple repeat the query. `email:mr.blue@reservoir.dogs email:mr.green@reservoir.dogs email:mr.black@reservoir.dogs` . The same trick works for selecting multiple phone numbers by using `phone:+1234567890` and repeating that query.
- `created:2024-05-21` if you want to send all new users
- `plan_type:platinum` for all users where the custom field *Plan Type* is platinum
And finally, after selecting the template and our audience, we're ready to send our message to all users:
# Limitations
The Relay app is developed via Zendesk Labs, and you notice this when using the app. It does what it promised to do, but it feels like a proof of concept and not as a fully developed and robust solution.
For some reason I can't send a message to my own personal WhatsApp number. But with only this helpful error message and no logs or console errors, it's just impossible to troubleshoot why.
Similarly, the process of manually encoding and adding the tokens, instead of leveraging a nice oAuth flow also feels a bit clumsy.
But, since this is a Labs app on the Marketplace and not a core part of Zendesk Suite like [Proactive Messages](https://internalnote.com/proactive-ticketing-for-messaging/) is, it's understandable. Although I do hope some of the setup and error experience gets some attention in future updates.
Speaking of future updates, there's a few features I'd love to see:
- Add scheduling. It would be nice to be able to schedule messages in advance.
- Add the ability to importing users in the audience step, and add these as end-users in Zendesk.
- Make these messages available as an action in native automations or triggers.
- Allow for setting placeholders in templates based on user fields, organization fields or linked custom objects.
# Conclusion
So yeah, that's Relay. A nice and **free** app on the Marketplace that allows for some simple outbound reach over WhatsApp (or SMS). It makes using the included Sunshine Conversation abilities available to all Zendesk admins, even those who can't or don't want to build their own integrations to handle this.
### Showing device Information in the customer panel in Zendesk Agent Workspace
URL: https://internalnote.com/device-information/
Last updated: 2025-09-08T06:41:12.000Z
In [May's Roundup](roundup-2024-05) I wrote about Zendesk's new Device Information section for the Customer Context panel. This new sidebar section, displayed next to Messaging tickets, shows information related to the customers' device to agents.
This information, like operation system, browser type and version is useful information when troubleshooting technical issues.
For example, if you know a certain feature of your app doesn't work on older Android phones, or that your web app requires the use of Chrome instead of Safari, these pieces of information are crucial to detect problems and alert the customer.
# How does it work
The device information section is enabled by default for all instances. By default is will show operating system and browser information, and hides IP address and location info from agents.
By toggling the options in *Admin Center > Workspaces > Context Panel* you can display either or both the IP Address and location to agent if that information is relevant for your agents.
Since you're actively collecting and displaying PII to agents, don't forget to update your privacy policy accordingly! If you want to handle cookies correctly, take a look [here](https://developer.zendesk.com/api-reference/widget-messaging/web/core/?ref=internalnote.com#set-cookies).
[babelforce - The #1 Most Flexible Call Center Softwarebabelforce is the #1 most flexible call center software for customer experience teams. Make customer service easy. Make growth easy.babelforce – The No-Code Contact Center Platform](https://babelforce.com/?utm%5Fsource=internalnote)
My thanks for babelforce for sponsoring this month's Internal Note newsletter.
## Relevance
The device information context is linked to the **end-user** and not to the actual ticket. This means that this data will update regularly at each moment the user interacts with your Zendesk instance.
The benefit of having a single datapoint that updates for the user is that you always have the latest information available. If a user has updated their system and replies to the conversation, you can see that new data right next to the ticket.
However be mindful that the data shown can be more recent than the question the customer has asked. If you're handling an older ticket it might be that the customer has already updated their system or moved locations before you reply.
Secondly, since the information shown is linked to the user, you can leverage Messaging Authentication to make sure all tickets a user starts over Messaging are mapped to the same profile.
[Messaging Authentication: Verified email and merging existing users based on emailZendesk introduced a new email verification flow to handle the mapping of authenticated Messaging users and exiting end-user profiles. It’s a lot so let’s dive in!Internal NoteThomas Verschoren](https://internalnote.com/messaging-authentication-identify-and-merge-existing-users/)
# Exporting the data
Having this data available in the Agent Workspace is useful to resolve tickets easily. But sometimes you also want to access that data outside of Zendesk for reporting purposes.
One way to capture the data is by leveraging Zendesk's' (new) GraphQL API located at `https://subdomain.zendesk.com/api/lotus/graphql`
By executing a `POST` to this endpoint with the payload and variables below, you can get a nice JSON object that contains all the user information shown next to a ticket.
Authentication is done via Basic Authentication: admin@domain.com/token and a Zendesk API token.
### GraphQL Body
```graphql
query UserDeviceMetadata($id: ID!) {
user(id: $id) {
id
... on Customer {
deviceMetadata {
devicePlatform
id
ipAddress
lastSeen
location {
city
country
countryCode
stateCode
__typename
}
os
osVersion
userAgent
__typename
}
__typename
}
__typename
}
}
```
### Variables
```JSON
{"id":"24820449367569"} //user id
```
### Returned Data
The returned data contains all the items shown next to tickets, so you can export City, Country, State, IP, OS and much more.
```JSON
{
"data": {
"user": {
"id": "24820449367569",
"deviceMetadata": {
"devicePlatform": "Macintosh",
"id": "66388e17dde783733a88950c",
"ipAddress": "42.42.42.42",
"lastSeen": "2024-05-06T08:00:44.350Z",
"location": {
"city": "Antwerp",
"country": "Belgium",
"countryCode": "BE",
"stateCode": "VLG",
"__typename": "DeviceMetadataLocation"
},
"os": "macOS",
"osVersion": "14.4.1",
"userAgent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
"__typename": "DeviceMetadata"
},
"__typename": "Customer"
}
},
"extensions": {
"depth": 4,
"potentialNodeCount": 3
}
}
```
## How would you use this?
One way to use this is by leveraging an external tool like Zapier or Made, or your own scripts and setup a webhook listener. Each time a messaging ticket created, use a Zendesk trigger to send the `{{ticket.requester.id}}` to your webhook listener.
Once triggers, your external tool can call this API and pull in the users' metadata and store it in a log, BigQuery, Google Sheet or whatever suits your needs to log data.
By sending both the `{{ticket.id}}` and `{{ticket.requester.id}}` in your webhook trigger, you can even map both those elements in your log for more data.
# Conclusion
It's nice to see that Zendesk makes these kinds of information available as both a native feature in the Agent Workspace, as well as available over API for developers and data managers. This way you can leverage these new features immediately, while also integrating them deeper in your flows.
From a feature set I'd love to see these values being made available in triggers and bot flows though. "A customer that complains about your app not working running a version of the operating system that's too old" might be handled via a custom bot answer by setting up a condition "if OS Version < 16", show bot message "This version if your system is not supported, please update!"
### Handling multiple support addresses in a support email
URL: https://internalnote.com/multiple-recipients/
Last updated: 2025-09-08T06:41:17.000Z
You might recognize this scenario: a customer contacts your support team and adds support@company.com, sales@company.com and finance@company.com all in cc on the same email.
This email arrives in Zendesk and only one tickets gets created, and the fact that two other support addresses were part of the same email is invisible for your agents.
Only the first support email address shows up
# How to fix this
One easy way to find out if an email has been send to multiple support addresses, is by opening a ticket and going to the *View Original Email* option. This however is not the most discoverable way, and you need to be actively looking for such emails.
## A better solution
A better method would be to somehow detect multiple support addresses in cc, and then clone the ticket for each of these addresses. That would create duplicate emails, but each email would be routed and handled as it if were emailed solely to that specific support address and assigned to the correct group.
The confusion that may come from those duplicated tickets is solved by those tickets showing up in the customers' *Interaction history* in the context panel, or would be picked up by the *Merge suggestions* of Zendesk Advanced AI.
# Flow
To build this flow we're going to need a Cloudflare Worker, or a platform like [Make.com](https://make.com/?ref=internalnote.com), Zapier or similar.
😫
I tried to make this flow with [make.com](https://make.com/?ref=internalnote.com) to do it as low code. Failed miserably because I'm just not familiar with those abstractions and faster in plain javascript code. If anyone knows how to loop through an array, let me know!
To make this flow work we need a few elements which we can all grab via the Zendesk API.
- If you want to detect all the support addresses in cc on an email ticket we can use the Audit Log (Events) of a ticket.
- To know if an address is a support address registered in Zendesk or a regular end-user, we need a list of all support addresses in the instance
- And we need a copy of the original Zendesk ticket in order to clone it.
Once we have these elements, we can use them in a flow like the one in the diagram below:
# Setup
## Webhook
Once you've created your Worker (see below) or other webhook endpoint we need to add a Webhook to Zendesk. Note that we need to pass the `ticket.id` to the worker, so make sure the request method is a POST.
## Trigger
Next, create a trigger that fires when a ticket is created, and only for the Email channel. This prevents unnecessary calls to our worker since only email tickets can contain those multiple support addresses.
In the actions, select the *Notify by | Active Webhook | Multiple recipients* web hook we just created and pass the following payload
```json
{"ticket": "{{ticket.id}}"}
```
💡
I opted to pass only the ticket id in this step, and retrieve all ticket info via API in the worker. You could pass the subject, description e.a. in this step too, but by retrieving it over API in the worker itself our trigger is easier to set up.
## Worker
All code for this flow is available on GitHub. You can fork the repository, and use the *Deploy to Cloudflare* button to create your own version of this script **for free**.
[GitHub - verschoren/multiple\_recipientsContribute to verschoren/multiple\_recipients development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/multiple%5Frecipients?ref=internalnote.com)
One you've deployed the script successfully there's two things you need to change on the Cloudflare Worker variable page:
- Replace the **SUBDOMAIN** with your own Zendesk subdomain
- Add a **TOKEN**, make it a secret, and use a base64 encoded string of `admin@domain.com/token:zendesktoken` where `zendesktoken` is an actual token taken from the Zendesk admin center, and `admin@domain.com` is a real admin in your instance.
# Code deep dive
## Get Data
The first step in our code is grabbing the POST payload from our trigger:
```javascript
var payload = await request.json();
const ticket_id = payload.ticket ? payload.ticket : "";
```
Next we have three functions to grab the ticket info, ticket audit log and support email addresses.
### Get ticket info
```javascript
async function getTicket(ticket_id, subdomain, headers) {
const init = {
method: "GET",
headers
};
const result = await fetch("https://subdomain.zendesk.com/api/v2/tickets/42.json", init);
let json = await result.json();
return json.ticket;
}
```
This function calls the Zendesk API and returns a JSON element with all ticket information:
```json
{
"ticket": {
"id": "42",
"subject":"What happens at 88mph?",
"description":"We're going back in time"
"requester_id":"123456789",
"recipient":"support@domain.com",
//...
}
}
```
### Get audit log
```javascript
async function getAudit(ticket_id, subdomain, headers) {
const init = {
method: "GET",
headers
};
const result = await fetch("https://subdomain.zendesk.com/api/v2/tickets/42/audits", init);
let json = await result.json();
let audits = json.audits;
for (const audit of audits) {
if (audit.via.channel == 'email'){
var recipients = audit.via.source.from.original_recipients;
return recipients;
}
}
}
```
This API returns the audit log of our ticket, which conveniently contains a list of `original_recipients` under the via|email channel key:
```json
{
"audits": [
{
"ticket_id": 1841,
"via": {
"channel": "email",
"source": {
"from": {
"address": "m.mcfly@example.com",
"name": "Marty McFly",
"original_recipients": [
"note@internalnote.com",
"support@internalnote.com",
"support@verschoren.com",
]
},
}
}
//....
}
]
}
```
### Get support addresses
The final step in our preparation is getting all the support addresses via the API. We parse the returned JSON and create a list of all support addresses in the instance called `support_emails`
```javascript
async function getSupportAddress(subdomain, headers) {
const init = {
method: "GET",
headers
};
const result = await fetch("https://subdomain.zendesk.com/api/v2/recipient_addresses", init);
let json = await result.json();
var support_emails = [];
for (const element of json.recipient_addresses) {
support_emails.push(element.email);
}
return support_emails;
}
```
## Compare
Now that we have our required data, we can start processing the ticket to check if we need to create duplicates for additional support addresses in the original email.
```javascript
for (const recipient of recipients) {
//run check
}
```
### Filter out current recipient
First check we want to do is filter out the support address that was used for the ticket that **did** get created. Since this email address got accepted by Zendesk, it's no use making another copy!
```javascript
if (ticket.recipient == recipient) {
console.log(`Skipped. Existing ticket for ${recipient}`);
}
```
### Check if support email
Secondly we want to filter out any email that isn't a support address. The customers could've cc'd a friend, spouse or colleague and those should not trigger a duplicate ticket.
```javascript
else if (!support_emails.includes(recipient)) {
console.log(`Skipped. ${recipient} is not a support email.`);
}
```
And finally, if the ticket is a support email and isn't the original recipient, we want to create a ticket!
```javascript
else {
var new_ticket = await createTicket(ticket, recipient, ticket_id, subdomain, headers);
}
```
## Create ticket
Since this is an email ticket we only care about the subject, description and requester. Email cannot set ticket fields or other metadata, so we don't need to pass these along.
💡
I did not add attachments to these duplicate tickets. Since Zendesk charges for data these days I opted to just reference the original ticket so you can look at any attachments in that ticket.
First step is to define our ticket.
- Note we used a via.channel "mail" element to make sure the duplicate ticket is also processed as an email ticket!
- The recipient value is the support addresses that passed our filters in the previous steps.
```javascript
var payload = {
"ticket": {
"subject": ticket.subject,
"comment": {
"html_body": `Split from ticket #${ticket_id}` to the Content Blocks feature. This month the list of tags got expanded – as announced [late last](https://internalnote.com/roundup-2024-01/#%F0%9F%94%8E-help-center-and-self-service) year – with a whole new set of supported elements, which makes building nice CTA blocks to be used across article that much easier!
```
`a`, `abbr`, `address`, `aside`, `bdi`, `bdo`, `blockquote`,
`br`, `caption`, `cite`, `code`, `col`, `colgroup`, `data`, `dd`,
`del`, `details`, `dfn`, `div`, `dl`, `dt`, `em`, `figcaption`,
`figure`, `h1`, `h2`, `h3`, `h4`, `h5`, `h6`, `hr`, `iframe`,
`img`, `ins`, `kbd`, `li`, `mark`, `ol`, `p`, `pre`, `q`, `rp`,
`rt`, `ruby`, `s`, `samp`, `small`, `source`, `span`, `strong`,
`sub`, `summary`, `sup`, `table`, `tbody`, `td`, `tfoot`, `th`,
`thead`, `time`, `tr`, `track`, `ul`, `var`, `video`
```
The [Generative AI feature for Zendesk Guide](https://internalnote.com/preview-of-the-new-generative-ai-for-knowledge-in-zendesk/) got expanded with a new Tone that will simplify the written text into shorter, less complex text. Perfect for turning an article written by your developers into something end-users actually understand 😅.
And to conclude, the Request List for end-users in the Help Center now supports the [Custom Status](https://internalnote.com/tag/custom-ticket-status/) feature.
## 🧱 Open and Flexible Platform
First off, it’s now possible for end-users to [enable 2FA](https://support.zendesk.com/hc/en-us/articles/6584372830362-Announcing-two-factor-authentication-for-end-users?ref=internalnote.com) on their accounts. This is enabled by default for all instances and doesn’t require any admin settings.
Secondly, for those using [Custom Objects](https://internalnote.com/tag/custom-objects/) you can now hide object types for your agents. So if you have objects that are only surfaced in a third party app, or are used to store data that should be directly interacted with by agents you can now hide those in the Object Viewer. Or similarly, if you have object types you don't use anymore, you can hide them instead of deleting them.
Zendesk AI got also expanded with a new Industry: Travel. This leaves HR, IT, Hospitality still on the announced roadmap. (Although I think most of us are waiting for fully custom models based on your own data like [Ultimate.ai](https://ultimate.ai/?utm%5Fsource=internalnote) offers)
# 💡Insights
## Demystifying LLMs
[Demystifying LLMs: Ultimate’s take on Next-Gen Conversational AI – Ultimate. RnD BlogThe rapidly evolving landscape of AI has witnessed an influx of new terminology, acronyms and buzzwords, such as GenAI, LLM, Prompt Engineering, and others. This article aims to demystify these terms, focussing on Large Language Models (LLMs), which have emerged as game-changers, blending technological prowess with vast capabilities. We will discuss their capabilities, how they are trained and look at both their risks and their potential as well as briefly touch on how Ultimate embraced these models to revolutionize customer Interactions.Meysam Asgari-Chenaghlu, Staff AI Researcher](https://rnd.ultimate.ai/blog/large-language-models-and-generative-ai?ref=internalnote.com)
The team at [Ultimate.ai](https://rnd.ultimate.ai/blog/large-language-models-and-generative-ai?ref=internalnote.com) has created a development focus blog where they dive into the technical side of their product. One of the first articles offers a deep dive on how LLMs for Conversational AI really works. Worth a read!
## Extreme brainstorming questions
[Extreme brainstorming questions to trigger new, better ideasWe know, “no idea is a bad idea,” but brainstorming is often unsuccessful. These prompts actually work. They could even lead to a unique business model.A Smart BearJason Cohen](https://longform.asmartbear.com/extreme-questions/?ref=internalnote.com)
This article focuses on brainstorming, but also posed an interesting question: what would you need to setup in order to remove the need of an actual customer support team?
> If you were never allowed to provide tech support, in any form, what would have to change?
My 3 big ticket items are:
- a knowledge base to gather documentation
- a way to have self-service based on that info on an FAQ or Chatbot
- and a feedback loop that detect gaps in your knowledge.
What are yours?
## Help Center PDF Export
The people at SwiftEQ wrote up a nice tutorial on how to add a print option to your Help Center that outputs nicely formatted PDFs with your company’s logo.
[How to Enable PDF Export for your Zendesk Help Center ArticlesImage by vectorjuice on Freepik Welcome to a step-by-step guide on enhancing your Zendesk Help Center! By the end of this tutorial, you’ll have successfully implemented an option for visitors to download or export articles as PDFs in a professional format using the printing function of their browser (Ctrl+P). This feature not only adds value for your customers by allowing them to easily download articles for offline use, sharing, or printing, but also enhances the overall user experience of yourswifteqSorin Alupoaie](https://www.swifteq.com/post/export-pdf-zendesk-help-center%20?ref=internalnote.com)
# 🎥 Videos
I spoke on a recent event from [Ultimate.ai](https://www.ultimate.ai/webinars/scale-smarter-with-generative-ai-conversation-design?utm%5Fsource=internalnote) and they just put the conversation on Youtube. Apologies for the terrible freeze-frame 😅.
# ⚠ Major Changes
### Legacy Social Messaging deactivation
> Effective July 31, 2024, the legacy Social Messaging app will cease to function and customers will no longer be able to send or receive messages using the legacy app. To retain messaging functionality, you will need to set up social messaging within Agent Workspace by following the instructions linked below. - [link](https://support.zendesk.com/hc/en-us/articles/6558217474458?ref=internalnote.com)
Ye be warned 🏴☠️
### Automatic activation of Omnichannel Routing
> Starting April 2024 Zendesk will start the automatic activation of Omnichannel Routing for specific accounts. Once enabled agents will have the use the new Unified status updates to set themselves available for chat, phone or ticketing assignments. - [link](https://support.zendesk.com/hc/en-us/articles/5716181238938?ref=internalnote.com)
Zendesk will automatically enable the feature, but as long as you don't setup a routing tag or tweak your triggers to make use of it, you won't notice any difference.
# 📝 Articles this month
[Managing Storage Limits in ZendeskThis article gives insights in how to manage your data usage limits in Zendesk, how to bulk delete specific tickets and how to remove attachments to keep your data usage under the allowed limits and reduce the risk of paying more!Internal NoteThomas Verschoren](https://internalnote.com/storage-limits/)
[➕ Zendesk Security ChecklistOne of the big Trends for 2024 is the idea that security no longer is an add-on but should be seamlessly incorporated throughout the customer journey. To get started the right way, I’ve written a Zendesk Security Checklist. It’s a practical approach to improve the security of your Zendesk instance.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-security-checklist/)
[Zendesk CX Trends 2024Zendesk just launched their Trends Report for 2024 giving insight into where Customer Experience and support is moving towards this year. I’ve written up my own overview of the trends and how I see the product moving forward influenced by these trends.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-cxtrends-2024/)
[Answer Linking for the Zendesk BotAs a first big release for the Zendesk Bot this year, Zendesk is introducing Answer Linking, which can turn answers into reusable blocks!Internal NoteThomas Verschoren](https://internalnote.com/answer-linking-for-the-zendesk-bot/)
[Restarting a Zendesk Messaging ConversationThis tutorial shows you how to build an Answer flow for Zendesk Messaging that allows visitors to restart the conversation.Internal NoteThomas Verschoren](https://internalnote.com/restarting-a-zendesk-chat-conversation/)
# And Finally...
> "We're building so many things in 2024, but one product my team is especially looking forward to building is Dark Mode as a native setting across Zendesk Support, one of our highest requested features from agents. It's not just about aesthetics. Agents are starting at our default bright white screen 8-hours-a-day, 5-days-a-week, which ultimately ends up causing strain and drains agent efficiency. This also opens up the door for theming - which is a way for customers to customize their entire Ul with their company branding/colours." - JJ Miclat, Group Product Manager Agent Workspace.
🎉
### Restarting a Zendesk Messaging Conversation
URL: https://internalnote.com/restarting-a-zendesk-chat-conversation/
Last updated: 2024-04-08T07:54:47.000Z
At the end of the [Answer Linking for Zendesk Bot](https://internalnote.com/answer-linking-for-the-zendesk-bot/) article I quickly mentioned it's now possible to restart a conversation in the Messaging Widget. I didn't really go into details, so enjoy this short bonus article that explains you how!
[Answer Linking for the Zendesk BotAs a first big release for the Zendesk Bot this year, Zendesk is introducing Answer Linking, which can turn answers into reusable blocks!Internal NoteThomas Verschoren](https://internalnote.com/answer-linking-for-the-zendesk-bot/)
## Sign up for Internal Note
The blog about Zendesk with a focus on Ticket Automation and AI
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
## How to restart a conversation
With the new Answer Linking you can link back to the beginning of the conversation. So if we create a custom Answer that only applies this step, we can restart the conversation for our customers.
Setting it up requires only a few short steps:
1. Create a new Answer for your bot, called *Restart Conversation*
2. Give it a few training phrases like *"Reset", "Restart Conversation", "Back to the beginning"*
3. The Bot flow exists out of one step: a "Link Answer" step that links back to the start of the conversation.
Once you set it up, whenever a customer interacts with your bot and indicates they want to restart the conversation, the conversation will restart with your welcome message.
Enjoy.
➕
If you like this kind of content and want to support this blog, take a look at [Internal Note Plus](https://internalnote.com/plus), a paid tier with more content.
### Answer Linking for the Zendesk Bot
URL: https://internalnote.com/answer-linking-for-the-zendesk-bot/
Last updated: 2024-08-19T20:34:40.000Z
Over the last year Zendesk has been investing heavily in their Zendesk Bot both from an Intent Mapping and from a Flow Builder standpoint.
At the end of last year they introduced [Dynamic Bot experiences](https://internalnote.com/dynamic-conversation-experiences/) which allow more powerful API integrations with support for variables, arrays and dynamic carousels. This release removed some big ticket items from my feature wish list.
This week, as the first big release for 2024, they are introducing [**Answer Linking**](https://support.zendesk.com/hc/en-us/articles/6636652095770-Announcing-answer-linking-for-Zendesk-bots?ref=internalnote.com) as a way to make big complex bot flows more easy to manage.
# Answers
When Zendesk introduced the Zendesk Bot – or Answer Bot as it was called then – they relied almost entirely on Flow Builder to build the bot with steps, flows and conditions. You can add comments, offer options and buttons, and have conditional flows to route customers to articles, external pages or, if needed, agents.
With the arrival of Zendesk AI they changed their approach to the bot and made it intent based. Now, out of the box, the bot will read a customers' comment and propose up to three Help Center articles, either as direct links with snippets, or as generated replies via Zendesk AI.
For more complex flows where we need more information from the customer like an order number, or want to route the customer to a series of questions to detect the right solution for their problem, Zendesk still offers these custom build flows under the section *Answers.* These answers can be triggered with specific keywords or intents and allow you to build more complex purpose-built flows for your customers.
This way, your self service approach can scale from very broad to very specific within the same bot.
The evolution of Zendesk Bot
Ideally you deflect the most requested (and easy) questions via Help Center articles and generated replies to those questions. Then, for more complex flows, you can build an Answer with steps and options to guide the customer to an answer, or to pull in external information. And for the most complex and unique questions that require a human touch, you still have your agents.
"*How do I reset my password?*", can be answered with an article. "*Where is my order?*" might require an Answer flow with an API call. "*My hamster escaped on the plane*" needs a real agent to assist. And hopefully we can move more and more customer questions from Agent towards automated articles or Answers.
# Custom Flows
When building custom flows to assist your customers you quickly run into the scenario where multiple answers require the same steps. Each question about a malfuncting smartphone requires a software update or reset. Each complex scenario you didn't account for requires an escalation to an agent. Each customer who needs to access your system might need the steps to reset a password or remember their account.
Until now there were two approaches to handle these repeated scenarios. You can copy-paste all the required steps across all the custom answers you build. Or you could end each flow with a "did this help? ⟶ "No" and ask them to rephrase the question, or tell the customer to ask for "how to reset my password"
However you turn it, it's not ideal. A change in the password reset flow requires you to update every answer that references it, and good luck remembering all of them.
# Linked Answers
This week Zendesk released a new *Step Type* for Flow Builder. This new type allows you to link to another Answer within another flow. This means we can take the flow above, that has repeated Agent Escalations steps and split it into two.
One flow contains the *"Talk to an agent"* steps, including a check for Business Hours, and an Ask for Details step to ask for name and email.
And the original complex flow replaces the entire Transfer to Agent flow with a new Linked Answer step that links to the new "Talk to an agent" flow.
And similarly, we can use this same "*Talk to an Agent"* step and use it in another flow
# Best Practices
Now that we can reuse flows and link to them, it might become best practice to go over your existing flows and identify any flows you re-use. Good candidates are escalation flows or lookups via API.
You can then pick one of these recurring flows and copy the steps to a new Answer. I like to prefix mine with \[BLOCK\] so I know this is a block I use a lot. Then go over all your flows once again and replace these recurring flows with your new blocks.
Similarly, if you have very big flows you can now break them up into smaller chunks and create one flow that handles the routing (e.g. question about iPhone, iPad, MacBook) and then link to the answer flows that handle these three types. Your flow builder trees become a lot smaller per Answer, and a lot more more handy to navigate.
# Caveats
There's a few caveats with this new step type:
- The *Transfer to Agent* steps allow you to add a `tag` to the escalation. Naturally, all answers that link to a transfer step with a tag will contain the same tag now.
- Any *Ask for Details* requested in previous steps are forwarded to the ticket if you escalate to an agent (e.g. Order Number). However, if you create a variable, those are not usable across Answers. So if you get the Order Status in your "Order Status" flow via API, you can't use that status as a variable in steps of a linked answer.
💡
In [comments](https://support.zendesk.com/hc/en-us/articles/6636652095770/comments/6646695864858?ref=internalnote.com#comment%5F6646695864858) on the original post the Product Manager gave this feedback:
"Currently, variables and tags can’t be carried across linked answers. However, we are already working on supporting this with our next release scheduled for early Q2."
- If you asked for data via *Ask for Details*, and the next flow references that same Field in its own *Ask for Details* field, it will already be prefilled with the previous data.
- You can't currently link an Answer to itself (luckily). You can however create a loop by linking Answer A to Answer B, Answer B to Answer C, and Answer C back to Answer A. So be careful not create a loop.
# Conclusion
This step type is a welcome addition to the bot building flow that will make many complex flows a lot smaller and easier maintainable.
Are there things I'd like to see? Sure, I'd like to put an Answer in the middle of a flow and have it process input (e.g. an API call to get order status might be used in the middle of multiple flows for different purposes). But even without that, I'll welcome this new Linked Step type with open arms!
💡
The Linked Answers step also allows you to link back to ****Start of the Conversation**! So if used wisely, it can be used as a reset button at the end of specific flows if the customer gets stuck.
### Zendesk CX Trends 2024
URL: https://internalnote.com/zendesk-cxtrends-2024/
Last updated: 2024-01-26T14:58:07.000Z
Zendesk just launched their [Trends Report for 2024](https://premiumplus.io/the-10-cx-trends-driving-2024-powered-by-zendesk/?utm%5Fsource=InternalNote&utm%5Fmedium=Article&utm%5Fcampaign=CXTrends) giving insight into where Customer Experience and support is moving towards this year.
Last year's report had 5 big trends:
1. AI experiences are becoming more evolved and seamless
2. Conversational experiences are empowering consumers
3. Customers are eager for deeper personalisation
4. Consumer well-being and sentiment are reshaping CX
5. CX teams are breaking down silos as they become more integrated
Looking back at those trends today, you can clearly see that Zendesk made big shifts and improvements in their platform to accommodate these trends. The launch of Zendesk AI, major expansions in their Messaging platform and bots, and improvements in Agent Workspace with regards to Teams integrations, Slack bots, and custom objects, made sure that Zendesk could offer a technical solution for each of the trends.
This year's report doubles the amount of trends with a whopping ten insights that will allow you to *Unlock the power of intelligent CX*.
The trends are split across three Innovation Areas: AI & intelligent Experiences, Data & Trustworthy Experiences, and Next Gen & Immersive Experiences.
Conveniently, these three areas map nicely into two new add-ons Zendesk launched in 2023: *Zendesk* *Advanced AI* and *Advanced Customer Data and Privacy Protection*, with the third area partially covered with the acquisition of Tymeshift last year.
Let's dive in.
# AI & Intelligent Experiences
## Trend 1: Generative AI will accelerate the delivery of a more humanised journey that feels personable & interactive.
> But can generative AI help businesses provide the kind of warm, human service that feels like interactions with local, family-owned businesses? More than two-thirds of CX organisations think that it will help their business provide that warmth and familiarity, even if they serve millions of customers.
One of the main benefits I find in these new Generative AI-driven bots like those from Zendesk or Ultimate is the fact that they work completely differently than the decision tree-based bots from before.
It used to be that a bot was the chat equivalent of a phone IVR. Do you want sales or support? Support. What kind of question do you have? "something broke", "product support", "upgrade questions" and so on, until you hopefully either make the right decisions and get an answer, get completely frustrated and restart the conversation until you find the right incantation, or start typing human, human, human, until you reach an agent.
Now with Generative AI bots instead of me searching for an answer, I can just ask a question "How do I update my iPhone?" and the bot will reply with the answer without the need of navigating a decision tree of predefined options.
That combined with a specific tone of voice, context from my previous questions and, hopefully, the data the company has on me, will result in better answers and quicker resolutions.
Is Zendesk there yet to fulfil the needs of this trend? Partially. Asking a question and getting data from your Help Center works like a charm, but there's no actual conversation happening here, and the data available to the bot is limited to the Help Center, the Zendesk AI bot cannot execute API calls or collect information from external or internal resources to give a richer answer with more context.
💡
Coincidentally, the Ultimate webinar where I [spoke](https://www.ultimate.ai/webinars/scale-smarter-with-generative-ai-conversation-design?ref=internalnote.com) last week did cover these kinds of richer hybrid AI + integrations approach that's currently lacking in Zendesk.
## Trend 2: Chatbots are rapidly transforming into digital agents that have the capacity to do more.
> Chatbots are doing a better job at surfacing information quickly, and they're also improving their ability to tailor responses to better fit where customers are in their journey.
A good customer care approach has always and will always revolve around self-service and knowledge management. Even before chatbots, having a knowledge base with your top 10 support inquiries and a system in place that can automatically reply with those topics will dramatically reduce the workload of your customer care team.
Chatbots and AI offer a more powerful version of this since it improves on two aspects:
1. It can turn a customer inquiry into an intent and have a better grasp of what the customer needs instead of doing a basic keyword search in the knowledge base.
2. It can turn the available knowledge sources into a custom answer tailored to the needs of the customer, instead of replying with a link to the article or the exact article contents.
It comes as no surprise that these two improvements are what make agents better at handling support than regular search, and now that bots have these same capabilities, even more customer questions can be handled by the bot without any human interactions.
What's missing is the more complex workflows. Autonomously handling refunds, booking changes, reservations... Some of these are already possible by using the API endpoints for the Zendesk Bot, and I'm sure we'll be seeing more of these items become natively available in the Zendesk Bot once elements like [Conversational Commerce](https://www.zendesk.com/blog/conversational-commerce/?ref=internalnote.com#:~:text=Conversational%20commerce%20refers%20to%20how,Last%20updated%20August%2016%2C%202023) become more widely available within Zendesk.
## Trend 3: Disconnect grows between CX leaders and agents on everything related to AI; strategy, tools, and role impact.
> The problem is that many agents aren't so sure. They look at the new tools with a mix of wariness and deflation, fearing what AI will mean for their job security.
The big promise of AI is that it can automate a lot more processes and lower the workload for agents. It generates insights by tagging each conversation and automates offering answers to customer inquiries.
This gain in efficiency can have two results which are on the opposite side of the spectrum. On one hand, you can start from the status quo. You're kind of happy with your CSAT and First Reply Time metrics and see the reduction in tickets as a way to reduce headcount. Less work means fewer people needed to do that work while not really impacting your customer experience in a negative way.
On the flip side is the idea that you can use that extra time to improve your customer support. You can give agents actual time to dive into the complex issues or you can use that time to look into processes and documentation. The former allows you to have deeper, more personal conversations with your customers and really solve their issue, the latter improves self-service and offers more knowledge to your bot to deflect even more tickets.
However you look at it, if you use the impact AI can have on your support interactions as a trigger to improve, you create room for agents to grow and become experts in handling cases or documentation.
So it's important that, as a company, you lay out your vision and choose to go for quality or quantity and be transparent on why you implement AI in your company and what you will do with the efficiency gain. You can choose the short-term solution of reducing headcount and freeing up costs. Or you can go for the long-term approach of creating a cycle of continuously improving your customer experience by shifting agents towards more complex roles.
## Trend 4: AI transparency and decision-making are now the rule, not the exception
> The need for trust through transparency becomes especially important when dealing with sensitive information that reveals a person's identity, their health history, and their financial status. No customer wants to feel like AI is snooping on them or making their data available to bad actors. As 58 percent of consumers told Zendesk, knowing how their data is collected, stored, and used plays an outsized role in whether they'll purchase a product or service from a company.
The answer to the question "who owns or runs your AI model" will become a hot topic this year.
Where 2023 was the year of "everyone uses OpenAI" as a way to hack AI into your company, I think 2024 will become the year where the question "who owns our model" is the big question to ask when enabling AI in your company. You want to be sure answers are based on actual facts, and you want to be sure that it's only you who can use your own data, and not a competitor.
Zendesk has its own AI models trained on Zendesk Data, and made available to their customers. The answers generated to reply to tickets are based **solely** on the data in your Help Center or your own ticket data in the case of suggested macros. It only uses OpenAI to generate a reply, but the data used to generate the reply is 100% controlled by Zendesk, and in the case of intent or sentiment mapping, OpenAI is not even involved.
As a company, this [Mark Zuckerberg](https://www.youtube.com/watch?v=9aCg7jH4S1w&ref=internalnote.com) vision is a sure thing within Zendesk. Your customers can only get answers based on data you made available to Zendesk.
> Yeah, so our view is that there’s actually going to be a lot of these that people talk to you for different things. \[...\] let’s say you’re a small business and you want to have an AI that can help you interface with customers to do sales and support. You want to be pretty confident that
>
> your AI isn’t going to be promoting your competitor’s products, right?"
As a customer, the question about trust is twofold. On one side it's the question "can I trust the answer I get from the bot", and on the other side it's the question "can I trust this bot or company to handle my data correctly".
One is about making sure that the answer "can I melt an egg" is replied to correctly. The other is about "if I enter my birthday here, how securely do they store it". I don't think we'll get to a point where customers will see the Zendesk Logo on a chatbot as a way to know "hey this is safe". But I do think that a quality bot with good answers will automatically infer a feeling of trust that will give customers assurance that they can *trust* the bot with their personal data.
# Data & Trustworthy Experiences
## Trend 5: Businesses are heavily focused on being able to instantly modify user experiences, putting increased pressure on leveraging data in real time.
> That preference means companies must focus on boosting their bot capabilities via Al, specifically using its power to capture and analyse sentiment and intent. Doing so will help businesses predict customer needs and resolve issues quickly and efficiently (including knowing when a bot needs to hand over an interaction to a human agent).
For me one of the main benefits of bots powered by Generative AI is the aforementioned idea that they allow customers to ask questions instead of customers navigating a flow to find an answer.
- A generated response based on knowledge base data (or other sources)
- An API powered flow that pulls in context (e.g. order status) to reply to the customer
- An escalation to the right agent based on intent, sentiment and conversation content
Most customer interactions possible today (In Zendesk) are very closely linked to indexed knowledge base content as a basis for an automated self service approach. But recent updates in e.g the Zendesk Bot for more [dynamic conversation experiences](https://internalnote.com/dynamic-conversation-experiences/) make the bot capable of pulling in external data to offer more complex answer flows.
I really hope Zendesk will keep improving that part of their bot in the next year, making it easier to pull in data from CRM, webshop or other systems and combine that data with Generative AI to have more complex and data-driven conversations with customers.
So now that the Knowledge Based bots are a solved matter from a technical standpoint, the focus this year will shift towards conversations powered by other types of data, further reducing the need for agents to do these repetitive lookups, and once again giving them more time to dive into the complex and unique questions.
## Trend 6: CX leaders are the new drivers of data privacy as AI & personalisation take on a greater role.
> They know that it's not good enough to simply have Al tools for personalisation; those solutions must keep customer data secure.
Looking back at Relate 2023 it's almost surprising this trend wasn't Trend 1 in this report. At the keynote they showed off this awesome slide which kinda sums up the entire trend.
There's a couple of benefits of powering your customer interactions with AI. One benefit is that you can better capture intent, which also means that instead of asking a lot of data in a form, you can ask for only the data needed to solve the inquiry, since you can closely match question and intent.
Within Zendesk we've now got [automatic ticket deletion](https://internalnote.com/storage-limits/), build-in redaction tools in the Agent Workspace and more secure integrations to make sure data is shared correctly across platforms.
Similarly, if we can have API integrations in the bot that pull in the data directly from the source, we need to pass less data to agents, or store less data in tickets, since we can resolve inquiries before they even reach the ticket stage.
It's all small changes, but they can amount to a big reduction in customer data stored, while not loosing the ability to correctly and swiftly resolve their inquiries.
## Trend 7: Security is no longer an add-on but is seamlessly incorporated throughout the customer journey.
> Thankfully, CX leaders have options for seamlessly integrating security measures into customer experiences, most of which aren't new: multi-factor authentication, encryption of service interactions, and being transparent with customers about security and data privacy practices.
Security and Privacy go hand in hand, and this is a trend one where Zendesk is has made big efforts in the last 12 months, starting with a renewed focus on Privacy and Security at [Relate 2023](https://internalnote.com/relate-2023/#%F0%9F%94%92-privacy) and the release of more complex role and permission settings, messaging and end-user authentication, an entire new security add-on and more.
If you're interesting in securing your Zendesk instance, take a look at my Security checklist published earlier this month.
[Zendesk Security ChecklistOne of the big Trends for 2024 is the idea that security no longer is an add-on but should be seamlessly incorporated throughout the customer journey. To get started the right way, I’ve written a Zendesk Security Checklist. It’s a practical approach to improve the security of your Zendesk instance.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-security-checklist/)
# Next Gen & Immersive Experiences
## Trend 8: Live and immersive experiences are now heavily influencing the future of online shopping.
> Now, 80 percent of consumers expect chat agents and support representatives to assist them with everything they need. The line between support and sales has begun to blur.
Remember [Zendesk Labs](https://zendesklabs.zendesk.com/hc/en-us?ref=internalnote.com) announced at Relate 2023? We've still not seen any actual feature releases or EAPs for their idea of conversation commerce, even though this concept is closely related to this eight trend.
As mentioned in one of the earlier trends, shifting conversations towards bots gives agents the room to dive into the unique issues. Similarly, the more complex scenarios bots can handle, the more is expected of the bot by consumers.
In a classic scenario I might get an article about the return policy, or in a AI powered scenario I might get a custom reply based on my question.
For example, if I booked a flight but want to move my return journey one day forward. My first step would be to open the airline's app and view my booking and try to change the booking there. If that doesn't work, I'd either open the in-app chat and ask for "how to change my return flight," or even better, I could get a proactive bot message asking me about my booking changes.
But even if the answer I get is "yes, this is possible" or "no, it's not possible, you need to book another flight," to actually resolve my question, the chatbot or agent would need context (who am I, what flight, which ticket type), would need power (update, cancel, reschedule bookings), and would need to do it right then and there in the chat.
Is this a sales scenario? Yes, since I need to potentially buy a new ticket or pay a surplus for the rescheduled journey. Or is this a support scenario? Yes, because I don't know if it's possible and how to reschedule my ticket.
## Trend 9: Voice is carving out a more advanced role focused on handling complex and escalated issues.
> When companies create a seamless transition from digital to voice channels for handling complex issues, consumer confidence in the former rises. That in turn becomes a virtuous cycle in which customers feel increasingly satisfied with reaching out via digital channels, thus lessening demand for phone options.
People who know me, know I'm not the biggest fan of voice as a customer care channel. Voice requires both the customer and the agent on the other side to make time **now** to handle and resolve an issue, whereas more often than not an asynchronous approach of asking a question and getting an informed response or solution offered an hour later is often way more rewarding.
This idea, sadly, starts from the negative idea that most customer care teams are not able to resolve issues now and then, and that the person picking up the phone is often not enabled to make the decisions or changes needed to resolve an issue.
However, like the trend report mentions, if a lot of customer interactions shift towards the automated chat channels I prefer, this actually frees up time for agents to handle only the complex issues that require an actual human interaction. And, just like with Siri or Alexa, asking via voice to add "coffee to the shopping list' is way faster than typing it out, and a phone call or conversation does allow for a more nuanced discussion often needed to handle these complex issues.
So if companies follow the idea I mentioned earlier of using the efficiency gain made possible via AI to give agents the room to dive into the complex issues, then voice might become the logical channel to have those longer and deeper conversations that actually solve my issue.[an easy way](https://internalnote.com/voice-api-for-zendesk/) to turn a chat conversation into a call via the web widget, and the new [Generative AI for Voice](https://internalnote.com/preview-of-the-new-generative-ai-for-voice/) turns long conversations into short summaries, allowing for easier escalation to the right person to handle the ticket.
## Trend 10: Predictive agent management tools are finally eclipsing traditional methods.
> However, leaders now have agent management tools at their disposal that can take a lot of the guesswork out of running a support operation. These tools can offer both operational and strategic foresight, helping managers make better staffing and training choices.
It's no coincidence that, in the last twelve months, Zendesk has bought both [Tymeshift](https://www.tymeshift.com/?ref=internalnote.com), an agent scheduling and forecast tool, and [Klaus](https://www.klausapp.com/?ref=internalnote.com), an AI-powered Quality Management platform. They both fill gaps in the platform that enable bigger customer care teams to get insight in what agents are doing, and make sure the right type and amount of agents are available to handle customer inquiries.
Both purchases are fairly recent (or in the case of Klaus not even fully completed) so there's not much to say about how the tools currently integrate deeply into the Zendesk platform, but given the wealth of historical data available in your own Zendesk, the insights they can pull based on all the thousands of other Zendesk customers, it's only logical that they leverage that data to give insight into agent availability natively, instead of handing that data off to an external party.
And if AI Bots shift agents towards more complex interactions, you want to make sure that, if a customer needs a human, the right person is available to chat, call, or reply to you.
# Conclusion
I have to admit, this was a tough article to write. Where my overview of the [Customer Service Trends](https://internalnote.com/ultimate-trends-2024/) by Ultimate last year went fairly easy due to the trends being practical and more technically focused, Zendesk's trends are a bit more *fluffy* and less grounded in technical reality.
That being said, they do force me to take a bird's eye view of the Zendesk Platform and how it plays in the broader ecosystem of customer care and companies, instead of writing weekly about the nitty-gritty feature releases launched by them.
**If I have to pick one favourite Trend, it's gotta be Trend 4 (AI transparency and decisioning are now the rule, not the exception).**
I'm a big proponent of "control your own data". It's the reason I write on my own website and not just post on LinkedIn. It's the reason why I use Apple products, since they have a very strict stance on privacy, and it's the reason why I like to work with Zendesk or Ultimate.
They both allow companies to build solutions based on their own data and tweak it to their own needs, while integrating with basically everything. But regardless if a customer contacts you via Facebook or email, searches your FAQ, or chats with your bot, you have to be sure you give answers based on the data you provide, and that you can report on all the interactions you have. And your customers have to be sure that the answers they get can be trusted.
### Zendesk Security Checklist
URL: https://internalnote.com/zendesk-security-checklist/
Last updated: 2025-09-08T06:41:23.000Z
One of the big Trends for 2024 is the idea that security no longer is an add-on but should be seamlessly incorporated throughout the customer journey.
To get started the right way, I've written a Zendesk Security Checklist. It's a practical approach to improve the security of your Zendesk instance.
# Email, Web and DNS
Zendesk comes secured out of the box with a fully validated set of email addresses and domains to get started. In my case that would be `support@internalnote.zendesk.com` for the support address, and a `https://internalnote.zendesk.com/hc` domain for my Help Center that's secured with an SSL Certificate.
However it's best practice to replace those with email accounts and a Help Center domain that are based of your own domain name. E.g. `note@internalnote.com` and `https://support.internalnote.com`. Not only do those look better and reflect mine or your brand, they also give you full control later if you ever want to move away from Zendesk, migrate to another instance, or want to control security yourself. Doing this does not remove Zendesk's build-in security measures, it just shifts the branding and domains towards your own.
## Email
When it comes to securing your email communication there's a few basic settings you need to get right. First off, make sure you're using a custom support address (e.g.`note@internalnote.com`) and it's set as the primary email address. This way you're never sending out emails with a `@subdomain.zendesk.com` email address, and your customers only receive emails from your trusted domain.
While you're in the list, it might also be a good idea to [remove](https://support.zendesk.com/hc/en-us/articles/360057705734-Decommissioning-email-support-addresses?ref=internalnote.com) any unwanted email addresses from the list. A cleaner list is easier to manage, and less change of bad configurations.
If you've added your own custom domain to Zendesk, you'll also want to validate the DNS settings for your support email addresses. The `SPF` record will make sure your email that gets send from Zendesk does not go into customers' spam.
The `zendeskverification` record binds your domain uniquely to that specific Zendesk instance, and the two `DKIM` records authenticate your outgoing emails.
| DNS | Type | Value | Comment |
| -------------------- | ---- | ------------------------------------ | ------------------------------------------------------------ |
| @ | TXT | v=spf1 include:mail.zendesk.com -all | You might have more values here |
| zendeskverification. | TXT | Random Text Value | Unique per Zendesk instance and domain. You can add multiple |
You can check if your settings are correct by validating if all items under *Admin Center > Channels > Emai*l are green. If some give an error, use [this article](https://support.zendesk.com/hc/en-us/articles/115014034108-Authenticating-incoming-email-SPF-DKIM-DMARC-?ref=internalnote.com) to fix your issues. Note that you need to do this for each (sub)domain added to your Zendesk instance.
After validating all your email settings, make sure to go all the way to the bottom of that page and enable a Custom Domain for [DKIM](https://support.zendesk.com/hc/en-us/articles/203663326-Digitally-signing-your-email-with-DKIM?ref=internalnote.com) too. This will authenticate your emails as being allowed to be send from Zendesk with your custom domain, and will further prevent them going into spam.
| DNS | Type | Value | Comment |
| --------------------- | ----- | -------------------------------- | ---------------------- |
| zendesk1.\_domainkey. | CNAME | zendesk1.\_domainkey.zendesk.com | Primary DKIM domain. |
| zendesk2.\_domainkey. | CNAME | zendesk2.\_domainkey.zendesk.com | Secondary DKIM domain. |
While you're in the Admin Center, also make sure to enable *Authenticate emails received with SPF, DKIM, and DMARC alignment*. This used to create a lot of suspended tickets, but now that DKIM and DMARC are more and more adopted, it actually helps reducing spam in your instance.
## Help Center
The next item on the list is mapping the default Zendesk domain to a custom subdomain of your own. This is done by adding the following record to your DNS table:
| DNS | Type | Value | Comment |
| -------- | ----- | --------------------- | -------------------- |
| support. | CNAME | subdomain.zendesk.com | Help Center hostname |
Once you've done so go to either:
- *Admin Panel > Account > Branding*
- *Admin Panel > Account > Brands > (choose a brand)* (Enterprise)
And [add your custom domain](https://support.zendesk.com/hc/en-us/articles/203664356-Host-mapping-Changing-the-URL-of-your-Help-Center?ref=internalnote.com#topic%5Fpwv%5Fln2%5Fv3) in the Host mapping field.
Next, go to *Admin Panel > Account > Security > More Settings > SSL* and enable a [Hosted SSL certificate](https://support.zendesk.com/hc/en-us/articles/203664356-Host-mapping-Changing-the-URL-of-your-Help-Center?ref=internalnote.com#topic%5Fpwv%5Fln2%5Fv3). Also make sure to enable HTTP Strict Transport Security to make sure you never fall back to HTTP
💡
Zendesk offers automatic SSL management via LetsEncrypt. You can use your own Certification chain if you want, but if you don't own or manage your own certificate chain, LetsEncrypt is a good way to get SSL enabled without any management of your own. (Or worrying about annual renewals)
## Web Widget
You can[ setup the Web Widget](https://support.zendesk.com/hc/en-us/articles/4500748175258-Installing-the-Web-Widget?ref=internalnote.com#topic%5Fedv%5Fxd5%5F2tb) to be available only on specific domains. This prevents others from copying your widget code and embedding the widget on their own website. Doing this wouldn't give them access to your tickets or data, but would lend credibility to the rest of their website.
# Agent Accounts
The second big block of security measures can be found in the way your end-users and agents sign into Zendesk. By default Zendesk allows anyone to create tickets, and users can create accounts with passwords stored in Zendesk itself.
You can choose to allow other login methods like social or business platforms, or, if you choose to allow Zendesk accounts, you can setup the security levels for passwords and enforce 2FA options to further secure your logins.
## Login Methods
When configuring the way your agents login to Zendesk there's two methods available:
1. You use the native Zendesk option and have agents login with credentials directly within Zendesk
2. You have an Enterprise SSO solution like Microsoft, Google Workspace or Okta enabled in your company and want agents to login that way.
### Enterprise SSO for Agents
The benefit of enabling any kind of SSO solution is that you centralise your agents' authentication flows within your company and control access to all your SAAS tools in once place. This makes on- and off boarding employees easier, you can setup a single account with security policies, and Zendesk, your email tool, your intranet and all other platforms linked will adhere to those policies.
Zendesk has some pretty good [documentation](https://support.zendesk.com/hc/en-us/articles/203663826-SSO-single-sign-on-options-in-Zendesk?ref=internalnote.com) on how to enable this.
> what to do with the Zendesk authentication once an external authentication method like an Enterprise SSO is enabled?"
There's two ways to approach this.
1. If you have external parties using your Zendesk, like e.g. contractors, who aren't in your Enterprise SSO, you can leave the Zendesk accounts enabled. This way Agents can login with their business account, and contractors can access Zendesk via `https://subdomain.zendesk.com/access/normal` to login with a password stored in Zendesk.
Obviously, this will also allow agents to login similarly, potentially circumventing your SSO policies.
2. If you never want someone to use a Zendesk password, you can disable Zendesk authentication. This way you can only login with your SSO account credentials. However, if your SSO goes offline, you're locked out of Zendesk too. Luckily admins can use the SSO Bypass via [https://subdomain.zendesk.com/access/sso\_bypass](https://d3v-verschoren.zendesk.com/access/sso%5Fbypass?ref=internalnote.com) to login with a one-time-password link emailed to them to gain access again and restore the Enterprise SSO options in that scenario.
### Zendesk Accounts
Since last month Zendesk now has a new Recommend setting for their accounts. This is enabled by default for all new customers, but existing customers might have a different option enabled. I recommend using this new setting by default since it checks for breached password matches via [HIBP](https://haveibeenpwned.com/?ref=internalnote.com).
If you go to *Admin Panel > Account > Security > Advanced* there's also a few settings you might want to validate:
- Disable "*Enable admins to set passwords*". This setting allows admins to set passwords and then email them as clear text to customers. Passwords should never be shared and sending a reset password link from the Agents' profile is a better practice.
- *Email notifications* on password change should be enabled so people are aware of unwanted actions.
- Enforce [2FA](https://support.zendesk.com/hc/en-us/articles/4408826974874?ref=internalnote.com) to further secure Accounts. This is not required if you use Enterprise SSO solutions since they often have their own 2FA policies.
## Roles and Permissions
Once you've setup the way agents get access to your instance there'a a few extra audits you can run to make sure agents have the right access and permissions.
### Groups
First off, make sure your agents only belong to the groups they should have access too. Often agents are temporarily added to other groups to cover for colleagues during holidays, or are still in their old department groups.
The easiest way to handle this is to go to the Groups settings in Admin Center > People and do a quick audit of each group.
### Ticket Access
Regardless of your instance having access to custom roles or not, you should check if [Agents](https://support.zendesk.com/hc/en-us/articles/4408886939930-Adding-agents-and-administrators?ref=internalnote.com#topic%5F3zw%5Fyl2%5Fyg) or their [roles](https://support.zendesk.com/hc/en-us/articles/4408882153882?ref=internalnote.com) (enterprise) have access to the right tickets.
You can set an agent's access to tickets as:
1. All tickets
2. Tickets in agent's groups
3. Tickets in agent's org
4. Assigned tickets only
It's best practice to set this to the second option since this way an agents can work as a team on their assigned tickets, without getting access to other departments' work (important in case of HR or Finance scenarios).
### Admin Access
If you're using Zendesk Enterprise you have the benefit of giving [specific user roles](https://support.zendesk.com/hc/en-us/articles/4408882153882?ref=internalnote.com) limited admin access. This way you can allow Team Leads to e.g. manage Views or Macros, while keeping the rest of the Admin Center locked to a specific set of Administrators.
Best practice in this scenario is to create custom roles for Agents, Team Leads (with minimal Admin permissions) and Admins (all permissions). This way you can allow for customisation, while retaining control of your instance and preventing too many people from having full access rights.
I've even seen customers with roles for Guide Admin, Reporting Admin, Agent Workspace Admin, ... and assign these bespoke custom roles to maybe just one or two people in the organisation.
### Agent Roles
Once you're done validating Agent group access, reviews your user roles and maybe created a few new "Light Admins", you need to review all your agents in Admin Center > People > Team members and make sure that only active employees are in that list. You can downgrade anyone else to end-user (or suspend them, see below). This is especially important in the case if light agents since they often get access to a lot of tickets for third line support roles, but are always outside of the Customer Care team and might be overlooked in reporting or audits.
Once your list is cleaned, go over the remaining agents and make sure they all have the right Ticket Access or Custom Role assigned to them.
# Triggers
This one might seem weird but disabling all triggers that alert agents of incoming tickets is a security measure.
Imagine a scenario where a potential scammer creates an urgent message via email to support@yourdomain.zendesk.com.
If they enter, for example, a message that ask you to reset om your password by clicking here, then all your agents will get an email from your instance that they should do so. Some will see this as legit, and will do so. Phishing attempt: successful!
So disable triggers that alert your agents and train them to go to Zendesk regularly.
# End-Users
💡
One topic I'm not handling in this article is the choice between an open Zendesk environment and a closed environment that only allows specific or [authenticated](https://support.zendesk.com/hc/en-us/articles/4408893912986-Permitting-only-users-with-approved-email-addresses-to-submit-tickets?ref=internalnote.com) users to submit tickets.
This choice is not really about security but is more about catering to specific use cases like Financial Institutions or Internal Help Desk, instead of supporting every potential customer like most B2C setups would do.
## Login Methods
Similar to Agents, one of the first things to look at are the login methods for end-users. But where the login for agents handles protecting your Zendesk instance and all the data within, authentication for end-users is important because it's the way to know if you're actually speaking to a specific customer, and to allow customers to find their past conversations.
By default Zendesk offers their own Zendesk Accounts and passwords. Here too I would recommend to pick the [Recommended](https://support.zendesk.com/hc/en-us/articles/4408887573274-Understanding-options-for-end-user-access-and-sign-in?ref=internalnote.com) option cause it protects end-users from falling into the trap of re-using the same (breached) passwords. As for 2FA, the ability for end-users to enable this has [just](https://support.zendesk.com/hc/en-us/articles/6584372830362-Announcing-two-factor-authentication-for-end-users?ref=internalnote.com) been announced!
Since good customer experiences is all about meeting the customer where they are and supporting them within the right context, it's important to also adopt that same process when it comes to authentication. So enabling social logins via Facebook, Google or Microsoft accounts offers the customer flexibility in choosing their channel of preference, while still making sure you get the right information: name, email and validation it's them.
Similarly, if you run your own login system for your webshop, product or platform, it might be worthwhile to look [into linking that](https://support.zendesk.com/hc/en-us/articles/4408885847962?ref=internalnote.com) to Zendesk so you can offer a transparant SSO flow to your customers.
## Messaging
Zendesk Messaging and the Web Widget allow you to authenticate users by passing along a JWT token. This gives agents the benefit they can see they're talking to the actual person, and removes the need for end-users to provide their name and email each time.
You can read all about enabling authentication for Zendesk Messaging in the widget and on your Help Center in the article below.
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging/)
## Data Access
When working with end-users there's a few additional options to check. In *Admin Center > People > End-users* you should make sure that end-users can always change their passwords (in case of a breach elsewhere).
While you're in that section it might also be good to enable the "Validate Phone numbers" option. This makes sure the numbers stored are actual phone numbers and not just some random digits, so if you ever need them to map your customers' WhatsApp profiles or incoming Talk tickets, you know the data is good. (This will only show an error when numbers are edited or added and will not remove any exiting data).
# Data integrity
So far we've handled communication and access to and from your Zendesk instance. This section will dive into the data that's actually in your Zendesk instance.
The first topic might be a controversial one, but I'm a big proponent of setting up some kind of auto-deletion rule for old and archived data because if the data isn't there, you never run the risk of exposing or leaking it. I wrote a full guide on this last week, so you can refer to that one if you want to clear some old data.
[Managing Storage Limits in ZendeskThis article gives insights in how to manage your data usage limits in Zendesk, how to bulk delete specific tickets and how to remove attachments to keep your data usage under the allowed limits and reduce the risk of paying more!Internal NoteThomas Verschoren](https://internalnote.com/storage-limits/)
## Redaction
Next up, make sure to enable the[ Automatic Ticket redaction](https://support.zendesk.com/hc/en-us/articles/203663876-Automatically-redacting-credit-card-numbers-from-tickets-Professional-and-Enterprise-?ref=internalnote.com) to make sure no credit card numbers show up in your system. It's also best to train agents to make use of the [native redaction tools](Agent workspace redaction, train agents - https://support.zendesk.com/hc/en-us/articles/4408846470170) in the Agent Workspace so no sensitive or personal information remains in your tickets while they're being handled.
## Marketplace Apps
Zendesk has a big marketplace with thousands of apps that add features to your Zendesk instance. Similar to how I go through the apps I linked to my personal [Google](https://support.google.com/accounts/answer/13533235?hl=en&ref=internalnote.com) account once a year, it's best to go over your Marketplace apps from time to time too.
Go to *Admin Panel > Apps and Integrations > Zendesk Support Apps* to find a list of active and enabled apps. If any apps are unused or no longer relevant you can [disable or uninstall](https://support.zendesk.com/hc/en-us/articles/4409155972378-Managing-your-installed-apps?ref=internalnote.com) them. Whatever option you choose, they no longer get loaded when looking at tickets, reducing the risk of data getting send to tools you no longer need.
💡
Uninstalling an app removes all settings and you no longer need to pay for it.
Disabling the app keeps its settings, and you might still need to pay for it, depending on its subscription model.
Once you've cleaned your app list you can dive into each app and enable [group and role restrictions](https://support.zendesk.com/hc/en-us/articles/4409155972378-Managing-your-installed-apps?ref=internalnote.com#topic%5Fxyf%5Fwmw%5Fqfb) for each app. These settings allow you to configure the apps so that only a specific set of your agents can access them. For example, an app that shows invoice and payment information for your customers might be useful for your Finance team, but shouldn't be visible to the IT team. Similarly an app that shows employee information is useful for HR, but not for your Customer Care team.
## App Authorizations
Ever wondered with external platforms have access to your Zendesk instance? With the [App Authorizations](https://www.zendesk.com/marketplace/apps/support/1011454/app-authorizations/?queryID=cedea5867e8df134b2b74924e82ae412&ref=internalnote.com) app from Zendesk Partner Sweethawk you get a nice overview of all tools you ever added to your Zendesk, and a one click option to remove their access.
# Zendesk API
## API Access
Zendesk has an extensive and powerful API that can be reached via a few authentication methods, not all of which should be enabled for it to work as expected.
By default Zendesk allows for password access and also has a Token access. If you've enabled 2FA like mentioned in the earlier recommendations Password access will not work anymore. I recommend [disabling it](https://support.zendesk.com/hc/en-us/articles/4408889192858-Managing-access-to-the-Zendesk-API?ref=internalnote.com#topic%5Fzbv%5Fck1%5F2yb) by default and switching to API tokens with a unique Token per integration. This reduces the risk of leaking passwords, and if one gets leaked, you can rotate that one to a new one without affecting the other integrations.
Next, go through your list of API tokens and remove all Tokens that have never been used, or have not been used in the last 3 months. You can always generate new ones if you need them.
If you only have a single token active but know you have more than one integration, I highly recommend adding additional tokens per integration and migrating them each to a unique token.
💡
I really hope in a future update I can also add a recommendation to limit the scope of your tokens to the applicable API paths. But sadly Zendesk' tokens are global and have access to your entire Zendesk instance.
## Webhooks and Targets
💡
Zendesk Targets will be [deprecated](https://support.zendesk.com/hc/en-us/articles/6468124845210-Announcing-the-deprecation-of-URL-targets-and-branded-targets?ref=internalnote.com#:~:text=Zendesk%20is%20discontinuing%20support%20for,targets%20in%20favor%20of%20webhooks.) soon, but even so it's recommended to check both Webhooks and Targets for the time being.
Webhooks are used to send data out of Zendesk to external platforms. They can be added by Admins, or can be installed as part of Apps and Integrations you installed via the Marketplace.
Normally, once you uninstall a Marketplace app, its linked Webhooks will be removed too. But it's best practice to go over this list and check if there are any unknown or unwanted Webhooks in the list. You can open a Webhook and look at the Activity tab to see if they've been recently used, and what data got sent out.
You can use the Actions dropdown top right to disable or delete any unneeded webhooks.
# Admin Center
## Audit Log
If you're using Zendesk Enterprise you get access to the [Audit Log](https://support.zendesk.com/hc/en-us/articles/203663196?ref=internalnote.com). This log gives you insight in any Admin changes made in your instance, as well as any ticket or user deletions that might occur. So if you're ever in a scenario where something weird happened with your data, or you want to know who made a critical change in your instance, that would be the first place to start looking.
## Security Alerts
Zendesk allows you to subscribe to security alerts for your instance by entering a security contact via the admin panel > Account > Security > More Settings.
You can also subscribe to status updates about your instance via [https://status.zendesk.com](https://status.zendesk.com/?ref=internalnote.com).
# Advanced Data Privacy and Protection add-on
We can't wrap up this security checklist without a mention of Zendesk's newest add-on: the Advanced Data Privacy and Protection add-on.
This add-on adds an extra layer of security on top of the existing features:
| Access Log (EAP) | Expands the Audit Log with insights into what each agent looked at in your instance |
| -------------------------------------- | ----------------------------------------------------------------------------------- |
| Advanced Data Retention Policies (EAP) | This allows you to setup more complex ticket deletion rules |
| Advanced Redaction (Announced) | Automatic redaction of more than just credit cards |
| Data Masking (Announced) | Hide specific ticket and user fields for specific sets of agents |
| Advanced Encryption (EAP) | BYOK Encryption |
Since most of these are announced and not generally available, I'll write more about it once I can get my hands on all the features.
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Managing Storage Limits in Zendesk
URL: https://internalnote.com/storage-limits/
Last updated: 2025-07-06T17:52:47.000Z
Zendesk is a SAAS product, meaning you buy a license of the product and they provide you with hosting, storage, software, security and all other elements you expect of a good software application.
When looking at SAAS offerings we can broadly put them in three buckets:
- **Seat based licensing,** where you pay per active user of the tool. Zendesk falls under this bucket with its Agent based licensing. You per pay active agent in your environment
- **Usage based licensing**, where you pay on how much you use the product. Mailchimp is such a tool, where you pay for the amount of recipients in your email list. Sunshine Conversation is another, where you pay per active monthly user.
- **Storage based licensing**, where you pay for the amount of storage used. Think Amazon S3, Google Drive or other solutions.
Most SAAS tools offer a combination of these. For example, in Google Workspace, each user license comes with a set amount of storage, but you can buy additional storage if you need more.
Zendesk has always been Seat Based and didn't really have limits imposed. A ten agent environment could serve 100, 1000 or a million end-users and your bill would be the same regardless.
But this has now changed. [Announced](https://support.zendesk.com/hc/en-us/articles/6340068721434-Announcing-the-Storage-usage-dashboard-in-Admin-Center-and-increased-storage-limits?ref=internalnote.com) at the end of November, Zendesk will now enforce a more strict storage usage on its platform, and will bill customers with an overage upon their next renewal.
# Your available storage
There's two types of storage measured:
- **File Storage** is the most straightforward. It's the attachments linked to your tickets, e.g. PDF files, images, signatures,....
If you receive an email with three 5MB attachments, those would count as 15MB of File Storage used.
- **Data Storage** is all other data in your Zendesk: the ticket object, comments, metadata (ticket fields, attachment metadata, tags,..), users, custom objects (2kb per object). On average, a ticket with three comments counts as about 5kb of data storage if I compare the usage across multiple instances.
Zendesk calculates your available storage as follows:
> \[BASE STORAGE for your plan\] + (\[AMOUNT of agent\] \* \[ADDITIONAL storage\])
So if you have Suite Professional with 10 agents, you would have 11GB of Data Storage, and 60GB of File Storage available.
## Overage
If you go over your allotted usage, Zendesk will show a notice in the Admin Center. You can then either reduce your usage by removing data or by buy additional capacity.
This is done by buying Storage Units (500 MB Data + 25 GB File Storage) and it's bought on a per-unit, per-month basis. But they don't come cheap!
# Viewing Storage Limits
Zendesk has added a new [Dashboard](https://support.zendesk.com/hc/en-us/articles/4408835043994?ref=internalnote.com) to the Admin Center to view your Storage usage. It gives you a high level overview of your current usage, and a timeline on how the data usage evolved over time.
The screenshot below shows the basic overview of storage usage in an instance that went over its File Storage:
Instance with 50.000 tickets
And if you drill down into the details, you can see a curve on how the data grew over time.
Instance with 50.000 tickets
💡
Note that the steep rise in usage starting in May is caused by the lack of indexed older data. Zendesk only started indexing storage usage around mid May 2023.
Sadly, the dashboard doesn't give you any more details. You can't see which tickets cause the usage, or where the big attachments are.
## How to find your overusage
Since we can't pinpoint the exact perpetrator, we can only fix this issue in bulk and work with averages. This is not an exact science, but since the amount of tickets is so big, we can assume the average per ticket is correct.
Take the example above. We know we have 50.000 tickets, and our File Usage is 60GB. This means that, on average, we have 1.2MB of File Storage per ticket.
If we want to reduce our File usage to get under the storage limit, we need to delete at least 10GB, so let's delete 15GB to be safe. That would equal deleting 12.500 tickets in our instance. This also frees up 2.5GB of Data usage at the same time.
💡
The easiest way to find out the amount of tickets in your instance is to look at the most recent ticket ID in your instance. Or, if you ever deleted tickets, you can go to Explore and open the Default Support dashboard and set the Time to show **All History*. The **Created Tickets* value excludes deleted tickets.
You can calculate your own via [this](https://demo.internalnote.com/storagelimits?ref=internalnote.com) calculator:
# Deleting Data
‼️
This article shows you ways to delete data in your Zendesk instance. Always double check your actions and make sure you don't delete something you don't want to.This article offers guidance, but I'm not responsible for any accidental data loss in your instance!
There are a couple of approaches to delete data. You can either do a deletion of tickets or delete only attachments in tickets.
If we delete tickets, that will also include any attachments linked to these tickets so we win storage in both tiers.
The downside is that you will lose reporting data on those tickets, so it might be better to delete a specific set of tickets (e.g. all except those Finance), or only delete attachments.
## Bulk Delete Tickets
### Via the Admin Center
To delete those 12.500 tickets, we can look up ticket #12500 in our Zendesk environment and note down its creation data. If we delete all tickets older than that date, we'll free up the necessary storage! For our example, let's assume ticket #12500 was created on December 13th 2021, and is thus approx 2 years, or 730 days old.
Deleting the required tickets can be done via the new [Timed Ticket Deletion](https://support.zendesk.com/hc/en-us/articles/6062884435866?ref=internalnote.com) (EAP). Go to the Admin Panel > Objects and Rules > Settings and enable Timed Ticket Deletion, and set the value to the required amount of days.
This will start the automatic deletion of all tickets that are older in your instance, and will run continuously.
### Via the API
Or, if you prefer the API route, we can use the [Bulk Delete Tickets](https://developer.zendesk.com/api-reference/ticketing/tickets/tickets/?ref=internalnote.com#bulk-delete-tickets) endpoint, and delete our tickets per hundred via the following command:
```bash
curl https://{subdomain}.zendesk.com/api/v2/tickets/destroy_many.json?ids=12500,12449,12448 -v -u {email_address}:{password} -X DELETE
```
⚠️
Zendesk has strict [API limits](https://developer.zendesk.com/api-reference/introduction/rate-limits/?ref=internalnote.com) so bulk deleting thousands of tickets has to be done over a long period of time. Don't go deleting 12500 tickets all at once cause this will impact your other integrations too.
## A more granular approach
The above example works fine if you want to just get rid of the data, but it doesn't work if you need to retain some tickets for regulatory reasons, or if you want your reporting data to remain available.
One solution to get a more specific set of tickets is to do a search that returns all tickets that are, e.g., not in a specific group and delete those tickets.
### Via the Admin Center
💡
Zendesk has announced multiple Deletion schedules as part of the Advanced Data Privacy and Security add-on, which will allow you to delete tickets on a more granular basis, e.g. deleting all tickets in the Support group after three years, and those in Finance after seven.
For now, this feature is not yet available, not even in an EAP. So when this becomes available I will update this article.
### Via the Zendesk API
In the API call below, we search for all tickets that are not assigned to Finance group with ID `1234567890` and are older than 13th December 2021, since that's the date ticket #12500 in our example was created:
```bash
curl 'https://{subdomain}.zendesk.com/api/v2/search.json?query=type%3Aticket%20-group%3A1234567890%20created%3C2021-12-13' -v -u {email_address}:{password} -X
```
This returns a list of tickets from which you can then extract the IDs. We can then use the Bulk Delete endpoint to remove these tickets.
### Via a Marketplace App
#### GDPR Search and Destroy
However, instead of writing the code yourself, we can also use an app like [GDPR Search and Destroy](https://www.zendesk.com/marketplace/apps/support/206749/gdpr-search--destroy/?ref=internalnote.com) by Zendesk Partner [Sparkly](https://sparkly.dev/?ref=internalnote.com) to do this same search but without the additional coding work.
Their app allows you to Bulk Delete tickets based on search queries, and run these deletions once, or as a continuous script. It's the perfect app to fix these Data usage charges, or to keep compliant with GDPR and similar regulations.
After installing the app, you can add a Task. Choose Ticket Search as the type and search for tickets with the query `-group:360001292600 created<2021-12-13` . This returns a list of matching tickets excluding the group Finance, and created before Dec 13th. Once you validate your results, you can *Process* these tickets to clear your storage.
#### Zendesk Attachment Storage Offload
If you don't want to delete the data, but rather offload the data to somewhere else so you can still reference it, you can look at solutions like [Attachment Storage Offload](https://www.zendesk.com/in/marketplace/apps/support/1002546/attachment-storage-offload/?ref=internalnote.com) which takes attachments and moves them to a cloud storage managed by them or yourself.
[Zendesk Attachment Storage OffloadAttachment Storage Offload for Zendesk is a simple app to keep Zendesk attachments accessible while reducing storage costs.EH7P](https://zd-external-attachment-storage.eh7p.com/about?ref=internalnote.com)
# What about attachments?
Bulk deleting tickets will fix your storage issue, but leaves a big gap in your historical reporting. And more often than not, the issue is not the Data Storage (tickets), but the File Storage (attachments).
So a smarter way to approach this issue would be to only delete attachments of those 12.500 tickets. That way both storage tiers are in the green again, while retaining historical data.
The Zendesk API does not have an easy *get all attachments* endpoint, and neither is there a *delete attachments* option. We do however have a [Redaction](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket-attachments/?ref=internalnote.com#redact-comment-attachment) endpoint which allows us to remove the File Storage element of an attachment, and replace it with a 5kb reference file for the original attachment. When you redact attachments, what's left is your ticket with a reference of the attachments, and all other data intact.
## Via API
Let's assume we did the work and got a list of all IDs of the tickets for which we want to remove attachments. This can be all tickets up to ticket 12500, or a more granular list that excludes some tickets in specific groups.
For **each** of these tickets we can call:
```bash
curl https://{subdomain}.zendesk.com/api/v2/tickets/{ticket_id}/comments.json?include_inline_images=true -H "Content-Type: application/json" -v -u {email_address}:{password}
```
This returns the Comments of that ticket, and an array of attachments:
```json
{
"comments":[
{
"id":16065263235986,
"type":"Comment",
"attachments":[
{
"id":16065246996882,
"file_name":"Screenshot 2023-07-04 at 21.36.50.png",
...
},
{
"id":16065239646098,
"file_name":"Screenshot_2023-07-04_at_21.36.50_thumb.png",
...
}
],
...
},
...
],
"next_page":null
,"previous_page":null,
"count":2
}
```
For **each** of the attachments we can then run the following command to redact them:
```bash
curl --request PUT 'https://{subdomain}.zendesk.com/api/v2/tickets/{ticket_id}/comments/{comment_id}/attachments/{attachment_id}/redact' -v -u {email_address}:{password}
```
But as you can probably guess, this gets big fast. If we assume a ticket has three comments, and each comment has 2 attachments, then we need to make a lot of API calls.
> 12500 tickets x 3 comments x 2 attachments = 75000 API calls!
## Via Marketplace Apps
Luckily this is another scenario where we can resort to Marketplace apps like [GDPR Search and Destroy](https://www.zendesk.com/marketplace/apps/support/206749/gdpr-search--destroy/?ref=internalnote.com) to run this redaction without worrying about coding this loop for us!
The app does require you to upload a list of IDs, so we still need to export all comments for the 12.500 tickets and create a matching CSV file first.
We could do this via API, or use the [Exporter](https://www.zendesk.com/marketplace/apps/support/89495/exporter/?ref=internalnote.com) app (Enterprise license or trial required). This application has a Comments export feature we can use to generate a list of comments and their attachments. When exporting, select CSV (Semi-Colon) as the export type, and make sure you select the `ID`, `Ticket ID` and `Attachments URLs` fields.
Once we get our export of attachments, we can upload that file in the GDPR app and run the deletion. Mind, this deletion runs client-side in your browser, so depending on the amount of data, can take a while!
# Conclusion
So there we have it, a couple of approaches to handle ticket deletion in Zendesk via native features or by making use of marketplace apps.
With regards to the Data Policies I can understand a company changing its policy on how they handle data usage. Data is expensive, and most of the archived tickets in Zendesk instances are never looked at, so that's basically storage used for data that's never used.
But regardless of usage, starting to charge customers for data without offering proper tools to handle these changes is not that great. The bulk deletion options in Zendesk are either too broad or locked behind a paid add-on.
I really hope we'll see a more nuanced data-management feature in the future where we can delete tickets and attachments easily to keep us within our data limits. But until that day comes (if ever) I'm glad we've got third party options or API solutions to handle this issue when it arises!
💡
This article mentions [Sparkly](https://apps.sparkly.dev/?ref=internalnote.com) and the [GDPR Search and Destroy](https://www.zendesk.com/marketplace/apps/support/206749/gdpr-search--destroy/?ref=internalnote.com) app as a good way to resolve your storage usage. Even though I use their apps on a daily basis at my day job, this article is not sponsored by them in any way, and I show their app because I know it's a good tool to fix the issue.
### Zendesk Roundup for January 2024
URL: https://internalnote.com/roundup-2024-01/
Last updated: 2024-01-02T07:21:38.000Z
Happy New Year and welcome to 2024 🎆🎇!
I'm starting the year fresh with a new roundup of the last releases of 2023, clearing the slate for a new year of exciting Zendesk announcements and product updates!
2023 clearly was the year of **Zendesk AI** with major improvements in the Zendesk Bot and Agent Workspace enabling automation, insights, and better self-service across the board.
Last year was also the year of the new Agent Home and Omnichannel Routing, enabling a brand new way to interact with your assigned tickets.
Custom Objects and Layout Builder offered ways to customise Zendesk to fit your business needs, and the focus on privacy and security resulted in more granular admin permissions and a brand new automated ticket deletion feature and the new Privacy and Security Add-On.
So, what will 2024 bring? I'd love to see some kind of co-pilot capability for agents, offering agents a similar self-service experience as end-users get, surfacing internal processes, information, and context to agents. From a developer standpoint, I hope they bring more API endpoints to Zendesk for managing Answers, Bots, and Explore dashboards via API, enabling bulk actions on Custom Objects and a more granular permission approach for API tokens, with options for read-only, or scoped to a specific set of objects or actions.
But the year has only just started, and [Zendesk Relate](https://www.zendeskrelate.com/event/819e790c-37ac-4207-bfef-ab965999c0ef/summary?utm%5Fcampaign=2024%5FRelate&utm%5Fmedium=crosspromo&utm%5Fsource=LuminariesHub) is still 4 months away, so let's dive into last year's final announcements and get ready for 2024!
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 🎉 New Releases
## 🤖 AI Powered Conversational Experiences
You can now [mention](https://support.zendesk.com/hc/en-us/articles/6297959490714-Announcing-Using-mentions-in-messaging-conversations?ref=internalnote.com) colleagues in messaging conversations. They'll be added as followers on the ticket and updates will appear in the Agent Home, as well as via notifications in the Agent Workspace.
## 👨🏻💻 Agent Workspace
The new [Agent Home](https://internalnote.com/tag/agent-home/) is now available for all Zendesk customers. It replaces the old dashboard view with a more actionable landing page for your agents. It shows agents all tickets assigned to them, including active conversations, gives quick links to followed or cc'd tickets, and provides a list of updates on tickets they worked on, or are mentioned in.
If you use Omnichannel Routing, or are an agent that used to work from a *My Tickets* view, I highly recommend setting Agent Home as your, well, Home Page, from now on!
## 🔎 Help Center and Self Service
It's now possible to pin both [internal](https://internalnote.com/zendesk-roundup-for-july-2023/) articles from your Help Center and [external](https://support.zendesk.com/hc/en-us/articles/6496294467738?ref=internalnote.com) articles indexed via Federated Search in the Knowledge Panel right next to tickets. Useful to point other agents to resources needed to resolve a specific ticket.
And if you use Content Blocks, they now support [tables](https://support.zendesk.com/hc/en-us/articles/6485855035674?ref=internalnote.com).
### Redesigning the editor experience.
In a series of articles, Zendesk announced a redesign of the article editor experience.
In a [first phase](https://support.zendesk.com/hc/en-us/articles/6486311589914?ref=internalnote.com), the editor will move publishing actions to the footer of the editor, similar to how Ticket Status currently works. This new footer will allow you to publish, draft, set review status, and preview the article.
This removes all publishing options from the sidebar, making the sidebar focused on organising your article in sections, setting permissions, and adding labels or promoted statuses. The sidebar is now only needed if you want to edit these settings and can be collapsed if you only want to focus on the content.
Later, the sidebar will be updated by moving all these options into different sections, making it more convenient and faster to update a specific setting of an article. And later in 2024, the articles themselves will also be expanded with additional HTML elements:
```html
abbr, address, cite, code, dd, del, dfn, div, dl, dt, em, figcaption, figure, hr, iframe, ins, kbd, pre, samp, small, source, span, sub, sup, track, var, video
```
## 🧱 Open and Flexible Platform
### Notion for Zendesk
I'm a big [Slite](https://refer.slite.com/cn8zn9oi7mha?ref=internalnote.com) fan myself, but for those of you who use Notion, there's a neat new integration available on the [Marketplace](https://www.zendesk.com/marketplace/apps/support/967901/notion/?queryID=e3b73edab1410e04ebee8d995051bceb&ref=internalnote.com).
## 🔐 Security and Privacy
Announced in October and finally available are the [advanced data retention policies](https://support.zendesk.com/hc/en-us/articles/6388095686682?ref=internalnote.com) which allow admins who purchased the Advanced Security Add-On to set up complex deletion policies for specific ticket types to comply with local privacy regulations, or to lower their Zendesk storage limits.
# 💡Insights
## Hybrid approach to Customer Support
Ultimate invited me to speak at their January 10th event about customer care, automation, and self-service. If you're interested, you can RSVP via [this link](https://www.ultimate.ai/webinars/scale-smarter-with-generative-ai-conversation-design?utm%5Fsource=internalnote).
# 🎥 Videos
Interesting end-of-year wrap-up interview with Zendesk CEO Tom Eggemeier.
> AI is a marathon, not a sprint.
You can watch it below on YouTube, or listen to the [Podcast](https://www.zendesk.com/blog/tom-eggemeier-podcast/?ref=internalnote.com).
# ⚠ Major Changes
Over a year ago, Zendesk released the new [Webhooks](https://internalnote.com/tag/webhooks/), which will replace the old URL Target functionality. They include more connection and security options, provide a better admin experience, and offer insight into their executions via a dashboard right within the Admin Panel.
Now, Zendesk has [announced](https://support.zendesk.com/hc/en-us/articles/6468124845210?ref=internalnote.com) the deprecation of these old URL Targets by October 28th, 2024, so take a look at your old Targets and move them over to Webhooks before it's too late!
# 📝 Articles this month
This month was a short month due to the Holidays:
- [Preview of the new Generative AI for Voice](https://internalnote.com/preview-of-the-new-generative-ai-for-voice/)
- [Dynamic Conversation Experiences in the Zendesk Bot](https://internalnote.com/dynamic-conversation-experiences/)
- [End of Year](https://internalnote.com/end-of-year)
# And Finally...
If you use liquid markup in your triggers and macros, you can use this bit of code to format your dates to a European show `DD-MM-YY` format.
```
{{ ticket.ticket_field_YourDateFieldID | date: "%e/%m/%Y" }}.
```
Or, if you work in Europe but want to convert the format to match your American customers;
```
{{ ticket.ticket_field_YourDateFieldID | date: "%m-%e-%Y" }}.
```
### End of Year
URL: https://internalnote.com/end-of-year/
Last updated: 2025-09-08T06:42:30.000Z
Welcome to my last newsletter of the year. This one will not contain any Zendesk news or updates, but will be a bit more of a meta, personal end-of-year update.
I'm writing this note from a couch with our Christmas tree in sight. End of year is about family, friends and tying up loose ends, but it's also about looking forward towards next year and how last year went.
The Internal Note project was started during the holiday break at the end of 2022 with three main motivations:
1. Keeping up to date with Zendesk releases is a big part of my 9 to 5 job, and this blog gives me a good way to test and validate the new features while also having a good logbook of my experiences with them.
2. Train different muscles was also a big part of the reason I wanted to blog. At work I scope Zendesk setups, offer consultancy for customers and develop integrations. But technical writing and maintaining a newsletter that forces me to write \~3k words a week is a whole new skill that I want to get better at.
3. I wanted to see if I could build a project that was self-sustaining and gets an audience.
My goal was to write an article a week and reach subscribers through organic growth. I'm not a fan of targeted ads via LinkedIn or Google, so I wanted to run this without that kind of advertising. Neither did I want to inject myself into active conversations on the Zendesk Community and link to this website in any relevant thread. The end result should be self-sustaining blog that exists because people like it and find it because it's useful.
The past year I published 57 newsletters while growing the audience for this blog from 1 to almost 500 subscribers, so for each of you who subscribed: a big thanks!
# My process
A few of you wrote in to ask how I managed to keep up with Zendesk's seemingly continuous stream of releases this last year. So, I thought it would be fun to use this final newsletter to give some insights in how I work, and what I plan to do different in 2024.
The best way to stay up to speed with anything Zendesk releases is to subscribe to the [Updates](https://support.zendesk.com/hc/en-us/sections/4405298833818-Announcements?ref=internalnote.com) category of the Zendesk Support website. This category contains all official releases, and includes a weekly [Releases Notes](https://support.zendesk.com/hc/en-us/sections/4405298847002?ref=internalnote.com) section. You can easily subscribe to them via the *Follow* button, or, just subscribe to this website to make your life easier 😇
However, that's just the tip of the iceberg since every release is accompanied by its own set of support documents. Often articles and features are added to Zendesk without an official announcement.
For example, Zendesk placeholders for Lookup Fields now support `{{ticket.ticket_field_<*field ID number*>.id}}`to capture the ID instead of the label, a feature was silently added and I only noticed because I follow updates to support.zendesk.com.
At the start of 2023 I subscribed to almost all sections, which resulted in hundreds of emails a week with updated articles. This was not scalable or maintainable, so halfway through the year I decided to approach this problem differently.
To make sure I got to see **every** change in Zendesk, I built a custom RSS feed to generate a list of updated or published articles, which I follow inside of an RSS reader.
It's powered by the [Help Center API](https://developer.zendesk.com/api-reference/help%5Fcenter/help-center-api/articles/?ref=internalnote.com) and refreshes on the hour. That way, every change gets logged somewhere, making it easy to stay up to date.
Another source of good information is the quarterly *What's new* presentation. They serve as are a reminder of everything announced in the past few months. Those are basically me screenshotting at 24fps and noting down the announcements to compare to what was already released And see what I missed.
I also subscribe to the [Upcoming Beta's ](https://support.zendesk.com/hc/en-us/articles/4408829663642-Current-and-upcoming-Zendesk-betas?ref=internalnote.com)page to request every EAP as soon as it becomes available. But getting the EAPs' quickly has been a bit of a hit and miss 😅
# What's next for 2024
> If it ain't broke, don't fix it
One thing I'm changing is the above mentioned RSS feed. Although it works, an RSS reader is not the best way to manage a large queue of data. I often move articles from my feed reader to a *read it later* list, just so I can remember to write about the topic later.
Similarly, you don't want to count the amount of times I accidentally marked *all as read* and had to scroll through pages of updates in order to check if I saw everything.
## Content Management
Last summer I read [Building a second brain](https://www.goodreads.com/book/show/59616977?ref=internalnote.com), a book about storing every snippet of info you need in one place to make it easily retrievable.
So with RSS feeds being too loose a format (read/unread is too limited), it was time to find another place to collect all Zendesk info and documentation.
These last few weeks I decided to start using [Slite](https://refer.slite.com/cn8zn9oi7mha?ref=internalnote.com) to manage my growing collection of Zendesk information. It's a knowledge management tool that works based on documents and articles, and offers powerful AI-powered search and summary tool, while also offering similar knowledge management features as Zendesk with verification, review and flagging features for articles.
[Slite - AI-powered knowledge baseDiscover Slite’s AI powered knowledge base for quick access to trusted company info. Simply ask Slite and get the answers you need. Start for free.Logo Bamboo HR](https://refer.slite.com/cn8zn9oi7mha?ref=internalnote.com)
### Unified Inbox
I updated my RSS feed script to push all Zendesk updates to Slite as unverified articles. This allows me to go over a view with all new articles and mark them as verified once I processed them. I can defer verification to a later date, or quickly copy over elements of documents into draft articles for the website.
Conveniently, updated articles in Zendesk will turn verified documents into unverified again, so detecting changes in documentation is quite easy now!
### Retrieving data
Since I want to use Slite as a reference for all my Zendesk related data I also imported all Internal Notes and the entire developer.zendesk.com website into Slite too.
So questions like "*What's Agent Home*", "*How to add lookup fields via API?*" or "*What did Zendesk release in October 2023*" are now very easy to answer.
I really hope that this new tool will make writing the roundups a lot more efficient this next year.
**But I promise: this blog will contain no AI written content.** 🤞
## Newsletter Structure
When I look at the newsletter, the content can be split into roundups, insights, previews and tutorials. Since a month has four weeks, and these are four kinds of content, you can expect one of each every month. With at least one of these exclusive to [Internal Note Plus](https://internalnote.com/plus) subscribers.
Thanks to, or due to, Zendesk efforts on Zendesk AI and the Zendesk Bot, the second half of 2023 was a little bit Chatbot heavy. For 2024 I want to have a better balance across all aspects of Zendesk.
- [Agent Efficiency](https://internalnote.com/tag/agent-efficiency/): Automations, Agent Workspace, Custom Objects,...
- [Bots and Messaging](https://internalnote.com/tag/bots-and-messaging/): Zendesk Bot, Autoreply, Sunshine Conversations,...
- [Knowledge Base](https://internalnote.com/tag/knowledge-base/): Self Service, Guide themes, Federated Search,...
- [Routing and Intelligence](https://internalnote.com/tag/routing-and-intelligence/): Omnichannel Routing, Zendesk AI, Custom statuses, SLA,...
## Zendesk Relate
In April Zendesk will hold their annual [Relate](https://www.zendeskrelate.com/?ref=internalnote.com) event, a big conference where they will hold product announcements, workshops and round tables with customers and experts alike.
My employer, [Premium Plus](https://premiumplus.io/?ref=internalnote.com), allowed a colleague of mine and me to join the conference in person this year, so that's already going to be a highlight of the year for me! I'll certainly write a same day round-up off all the announcements of the event, but will also try to give some impressions of the event itself.
And if I have to give myself one goal for the event it's to:
Write an article with answers, instead of an article with a lot of questions about the announcements.
# Wrap up
So, there we have it, a roundup of what I did in 2023, and a view of what's planned for 2024.
If you've made it this far in the article, can I ask you one favour?
🎁
Share this blog with at least one colleague or friend who's interested in Zendesk and ask them to subscribe as either a free or [Plus](https://internalnote.com/plus) member.
Enjoy the holidays and see you in 2024! 🎄
### Preview of the new Generative AI for Zendesk Voice
URL: https://internalnote.com/preview-of-the-new-generative-ai-for-voice/
Last updated: 2024-08-19T20:33:11.000Z
One of the latest releases in a line of Generative AI releases for Zendesk is the new [integration for Voice](https://support.zendesk.com/hc/en-us/articles/6115911331226-Zendesk-Generative-AI-EAP-capabilities-overview?ref=internalnote.com#topic%5Fw1n%5F32x%5Fvyb).
Phone channels are the oldest channels in a customer care team's tool belt, but it's often also the least efficient. An agent can only take 1 phone call at a time, which makes it difficult to scale. Historically, there hasn't really been a way to offer decent self-service or deflection, other than losing the customer in a labyrint of IVR choices, and if a call needs escalation of follow-up, it's not efficient to know what's already been discussed with the customer.
If an agent wants to escalate a conversation to another team or a team lead, they need to take notes on the conversation, which makes the wrap-up time for a call sometimes as long as the actual conversation itself. And getting the nuance of a conversation written down means the agent either takes notes during the call, which means less attention for the customer, or risks missing elements from the call.
# Generative AI for Voice
This is where the new AI features for Voice come in. By enabling this feature, Zendesk will automatically transcribe the entire conversation and add an additional summary at the end of the call.
Whenever a call wraps up and needs to be escalated, or if you want to revisit the ticket to check how a call's been handled, to check for similar issues or what have you, agents can now read the summary, instead of relistening to a minutes long recording of the call.
Even better, you can now delete the recordings (faster) to comply with local privacy laws, while still retaining the ticket information and conversation in a written way.
# Setup
Like most things Zendesk AI, enabling Generative AI for Voice is a matter of checking a box.
🤔
Am I the only one who is getting confused with how to enable all the AI features? Some are in the Bot options, some are their own Admin Section like Generative AI, and others are hidden under intelligent triage.
The settings are hidden in *Admin Panel > Channels > Talk > Settings* and consistof two options:
1. Enable transcribe and summarise
2. Disable transcripts
There's no other settings available, so it's a global on/off for all lines available in your instance.
[Announcing generative AI-powered call summarization for Zendesk Talk (EAP)Announced on Rollout starts Rollout ends November 29, 2023 November 29, 2023 December 20, 2023 We are excited to announce the launch of the EAP for call summarization using generative AI. Thi…Zendesk helpNova Dawn](https://support.zendesk.com/hc/en-us/articles/6330415782170-Announcing-generative-AI-powered-call-summarization-for-Zendesk-Talk-EAP-?ref=internalnote.com)
## Fine Tuning
For testing, I recommend enabling both the transcript and the summary initially. This way, you can validate the feature (it's currently an EAP so it might have some quirks).
Later, you can disable the full transcripts if you want just the summary, but don't want to store the entire transcript for efficiency reasons.
# Testing
No better way to test a Voice AI feature than to give it the world's most annoying Customer Care interaction!
So, how did Zendesk AI do?
## Transcribing the conversation
First, let's look at the call transcript:
#### Customer Transcript
00:00 Hello. Do you wish for technical assistance?
00:16 Oh God, I hate these.
00:18 Sorry?
00:20 What?
00:22 How dare you speak to me like that?
00:26 Sorry. Can I have technical assistance, please?
00:30 Cutting you through.
00:32 Coward.
00:36 Hiya.
00:37 Hello.
00:38 Can I have assistance?
00:41 What?
00:42 Can I have assistance?
00:45 I've got one of your laptops. It won't get past the loading screen.
00:50 Can you put it on again?
00:53 Yes. Look, is there someone else I can talk to?
00:55 There is no less valuable.
00:57 Oh, right.
00:58 With a model device.
01:00 What model do I have?
01:02 Yes.
01:03 It's a Zubion, I think.
01:05 Oh, Zubion, you're too serious.
01:07 Do you not think of me as a Zubion?
01:09 Can you put it on again?
01:11 I didn't get that.
01:13 Can you press the little wheel?
01:17 No, that would be worse somehow.
01:19 Can?
01:21 Can?
01:22 Yes. Can you press?
01:24 Delete?
01:25 Delete. Can I press delete?
01:27 Shh.
01:28 Shh.
01:29 Shh.
01:30 Shh.
01:31 Shh.
01:32 Shh.
01:33 Shh.
01:34 Shh.
01:35 Shh.
01:36 Shh.
01:37 While, while.
01:38 Yes.
01:39 Shh.
01:40 Shh.
01:41 Hello?
01:42 Hello?
As you can see, we get a pretty good summary of the conversation, but reading this after the call to assist the agent with this ticket is almost impossible due to the nature of phone calls. Half sentences, repeated words, some of the transcription (obviously) went wrong,...
## Summarising the call
The main feature of this new AI option is a summary of the call. Once my call wrapped up, the transcript above appeared after \~10 seconds, with the summary appearing a few moments later.
And well, you can judge the results yourselves, but I find this pretty accurate!
#### ****Call summary**
- The customer is requesting technical assistance for a laptop that won't get past the loading screen.
- They mention having a Zubion laptop model.
- The customer is frustrated and requests to speak to someone else.
- There is a mention of pressing the Delete key, but it is unclear if the customer was able to do so.
- The conversation ends abruptly and it's unclear if the issue was resolved.
# What's next?
If you use Zendesk Talk and are thinking about buying the Advanced AI add-on, this feature seems like a no-brainer. The amount of time that can be saved this way for agents is pretty obvious, let alone the gains on reporting and insights.
That being said, one major feature is missing currently for me, and that is the lack of integration with the other Zendesk AI and Intelligent triage features.
If you have a summary of the call, why doesn't this also feed the Summary option in the Intelligence Panel? And why can't we see the Intent or Sentiment of the call? All the data is obviously there. And if I copy the conversation transcript into a new ticket, I get the results on the right.
But, taking into account that feature is in an EAP, I'm sure this will all be nicely integrated once the feature work wraps up and the internal data gets hooked up.
### Dynamic Conversation Experiences in the Zendesk Bot
URL: https://internalnote.com/dynamic-conversation-experiences/
Last updated: 2025-09-08T06:41:39.000Z
Last year I wrote an article about using a combination of *Ask for Details* and *API Calls* in Zendesk Bot Builder to built a little Movie Bot based on the Open IMDB API.
[Flow Builder - Ask for detailsThe new Ask For Details option in Flow Builder allows you to pull in contextual information via API into your Zendesk Chat Bot.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-ask-for-details/)
That bot turned out great, but it was always limited by the features available in Bot Builder. Or more specifically, the lack of support for arrays and variables inside of the API calls made it impossible to chain API commands. So once a user picked a movie, I couldn‘t do anything based on that choice with another API call.
Now, almost a year later, Zendesk announced [Dynamic Conversation Experiences](https://support.zendesk.com/hc/en-us/articles/6405324545050-Announcing-dynamic-conversation-experience-with-Zendesk-bots?ref=internalnote.com) for the bot, which greatly expand the features available in API calls with the addition of three new options:
- Support for Arrays in the API call so that we can more easily iterate over a bunch of results
- The Carousel step can now handle variables and arrays
- We can use variables inside of API calls
So let's dive in and build a better version of the Movie Bot!
# What we're building
We're going to build a Zendesk Bot that will first ask the customer for a movie title.
We’ll then use IMDB to retrieve a list of movies that match. When the user selects their movie, we'll show the genre, director and a short summary of the movie, and ask if this resolved their question.
Under all this we make use of a couple of Zendesk Bot [step types](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-step-types?ref=internalnote.com):
- We use *Ask for Details* to ask for a movie title and store it in a custom field
- We use the *API call* step to make two API calls to search and get details from IMDB
- We use a *Carousel* step to show a list of search results, and use variables to set images, buttons and content
- We use *Send Message* with its image and button support to show the movie details and guide the customer
- We use *Ask if question resolved* to wrap up the Answer flow.
# Preparing your setup
We assume you already have a working Zendesk Bot setup for this tutorial.
## Ask for Details
This Answer Flow requires one custom ticket field to get started. So create a text field called "Movie", that's editable for end-users. We'll use this field to store the search query of the customer. E.g. `Alien`.
Bot Builder has a nice visual editor, so we don't need to store or remember the Field ID this time!
## The API Calls
This bot will make use of [Open IMDB](https://www.omdbapi.com/?ref=internalnote.com) to search and query IMDB and retrieve movie info. You can retrieve a free API key from their website.
We're going to make two API calls to first search for a movie based on the input in our custom field, and then get the details for that movie.
### Search for a movie
A `GET` to `https://www.omdbapi.com/?s=star%20wars&apikey=abcd1234` with `s=star%20wars` as your search query and `abcd1234` as your OMDB API Key.
The API call returns an array of results. With the new Dynamic Conversation release this array can now be stored as a variable, and we can show a list of all results!
```json
{
"Search": [
{
"Title": "Star Wars: Episode IV - A New Hope",
"Year": "1977",
"imdbID": "tt0076759",
"Poster": "https://m.media-amazon.com/images/M/MV5BOTA5NjhiOTAtZWM0ZC00MWNhLThiMzEtZDFkOTk2OTU1ZDJkXkEyXkFqcGdeQXVyMTA4NDI1NTQx._V1_SX300.jpg"
},
{
"Title": "Star Wars: Episode V - The Empire Strikes Back",
"Year": "1980",
"imdbID": "tt0080684",
"Poster": "https://m.media-amazon.com/images/M/MV5BYmU1NDRjNDgtMzhiMi00NjZmLTg5NGItZDNiZjU5NTU4OTE0XkEyXkFqcGdeQXVyNzkwMjQ5NzM@._V1_SX300.jpg"
},
...
]
}
```
### Get movie details
A `GET` to `https://www.omdbapi.com/?i=tt123456&apikey=abcd1234` with `tt123456` being the `IMDB ID` of the movie we found in the first step, and `abcd1234` your OMDB API Key.
Similar, we can now use the `imdbID` of whatever the user choose in the carousel from the search results in our API call, something new that's only become possible since this new release.
The API call returns movie details:
```json
{
"Title": "Star Wars: Episode IV - A New Hope",
"Genre": "Action, Adventure, Fantasy",
"Director": "George Lucas",
"Plot": "Luke Skywalker joins forces with a Jedi Knight, a cocky pilot, a Wookiee and two droids to save the galaxy from the Empire's world-destroying battle station, while also attempting to rescue Princess Leia from the mysterious Darth ...",
...
}
```
# Setting up the bot
Now that we've got the basics figured out, it's time to setup the bot.
## Starting
To get started, create a new Answer in your Zendesk Bot.
After the welcome greeting, add an *Ask for Details* step and select your *Movie* Ticket Field we created earlier.
## Searching for movies
Next, add a *Make an API call* step that has the following endpoint URL: `https://www.omdbapi.com/?s=&apikey=abcd1234`. Next, put your cursor right after `s=` and click the {+} button to select your `Movie` custom field.
You can then use the Make API Call button to test the API call. If successful, you'll get a list of variables. Save the `Search` Array and give it a name like results. This variable will contain a list of results, and can be one, two or (any) length.
Once you created the array, you can set a specific name for each internal element of the array via the *Name - Value* fields. You can define as many items as needed, but we'll use these 4 in the next step: `title`, `year`, `id` and `poster`.
## Displaying the results
Before the new Dynamic features we could not show or handle an array of data, especially when that array had an unknown length. Like shown in my original [Ask For Details](https://internalnote.com/flow-builder-ask-for-details/) article, we could only store e.g. `results[0].title` as a variable, but couldn't account for `results[i]`.
Now we can by making use of the support for variables and arrays in the *Show Carousel* step:
When you add a Carousel, you can now choose to populate the data with a Dynamic message based on API Data. Select *Use array variable*, and select your `results` variable. You can then access all elements in the array to fill out your carousel, by clicking the {+} button in each field.
- Title: `results.title` will create a carousel element for each movie in our results and set the title to the, well, title.
- Description: `results.year` to display the release year
- Image: this is a big one, cause we never were able to use variables in images before either: `search.poster`, containing the URL of the movie poster.
To complete the carousel we'll also add two buttons:
- *See on IMDB* will concat `https://www.imdb.com/title/` and `search.id` to show how we can combine variables and regular text to create dynamic links
- *View Details* will **save** the option chosen by the customer in a new variable `chosen_movie` which we'll use in the next step. You can do this by choosing *Save Response* in the Button action dropdown, and setting the name and value of the new variabel to `chosen_movie` and `search.id` via the {+} respectively.
The result is a carousel of movie posters, with the title and two buttons below it.
## Getting movie details
Once the customer has chosen an option in our carousel by clicking on the *View Details* button, we can use the new `chosen_movie` variable to make a second API call and pull in more movie details.
Use [https://www.omdbapi.com/?i=&apikey=abcd1234](https://www.omdbapi.com/?s=&apikey=abcd1234&ref=internalnote.com) as the endpoint and use the {+} button to add `chosen_movie` as a parameter after the `?I=` part of the URL. This will make our API request dynamic by referencing the earlier customer choice.
💡
One thing to note here is that the second API call will fail during setup since we can not pass a variable for chosen\_movie in the setup flow. So it's best to setup this API call with a known movie ID like tt0083658 and make a sample call. Once you stored the required variables, you can then replace the placeholder with the actual variable via the {+} button.
Once you make a sample API call, we can store the variables we need to show them to the customer. Note that this API call does not return an array but a single movie, so we can store variables directly but can't define an array this time.
I stored the `genre`, `director`, `title` and `plot` as variables to use.
## Showing the details
Finally we come to the end of the bot flow. We can use the *Show Message* step twice to show the customer information about our movie, including its plot in a second message.
## The result
The end result is a more complex Zendesk Bot flow that can account for variable sets of returned data, and interact with customer choices to make new API calls.
This demo makes use of the new features to show movies, but you can use this flow to get a list of orders for your customer, and allow them to then chose an order and cancel, refund, or get delivery statuses for that order.
Or you can search your webshop for matching products, and show a carousel of products. When a customer clicks on one in the carousel you can then link them to the buying page, show more information, or offer a way to contact your Support team.
You can test the full flow via the proactive message that should have appeared bottom right, or by choosing the *Movie Bot* 🍿 in the suggested answers list of the widget.
# Conclusion
I've build a lot of bots on this blog and for customers these last few years, and where the Zendesk Bot used to be a very limited hardcoded flow, the additions of metadata, authentication, carousels and API calls gradually made the Zendesk Bot a lot more powerful.
This last new addition brings yet another set of API features that were sorely missing to interact with external APIs.
Is there stuff missing? Sure. I would love to be able to filter what's shown in a carousel, or to transform returned data (capitalise text, replace strings, ...) or map data to other values before showing it to the customer. Or writing a customers' choice in a carousel to a custom field so it's visible to agents and actionable in triggers upon escalation.
But I'm sure we'll see some of these remaining requests being fulfilled sooner rather than later, given the current speed of innovation in the Zendesk Bot.
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Zendesk Roundup for November 2023
URL: https://internalnote.com/zendesk-roundup-for-november-2023/
Last updated: 2023-12-05T07:30:55.000Z
Black Friday has passed, it's getting colder, and this is the final Roundup for 2023 with the Zendesk releases and announcements for November.
Earlier this month Zendesk held their [What's new for Q4 2023](https://www.zendesk.com/whats-new/?ref=internalnote.com) with a nice recap of all the new announcements from this last quarter, most of which I already covered in earlier [Roundups](https://internalnote.com/tag/zendesk-roundup/) this year. That being said, the focus of Zendesk of iterating and releasing new stuff this last year has been relentless.
Even this last month has seen the release of 2 new Zendesk AI EAPs for Guide and Voice (announced in October), a nice new GDPR feature and a big expansion of the API capabilities for the Zendesk Bot.
🎄
Last roundup for the year so I want to thank all my readers and subscribers for an awesome year. 2023 was the first full year of me writing on this blog, and the response exceeded my expactions! So thanks! 🥳
Since December is a month of festivities, I've still got articles planned for the 5th, 12th and 19th of December but will be taking a break the week of Christmas, so there will be no Internal Note for the week of December 26th.
The next article (a Zendesk roundup) will be on January 2nd.
Happy holidays!
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 🎉 New Releases
## 🤖 AI Powered Conversational Experiences
The Zendesk Bot got three nice additions this month.
First off, it's now possible to set [a custom Avatar](https://support.zendesk.com/hc/en-us/articles/6360125630618-Announcing-custom-bot-avatars?ref=internalnote.com) for your Zendesk Bot, replacing the Zendesk eyes with a more *on brand* image. (Side-note: I noticed it works best with a non-transparant icon with at least a 5 px padding to account for the corner radius).
Secondly, Zendesk made[ two new Zendesk AI intents available](https://support.zendesk.com/hc/en-us/articles/6437858147482-Announcing-employee-experience-intents-and-other-intent-updates-for-intelligent-triage-and-advanced-bots?ref=internalnote.com) for Employee Experience, as announced at the [AI drop event](https://internalnote.com/zendesk-ai-drop-keynote/): HR and IT, while also updating the models for all other available industries. So if you're using Zendesk AI, you now should see better results.
And finally, the Zendesk Bot Builder got a major upgrading by enabling [Dynamic Conversation Experiences](https://support.zendesk.com/hc/en-us/articles/6405324545050-Announcing-dynamic-conversation-experience-with-Zendesk-bots?ref=internalnote.com). This allows you to better interact with API endpoints and show carousels of data or use API results in other steps in your flow. I've got an entire tutorial written that will go live next week, so [subscribe](https://internalnote.com/plus) to the blog if you're interested in learning more!
## 👨🏻💻 Agent Workspace
The biggest release for this month is the availability of [Layout Builder](https://support.zendesk.com/hc/en-us/articles/6380051925018?ref=internalnote.com) for all Suite Enterprise customers. This new feature allows you to customize the Agent Workspace by adding or removing apps, fields or views, and rearrange the order in which all elements are shown on screen.
It's a shame this feature is only available to Enterprise users. I think it's warranted to say that Suite Professional users should at least get access to **one** Layout to customize, while leaving the ability to have multiple an Enterprise only feature. Zendesk customers have been asking for more flexibility for years now, so to lock this behind the most expensive plan only feels like a missed opportunity to me.
With the arrival of Layout Builder, side conversations also got an upgrade. The feature has moved from the top of the ticket view to its own dedicated context panel. This move has been enabled for all environments, but annoyingly requires admins to [confirm the setting in the admin center](https://support.zendesk.com/hc/en-us/articles/4408828503450?ref=internalnote.com#topic%5Fwzc%5Frjq%5Fw5b) first before it's usable, even if the feature has already moved.
## Zendesk AI for Agents
There's two AI features that are now available to enable in your Zendesk environment:
Macro Suggestions for Admins, the feature that shows possible macros for Admins to create, is no longer an EAP but is now available to all Zendesk AI customers with the added ability to filter the suggestions and see more related tickets on which the suggestion was based.
And for customers who requested the [Zendesk Talk AI EAP](https://support.zendesk.com/hc/en-us/articles/6330415782170-Announcing-generative-AI-powered-call-summarization-for-Zendesk-Talk-EAP-?ref=internalnote.com): that feature is now available to test. I played around with it earlier this week and it works like a charm: each phone call gets a fully written transcript, that, moments after the call wraps up, turns into a call summary with the flow of the conversation nicely summed up. Currently this summary lives outside of the Intelligence panel and doesn't interact with the Intent, Summary and Sentiment features shown there, but that's apparently planned to be integrated sooner rather than later.
### IVR Tagging
Speaking of Zendesk Talk, it got another release this month: [you can now tag IVR choices made by your customers](https://support.zendesk.com/hc/en-us/search/click?data=BAh7DjoHaWRsKwiaGNg7zAU6D2FjY291bnRfaWRpA9kYoToJdHlwZUkiDGFydGljbGUGOgZFVDoIdXJsSSIBg2h0dHBzOi8vc3VwcG9ydC56ZW5kZXNrLmNvbS9oYy9lbi11cy9hcnRpY2xlcy82Mzc0NzM1NDg1MDgyLUFubm91bmNpbmctSVZSLWtleXByZXNzLXRhZ2dpbmctaW4tWmVuZGVzay1UYWxrLWZvci1vbW5pY2hhbm5lbC1yb3V0aW5nBjsIVDoOc2VhcmNoX2lkSSIpZjc3NjAxZjMtMDE2Zi00MzM0LTgwYmItNTQxMGYyNTM3ODNkBjsIRjoJcmFua2kGOgtsb2NhbGVJIgplbi11cwY7CFQ6CnF1ZXJ5SSIMSVZSIHRhZwY7CFQ6EnJlc3VsdHNfY291bnRpUA%3D%3D--2477ca45c312c09a18706651c8a08e54f4a3906e&ref=internalnote.com). This request is another long missing feature that will make more complex routing based on IVR choices possible.
Each keypress made by a customer can be tagged, so that instead of routing a call to a specific group, you can now also use the tags to route to agents or groups with specific skills or responsibilities.
## 🔎 Help Center and Self Service
Zendesk Guide got dealt the short stick this month and only received a single update this month, making improvements in the way copying and pasting attachments from external editors like Google Docs or Word are handled. You no longer need to add them to the media library in order to use them in your article. They'll be uploaded automatically.
[Announcing improved integration with third party editors for Guide media libraryAnnounced on Rollout starts Rollout ends November 16, 2023 November 16, 2023 November 16, 2023 We are excited to announce improved integration with third party editors such as Google docs and…Zendesk helpKatarzyna Karpinska](https://support.zendesk.com/hc/en-us/articles/6382274865306?ref=internalnote.com)
## 🧱 Open and Flexible Platform
Last month saw the release of [Custom Objects](https://internalnote.com/announcing-the-custom-objects-series/), and already this got expanded with a new EAP this month.
The new Custom Objects Triggers EAP allows you to react to changes in a record and update tickets, notify users, or alert agents when a change happens.
For example: if you have a Custom Object that stores Support Contracts, you can now automatically notify a customer or account manager when that contract is about to expire. Or if you store repairs in object, you can update the ticket whenever the status of a repair has changed from 'in progress' to 'repaired'
[Creating and using object triggers (EAP)What’s my plan? Note: Object triggers for custom objects are currently in an Early Access Program (EAP). You can sign up for the EAP here. In addition to using custom objects related to tickets i…Zendesk helpJacquelyn Brewer](https://support.zendesk.com/hc/en-us/articles/6294230624410?ref=internalnote.com)
Additionally, Zendesk has [expanded](https://support.zendesk.com/hc/en-us/articles/6340850168346-Announcing-new-conditions-for-group-SLAs?ref=internalnote.com) the conditions available for group SLA policies to include all the existing conditions for SLA policies. Previously you could only filter by group, but now we can use any set of parameters to define these SLA policies.
## 🔐 Trust and security
Slowly but surely Zendesk is shifting from a platform that handled only Customer Support to one that can take care of both your customers and employees. The release of the HR and IT intents is one part of that strategy, the focus on Slack and Teams integrations another.
But by turning a customer experience platform into one that also offers employee experiences Zendesk does run into its own design from time to time. One example is the fact that originally customers always lived outside of Zendesk, and Support Agents work within Zendesk.
But in an employee support environment you run into the scenario where an employee who works in IT does their job in Zendesk, while also being outside of Zendesk if they need to contact IT. You want to prevent, for example, employees with access to the Agent Workspace getting insight in the way HR discusses their ticket internally. To handle this scenario Zendesk has now updated the way agent initiated tickets for which they are the requester, by allowing admins to choose between three visibly levels:
> \- Show all internal notes
> \- Hide internal notes on tickets assigned to private groups
> \- Hide all internal notes
## Timed Ticket Deletion
Zendesk releases a beta of the new [automated ticket deletion feature](https://support.zendesk.com/hc/en-us/articles/6062884435866?ref=internalnote.com) announced as part of their new Advanced Security and Privacy add-on.
It's a pretty basic feature available to everyone: you enable the feature and pick an amount of days, from which Zendesk will delete every closed and unmodified ticket older than that date. It's a very basic way to comply with GDPR or other privacy laws, but if you want more nuance, you're better of with [Sparkly's GDPR App](https://www.zendesk.com/marketplace/apps/support/206749/gdpr-search--destroy/?ref=internalnote.com), or move to the new Security add-on.
# 💡Insights
## Zendesk Engineering
Did you know Zendesk has an entire blog where their engineering team writes about how they build Zendesk?
[Moving from DynamoDB to tiered storage with MySQL+S3Originally we implemented a feature to persist an event-stream into DynamoDB to allow customers to retrieve them. This proved effective…Zendesk EngineeringShane Hender](https://zendesk.engineering/moving-from-dynamodb-to-tiered-storage-with-mysql-s3-cb3dc9bf813a?ref=internalnote.com)
This last one dives deep into handling large data streams and querying them.
I wonder if this new infrastructure was build for the new [Access Log API's](https://support.zendesk.com/hc/en-us/articles/6066010357530-Monitoring-agent-activity-with-the-Access-Log-API?ref=internalnote.com) as part of the new Advanced Data Security and Privacy add-on? 🤔
## Moving from Gorgias to Zendesk
This article has a nice overview of the steps requires to move from one CX platform to another. It comes with a nice spreadsheet that shows all the steps required.
[From Gorgias To Zendesk: Ben Segal’s 5-Week Transition GuideFrom Gorgias To Zendesk: Ben Segal’s 5-Week Transition Guide. Discover how Ben Segal, VP at Thesis, strategically transitioned from Gorgias to Zendesk, enhancing customer experience and operational efficiency in just five weeks. . | By Ben SegalLang.ai logo](https://www.lang.ai/blog/from-gorgias-to-zendesk-ben-segals-5-week-transition-guide?ref=internalnote.com)
# 🎥 Videos
# ⚠ Major Changes
## New Storage Limits
These coming weeks Zendesk will enable a new Storage Usage dashboard inside of the Admin Center that gives you insight in how you use the allotted storage in your Zendesk instance.
This dashboard might seem innocuous at first, but there's more at play here. Starting this year Zendesk will start charging customers for data-over-usage, and from what I've heard, the fees are not low.
So when you get access to the dashboard, take a look, and check where you can optimise your data-usage. Maybe that Data Deletion features announced this month can come in handy!
I'll surely write an article about this topic early next year to teach you how to reduce storage in your instance. Subscribe to learn more!
[Announcing the Storage usage dashboard in Admin Center and increased storage limitsAnnounced on Rollout starts\* Rollout ends November 28, 2023 November 28, 2023 December 7, 2023 \*Dashboard rollout for self-service customers occurs at a later date. See below for details.…Zendesk helpSean Newton](https://support.zendesk.com/hc/en-us/articles/6340068721434-Announcing-the-Storage-usage-dashboard-in-Admin-Center-and-increased-storage-limits?ref=internalnote.com)
# 📝 Articles this month
[Customer Service Trends 2024 - Zendesk ScorecardUltimate has released their Customer Service Trends Report 2024 today. Let’s see how close Zendesk aligns with their trends.Internal NoteThomas Verschoren](https://internalnote.com/ultimate-trends-2024/)
[Preview of the new Customisable CSAT EAP for ZendeskThe new Customisable CSAT EAP for Zendesk has arrived, finally allowing you to change your rating scale, choose emoji, numbers or labels, and customize your follow-up questions. This article contains an initial overview of the new feature, and shows how it works with existing API integrations.Internal NoteThomas Verschoren](https://internalnote.com/preview-of-the-new-customisable-csat-for-zendesk/)
[Preview of the new Generative AI for the Zendesk Help CenterThis article previews the new Generative AI for Help Center content that’s been made available in EAP by Zendesk.Internal NoteThomas Verschoren](https://internalnote.com/preview-of-the-new-generative-ai-for-knowledge-in-zendesk/)
[Announcing the Custom Objects series for ZendeskIntroducing is a four-part series on Zendesk’s new Custom Objects feature. These articles cover setup, data import, using Custom Objects in forms and with agents, expanding user profiles, and displaying Custom Objects in Help Center forms.Internal NoteThomas Verschoren](https://internalnote.com/announcing-the-custom-objects-series/)
✏️
I updated the [Return to sender](https://internalnote.com/return-to-sender/) article to make use of the new Lookup Field Placeholders by ID feature.
# And Finally
Good to know:
> Inbound emails are [limited to 64kB](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket%5Fcomments/?ref=internalnote.com#:~:text=in%20Agent%20Workspace.-,Body%20size%20limits,-Ticket%20comment%20bodies). For any email tickets that contain a table, spreadsheet, or text where the total size of the comment is greater than 64KB, the comment will be cut off to the point it reaches 64KB.
[Why am I unable to see the full body of an inbound email?Question Sometimes spreadsheets pasted in an email appear to be cut off. Why am I unable to see the full body of an inbound email in a ticket? Answer Inbound emails are limited to 64kB. For any ema…Zendesk helpJulio H](https://support.zendesk.com/hc/en-us/articles/6365792641946?ref=internalnote.com)
### Customer Service Trends 2024 - Zendesk Scorecard
URL: https://internalnote.com/ultimate-trends-2024/
Last updated: 2026-03-25T13:56:15.000Z
[Ultimate](https://ultimate.ai/?utm%5Fsource=internalnote&utm%5Fcampaign=trendsreport) is a customer support automation platform powered by AI known for its powerful bot and ticket automation capabilities. They have released their Customer Service Trends Report 2024 today highlighting trends in CX for the coming year.
Similar to [Zendesk's trend report](https://premiumplus.io/2023/02/02/leveraging-the-zendesk-cx-trends-2023-report/?ref=internalnote.com) last year, this report serves as a guide on how you can start building your CX experience vision for next year, and see where your company is up to speed, or lagging behind.
[Customer Service Trends 2024Double-down on efficiency, refocus on retention, stay ahead of the competition, and unlock revenue with Ultimate’s 2024 customer service trends guide.Ultimate.](https://www.ultimate.ai/guides/customer-service-trends-2024?ref=internalnote.com)
I always find these reports interesting to read since they help place feature releases into a bigger picture and help frame technology as solutions instead of just gimmicks.
Even though most of these kind of reports are build on actual data, they still serve as a marketing tool that directly helps the vendor that releases the insights. There's no better way to promote yourself to potential customers by combining problem, solution and justification in one document. And justifiably so.
So instead of looking at Ultimate's report through the eyes of Ultimate, I thought it would be a fun exercise to take their 7 trends, and see how they compare to Zendesk's current offering. Where can Zendesk assist in following the trends, where does it run short, or where is it a clear winner?
# Seven trends for 2024
The report starts with some preamble about how AI and especially Generative AI has shaken up the CX industry in these last twelve months, and then lists the seven trends according to Ultimate, their customers and a panel of CX Experts that could provide input for the report.
1. Customer service becomes the brand
2. Try before you buy
3. Agent experience takes center stage
4. Hybrid reigns supreme
5. KB management is key
6. Don’t automate at all costs
7. AI innovation hits inflection point
Let's see how Zendesk scores for each of these.
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# Customer service becomes the voice and vision of your brand
> By using AI to provide instant support and self-serve tools. By creating seamless experiences across channels. And by ensuring a human is always on hand to provide empathetic care. Support conversations are a crucial touchpoint — and an opportunity to let your brand identity shine. When agents deliver CX that embodies your values, you’ll win first place in customers’ hearts.
This first trend is all about making sure customers get the right experience and support from your company. A good customer experience can make someone an ambassador, a bad experience a detractor. People share the good and the bad with friends, and a trusted brand is a brand where you return too, time and time again.
I think Zendesk is a perfect fit for this trend. The help center, asynchronous conversations, a hybrid model of bots, autoreply and actual human agents make sure customers can both easily self serve their issues, as well as get escalations to the right agent via skill based routing.
And by using both custom Help Center themes, a branded both, and availability of your support on your website, social channels and traditional email or phone, your support can be wherever customers interact with your brand.
**My score: A+**
# Generative AI lowers the barrier to entry for trialing automation tools
> Generative AI-powered chatbots work straight out of the box, and more providers are offering businesses free trials so they can test the benefits before committing. And CX leaders are taking them up on it.
If you would have asked me six months ago if Zendesk AI was up to the challenge I would probably have set no. But with the latest [AI Drop](https://internalnote.com/zendesk-ai-drop-keynote/) Zendesk is quickly catching up with the existing big AI Chatbot players.
One of the biggest risks with using any Generative AI model inside of your chatbot is the risk of the model pulling in data and replying with bad or wrong data. I keep going back to this [quote](https://www.youtube.com/watch?v=9aCg7jH4S1w&ref=internalnote.com) from Zuckerberg:
> Yeah, so our view is that there’s actually going to be a lot of these that people talk to you for different things. \[...\] let’s say you’re a small business and you want to have an AI that can help you interface with customers to do sales and support. You want to be pretty confident that your AI isn’t going to be promoting your competitor’s products, right?"
One thing that Zendesk does well is that it purposefully limits the bot to your Help Center and only your Help Center. This gives you two benefits:
1. The bot works out of the bot with your data. No training, no content management, if you did your homework with your traditional FAQ, the bot can use that content to answer
2. Enabling Generative AI is 2 checkboxes: enabling it + choosing a persona.
There's some limiting factors still at play though: Zendesk's intent models are fixed for an industry and copied across customers with no option for real customisation or training on your actual data. from my testing the models work well if you're in one of the supported industries, but custom models will still give you better results, especially for unique or complex industries.
**My score: B+**
# Agent experience takes center stage as providers build with them in mind
> These include generating on-brand replies, analysing message sentiment to help agents adjust the tone of their response, or creating summaries of long message threads \[...\] More importantly, gen AI makes support teams’ roles more rewarding too. Automating repetitive tasks frees agents to focus on complex cases that require their talents
I think this trend is one where Zendesk is at home, and every Bot company is playing catch up, if we can even consider them playing in the same league.
Zendesk has always been an agent-first platform with the Agent Workspace. They have an omnichannel approach where every channel is turned into a ticket handled in the same interface with context around it.
Now with Zendesk AI you get summary, intelligent triage, suggested tickets and macros, expand and tone shift. These are all items an agent traditionally would need to do manually: categorise, write a summary for team leaders upon escalation, search for related content to resolve an issue, it's all busy work that needed to be done per ticket. When using Zendesk AI I've already seen agent being freed up to only spend time on actually resolving issues, leaving the busy work to the system.
For me, ticket deflection with self service and bots, and agent automation with context and intelligent triage are really two sides of the same coin. When the easy stuff gets solved by the system, you also need to enable your agents to handle the complex stuff. Both in training and resources, as well as making sure they have and get the time to dive deep into these tickets, without wasting time on stuff that AI can automate.
**My score: A+**
# The hybrid approach to support automation unlocks its full potential
> Generative AI has taken the world by storm — but that doesn't mean conversation design is out in the cold. The real magic happens when you combine the ease and flexibility of gen AI with the precision and control of conversation design.
I have to admit, I had to read this section of the report twice before I got it. When I glanced over the report the first time I assumed hybrid was about human vs bot. But no, it's about the choice between 100% bot driven, vs human curated flows.
If you rewind a year Zendesk's (and every other bot) approach to flow building was basically this: someone needed to build a flow that, step by step, guided the customer through a fixed path until they reached either a resolution, or got escalated to an agent. We all know these bots: they start with a set of options, and once you ask your question you go down the rabid hole of questions and answers until you get stuck and frustrated. Or, if the intent mapping is done well, and the company has the insights to know where customers get stuck, you get bots with good flows that follow customer logic.
During the AI Drop event last month, Zendesk talked about how the best approach to building a bot is not "go build a hundred flows to tackle each scenario". That doesn't scale and isn't always worth the effort. A better approach is to leverage your existing knowledge base content and generative AI.
Your customers ask any type of question to the bot. The bot uses AI to map the question to an intent, and cross checks with your FAQ content to pull the right information from your articles. It then uses Generative AI to reply with an answer.
Next, you look at your reporting to see what intents are triggered, and where customers clearly are helped or satisfied with the offered bot answers. These flows are ideal for manually crafted flows: you can offer nuance, you can pull in external API date (e.g. Where's my order actually returns a delivery date instead of an article) and prevent issues.
This combination of bot and manual flows is scalable. You can quickly write an article to respond to a question, and moments later the bot will start serving that response to customers. Or you can spend the time to build a full flow for the bigger scenarios.
Zendesk AI has this approach build in with the [Intent Suggestions](https://support.zendesk.com/hc/en-us/articles/5537827011994-Using-AI-powered-intents-with-conversation-bots?ref=internalnote.com#:~:text=Understanding%20intent%20suggestions,bot%20the%20questions%20came%20through.), so they're clearly on the right path here. But since they only offer a fixed set of intents, and you can't add your own, you might be blindsided in your reporting for the stuff Zendesk can't map.
**My score: B+**
# #1 new support skill: Knowledge base management and optimisation
> Knowledge base management empowers CS agents to offer better, more streamlined support while also enabling customers to help themselves when and wherever it suits them. And in the age of generative AI, it is absolutely essential your knowledge base is in top form.
For a while now I've made the difference of talking about Knowledge Base when talking about managing content and articles, and Help Center when talking about the website where customers can read content.
Within Zendesk your articles feed your Help Center, are available to Agents in the Context Panel, are offered via email with autoreply to your customers, and are also used to offer a list of articles, or generated replies in your web widget and Zendesk Bot.
So detecting gaps in your knowledge, being able to translate problems into steps to resolve, and keeping the data up to date are foundational skills needed to offer good support to both your customers and agents.
Or to paraphrase a terrible CGI creation from a galaxy far far away:
> No content, no articles, no self-serve.
Zendesk Guide has long been a cornerstone of Zendesk's offering, even before Zendesk Suite was a thing, and with content cues and the additional review and publishing features in Suite Enterprise, you've got the basis to manage your content right within our CX toolkit.
My approach to this has been to create a report that pulls in search queries, ticket escalations (so self service failed) and categorised tickets. Then create a weekly report that shows the top 10 and bottom 10 of these there datapoints. And then sure your Help Center captures both the top and bottom ten for each of those datapoints:
- Top 10: the Frequent in FAQ.
- Bottom 10: make sure unique cases are also served so you show your expertise.
**My score: A-**
# Ticket automation becomes a differentiator for the most human CX teams
> AI can free your agents from their most repetitive, mundane tasks so they have more capacity to provide better, empathetic support to your customers. \[...\] the area in greatest need of improvement is eliminating manual tasks like routing and tagging tickets.
I kinda pointed to it in an earlier trend already, but agents should spend time helping customers, and shouldn't lost time doing repeated tasks.
If your platform removed repeated questions from agent queues, makes sure the agents get tickets aligned to their skills (reducing the need for rerouting) and offer full context with intent, sentiment, related tickets and things like purchase history, customer profile e.a., then the agent job is basically solving the issue, and not so much doing a lot of bookkeeping and filling in fields.
Zendesk's omnichannel routing, intelligent triage, Agent Home, triggers and automations are all pieces of the puzzle that forms Agent Workspace.
**My score: A+**
# AI innovation hits inflection point, fast-tracking global business growth
> The AI market is rapidly maturing. And the pace of change means we’ve hit a tipping point. Today, the effort-to-reward ratio has dramatically shifted — with AI-powered automation making it easier than ever for brands to scale.
This last one for me is where trends meets marketing. Naturally, every report this year will point out how every tech company threw out their roadmap and realigned their development to include OpenAI's platform into their product.
So it's only natural that the expected trend is that those companies get a return on their work and investment by now pushing customers to start using these AI powered tools. Even when not all of them have proven to actually work, provide good results or what the impact will be in 1, 5 or 10 years of the choses made today.
Criticism aside, I do see the same results and short-term impact that these new AI technologies have made possible, so its an obvious thing to put in a trends report cause it's true, no?
Zendesk itself also jumped on the bandwagon with their two-part Zendesk AI release this year, it remains to be seen how the platform will evolve and integrate more AI tools across their product suite.
# Conclusion
I joked (kinda) at the start of this article that most of these trends reports are self-serving to the publisher. But I have to give points to Ultimate for including this one:
> **Shop around for the right provider** \- Take a few generative bots for a test ride before committing. Choose an established automation partner that can handle complex use cases, has experience with your industry, and has the flexibility to grow with your company.
It's indeed true that not every provider is the same, and that a single platform can't solve all needs.
When looking at the trends Ultimate proposed my end-score for Zendesk is a solid A-, with the biggest weaknesses in the product the fact that Zendesk only offered prebuilt AI models scoped to a small set of predefined industries: Retail, Tech, Finance, HR, IT and Travel (announced).
These models are based on Zendesk's billions of tickets, but lack the ability to align with the actual intents of a company. Imagine you run public transport company that kinda does retail (tickets, subscriptions), but is also kinda travel (destinations, itineraries)? If that's the case, Zendesk AI's approach isn't a really good fit since the intents kinda align, but don't fully overlap with your actual ticket categories. So that's where purpose-build bot builders with custom intent models currently take a lead.
But when it comes to the other side of the coin, handling escalations to real agents and making their work easier and more rewarding, that's where Zendesk has a clear lead over other platforms.
However you take it, it's nice to see how Zendesk and the market overall seem to be running in the same direction. It gives me confidence in the route the platform has taken.
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Zendesk Custom Objects - Part 4: End-User and Forms
URL: https://internalnote.com/custom-objects-part-4-end-user-and-forms/
Last updated: 2025-09-08T06:41:44.000Z
In this four part series we'll explore the new Custom Objects from an admin, agent and end-user standpoint.
#### Custom Object Series
1. [Introduction to Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/)
2. [Custom Objects and Tickets](https://internalnote.com/custom-objects-part2-tickets/)
3. [Custom Objects and Users](https://internalnote.com/custom-objects-part-3-users/)
4. [Making Custom Objects available to End-Users](https://internalnote.com/custom-objects-part-4-end-user-and-forms)
The first two articles showed you how to set up custom objects and how agent can interact with them. The previous expanded user profiles with Custom Objects and showed how you can use them to trigger different priorities based on linked records.
This article will wrap up the series by solving a missing feature in Zendesk: Showing Custom Objects in Help Center forms and make the selected record available to agents.
[End-User access for Custom Objects and Lookup Fields (EAP Preview)Zendesk’s Custom Objects, released last year, enable you to expand platform data, like linking assets or contracts to tickets. Initially, end-users couldn’t interact with custom objects, but a recent update allows you to add Lookup Fields in forms.Internal NoteThomas Verschoren](https://internalnote.com/end-user-access-for-custom-objects/)
Zendesk has released a native solution for this feature. The steps in this article still work, but I recommend migrating to the native solution!
# Custom Objects for End-users
The previous Legacy version of Custom Objects [allowed us](https://developer.zendesk.com/api-reference/custom-data/custom-objects-api/permissions/?ref=internalnote.com) to set permissions for both agents, admins and end-users. This way you could make the objects available as read-only elements for (anonymous) end-users and use the API to inject custom objects on your Help Center.
This new version of Custom Objects manages the permissions entire via the Admin Panel. We've previously used the permissions to give our agents edit rights to link Pokémon to a specific end-user in the previous article.
Sadly, with the migration to this new version, the ability to give end-users access to the objects has been lost. This is partially due to an inherit security risk with the ability for Custom Objects to link users and organisation via Lookup Fields. If not done carefully, this might make your entire customer list available in e.g. a Lookup Field on a form. For this reason, the end-user is currently not available out of the box.
There are scenario's where it would be useful to have these objects available though. Maybe you want your customer to select a specific device when filling out a form. Or you want to show a list of retail locations to handle a return, based on a Location Object in your Zendesk instance.
This article will show you a workaround to fix this limitation.
# What we're building
In this article we'll build the following flow:
1. A user opens a specific Help Center form on the Help Center
2. We load a dropdown of all Pokémon for them to select
3. Once they choose a Pokémon we pass the ID to a hidden intermediate ticket field
4. When the ticket is submitted we use that field to set our Pokémon Lookup Field
5. The agent sees the linked object in the Agent Workspace and can view it's data next to the ticket.
# Getting the data to the Help Center
The easiest way to fix the end-user limitation is to use an external service as a proxy to (securely) load the API.
This proxy would be reachable from your Help Center, gather the records via API remotely, and return the JSON data back to the Help Center to use in your code.
## Securing the connection
Since we are proxying the data from our Zendesk instance to a public Help Center it's important to consider the security implications of this.
### Only allow your website
Zendesk recently introduced [Help Center JWTs](https://developer.zendesk.com/api-reference/help%5Fcenter/help-center-api/help%5Fcenter%5Fjwts/?ref=internalnote.com) to secure outgoing connections to external services. They wrote a clear example [in this article](https://developer.zendesk.com/documentation/help%5Fcenter/help-center-api/secured-requests/?ref=internalnote.com#making-third-party-api-requests-with-help-center-jwts) on how to use it.
🔐
In our example we'll pass the JWT token, but since handling JWT validation is an entire exercise on its own, we'll use `Access-Control-Allow-Origin` to lock the API to just our environment. This article focusses on leveraging the API after all and your deployment might not necessarily use the same Cloudflare environment as me.
### Scope the API
My proxy only supports the `/api/v2/custom_objects/pokemon` endpoint, and only allows for `GET` requests. This way the exposed data is limited to known data I don't mind being public.
It's going to be public by design once I add it to my webforms anyhow.
## The Proxy Code
I added a sample worker to this repository in the `/worker/index.js` folder.
[GitHub - verschoren/pokedexContribute to verschoren/pokedex development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/pokedex?ref=internalnote.com)
This worker takes the incoming URL and uses the path and parameters to create request that goes to Zendesk.
If I make a request to `https://myworker.workers.dev/api/v2/custom_objects/pokemon/records.json` the worker will take the URL and turn that into `https://subdomain.zendesk.com/api/v2/custom_objects/pokemon/records.json` and add the necessary authentication tokens.
We'll also filter the incoming request to only support `GET` requests and filter the pathname to only allow for `custom_objects/pokemon` .
Do deploy this code you can go to [https://workers.new](https://workers.new/?ref=internalnote.com) and create a new Cloudflare Worker. Copy-Paste the code from the repository in the editor.
You'll need to change the following four parameters:
```javascript
const zendesk = 'https://internalnote.zendesk.com'; //your zendesk domain
const custom_domain = 'https://support.internalnote.com'; //your help center custom domain
const authentication = '' //base64 encoded username/token:api_token
const object_name = 'pokemon' //name of your custom_object
```
💡
I really hope we can get access to an end-user permissions options again to make this step redundant.
Once you've deployed you'll need to copy the URL of the worker. That will be similar to [https://customobjectsproxy.verschoren.workers.dev](https://customobjectsproxy.verschoren.workers.dev/api/v2/custom%5Fobjects/pokemon/records.json?per%5Fpage=100&ref=internalnote.com).
# Setting up the flow
Setting up the rest of the flow is done entirely in Zendesk. We assume that you've already read the previous articles in this series. This means you have a form for your Pokémon inquiries, and have created a Lookup field that links to a Pokémon Custom Object.
## Create an intermediate field
Zendesk does not allow us to create Lookup Fields for end-users, which is the reason we need to build this in custom code.
To be able to pass the chosen Pokémon from the webform to the agents, we'll need to use an intermediate text-field that will (invisibly) store the ID of the chosen Pokémon.
To do this create a new Ticket Field of Type text. Call it `Intermediate Pokémon` and copy down the `ticket id`.
Make sure customers can edit the field, and add it to the your Webform.
## Setup a webhook
The end result of this flow should be the a filled in Lookup Field that shows the agents which Pokémon the end-user is talking about.
Since end-users can't fill in Lookup Fields, we want them to fill in the Intermediate Field and we'll use a webhook and trigger to fill in the Lookup field for them.
First off we need an 'Update Ticket Webhook'.
Go to the *Admin Panel > Apps and Integrations > Webhooks* and create a new Trigger based webhook
- Name: Update Ticket
- Endpoint URL: [https://d3v-verschoren.zendesk.com/api/v2/{{ticket.id}}.json](https://d3v-verschoren.zendesk.com/api/v2/%7B%7Bticket.id%7D%7D.json?ref=internalnote.com)
- Request Method: `PUT`
- Authentication: Basic Authentication: `admin@domain.com/token:zendesk_token`
## Setup a trigger
And to wrap up the Admin setup of this flow, we'll need a trigger that reacts whenever a ticket is created and puts the value of the Intermediate Field into our Lookup Field.
Create a trigger with the following conditions: we want to only run this when a ticket is created in our Pokémon Form, and we want a value to be present in the intermediate field (meaning a Pokémon has been chosen)
Once these conditions are met, we want to execute a *Webhook* that calls the Update Ticket webhook via *Notify > Active Webhook.*
Add the following payload to the webhook. Note that `id` should be the ID of your `Lookup Field`, and value should be the `id` of your `intermediate field`.
```json
{
"ticket": {
"custom_fields": [
{
"id": 14502103878802,
"value": "{{ticket.ticket_field_14512152430226}}"
}
]
}
}
```
# The Help Center
We've now got two out of three pieces setup. We have a way to get the Custom Object records for end-users, and we have a way for the object the end-user has chosen to be added to our Lookup Field.
We do miss one piece of the puzzle: we need to show a list of Pokémon to our end-users, and set the chosen option to our Intermediate Field so our trigger can use it.
The way we'll do this is by replacing the `intermediate field` with a custom dropdown of our own. Whenever the customer chooses an option in the dropdown, we'll fill in its value in the intermediate field so our trigger can use it.
The code below should be added to the `new_requests_page.hbs` of your Help Center theme.
## Setup
We start by checking which form is displayed so we're sure we only change our Pokémon form. This also keeps the code sorted per form if we modify multiple forms.
```javascript
$( document ).ready(async function() {
const urlParams = new URLSearchParams(window.location.search);
const ticket_form_id = urlParams.get('ticket_form_id')
if (ticket_form_id == '10992076994578'){
....
}
});
```
Next we'll hide our Intermediate Field and add an empty dropdown which we'll later fill with the Custom Object Records.
```javascript
const intermediate = '14512152430226' //ID of our ticket field
$('#request_custom_fields_'+intermediate).hide(); //hide custom fields
$('#request_custom_fields_'+intermediate).after(`
--
`);
```
Since we might want to use the new Zendesk Guide JWT option we also need to generate a JWT token
```javascript
const jwt = await getJwt();
async function getJwt(){
fetch("/api/v2/help_center/integration/token.json")
.then(response => {
var json = response.json();
return json.token;
})
}
```
We need to load the data from our Worker to get access to the records.
```javascript
const base = 'https://customobjectsproxy.verschoren.workers.dev';
var api_url = "/api/v2/custom_objects/pokemon/records.json?per_page=100&sort=id"; //the sort value makes sure we sort by the external_id
await getRecords(api_url);
async function getRecords(path) {
var url = base + path;
$.ajax({
url: url,
dataType: 'json',
headers: {
"Authorization": `Bearer ${jwt}`,
"Content-Type": "application/json"
},
success: function(response) {
//handle response
}
});
}
```
```json
//returned response
{
"custom_object_records": [
{
"id": "01HD0W1CWVCW3JYQFM91MWM1GY",
"name": "Charmander",
"external_id": "4"
...
},
...
]
}
```
Once we've captured the data by fetching the `api_url` we'll loop through the records to append them to our dropdown.
```javascript
//add to dropdown
$.each(response.custom_object_records, function (i, record) {
$('#select_pokemon').append(`
${record.name}
`);
});
```
Since it's possible we have more than a 100 records, we can use the `meta.has_more` key to check if there are more records, and if there are, we'll run our code again.
Since we proxy the requests to our worker we need to change the `links.next` url to use our worker `base` URL instead of the Zendesk URL.
```javascript
//check if more than 100
if (response.meta.has_more == true) {
var next = response.links.next.replaceAll(/https:\/\/(\w+\.)*zendesk\.com/g,'');
getRecords(next);
}
```
And finally (finally!) we need to react to changes in our dropdown to set the intermediate field:
```javascript
$('body').on('change', '#select_pokemon', function(){
$('#request_custom_fields_'+intermediate).val(this.value);
});
```
# The Result
Now that we've done the work, let's see how this all combines into one flow for our customers.
In the screenshots below you'll see a trainer go to our form, select his Pokémon type from the dropdown, fill in all other fields and submit the form.
On the Agents' site, they see a filled in Lookup Field linked to a Custom Object. The sidebar app we build in a previous article shows the Pokémon's information, and the native Record Preview also shows our record information.
## Where we take it from here
This flow is a bit long and complex, but in the end allows you to present custom objects to end-users and have a native and automated experience for your agents.
You can expand the Proxy Worker to allow for more types of objects, and then clone the trigger to allow for more intermediate-to-lookup field mapping. Off course, this also requires you to update the Help Center code to get the right objects and fill in the right intermediate field.
Like I mentioned earlier, I **really** hope we get native lookup fields sooner rather than later so I can delete this fourth article and replace it with a native flow. But until then, this offers a working solution.
# Wrap up
This article wraps up the Custom Object series. In the series we started with an introduction to Custom Objects and showed how to create object types and import records.
We then moved to how Agents interact with records and added a custom app that displays the data more nicely.
The series then moved to showing how we can augment user profiles with record data, and allow for priority changes based on the linked records.
And finally we allowed customers to interact with objects and choose values in ticket forms.
So now, I pass the control to. Which types of custom objects will you add to your Zendesk instance? Leave a comment to the article or send me an email with your results!
And if this series was useful, please consider sharing it with your colleagues and partners!
### Overview of the new Customizable CSAT for Zendesk
URL: https://internalnote.com/preview-of-the-new-customisable-csat-for-zendesk/
Last updated: 2024-10-14T18:21:44.000Z
At the closure of every ticket, Zendesk has always offered you the ability to send out a CSAT survey to your customers. This survey allowed customers to give a "Good, I'm satisfied", or "Bad, I'm unsatisfied" rating, accompanied with a comment field, and an optional list of predefined reasons why they're unhappy.
I've seen plenty of Zendesk customers who used [placeholders](https://support.zendesk.com/hc/en-us/articles/4408886858138-Zendesk-Support-placeholders-reference?ref=internalnote.com#topic%5Fnnz%5Fopl%5Frc) to turn these two options into images, emojis or other designs, like so:
```html
🥳 I'm happy .
😔 I'm sad .
```
But the real thing most Zendesk users actually wanted was the ability to customize the CSAT feedback. More options, different scales, a different set of reasons,... you name it.
So, this is where the new Customisable CSAT feature comes in.
# What's Customisable CSAT?
In short, this new feature allows you to customise your entire CSAT survey:
- You can choose between a rating scale with 2, 3 or 5 options.
- You can choose between numbers, emoji or plain text to differentiate between options.
- You can add custom labels to each option.
Once a customer has rated your ticket, they can then add additional feedback via the comment field, or give a reason why they gave bad feedback:
If they've given feedback, they have 48 hours to come back and change their feedback if they want to. All feedback changes are logged in the tickets' event timeline.
## Setup
Setting up the new CSAT features is pretty straightforward. Instead of being hidden under *Admin Panel > End Users > Settings*, they now get a nice new section in the *Objects and Rules* part of the Admin Panel.
Enabling the customisable CSAT is done with a simple toggle, which shows the setup wizard with three configuration sections:
1. CSAT survey, where you can set the Survey question, scale and type, and change labels.
2. a Follow-up dropdown to capture the reason for negative feedback.
3. Configure a comment field.
### Survey Questions
When setting up your CSAT survey you can choose the question asked (How would you rate our service?) and choose a scale and type.
One curious thing is that for each scale Zendesk predefines the amount of positive and negative options you get.
You get a satisfied/unsatisfied when going for the classic 2 option CSAT, you get 2 negative ones when going for 3 buttons, and get 2 positive options when going for the maximum of 5 options.
🤔
I find it a bit curious how negative options are more represented than positive feedback, but I assume that most customers leave negative feedback anyhow when they fill in such a questionnaire?
Once you choose your scale, you can pick a type of layout. You can go for the aforementioned numeric scale, you can pick a custom emoji for each option, or go for plain text buttons.
One thing to note here is that, even though customers might give a 2,3 or 5 rating, in the end it all gets converted into **the same good/bad binary** **options** from before. So you won't see any difference in your reporting, it's just the customer facing labels that differ.
### Follow-up questions
Once you've setup your CSAT type, you can edit the two Follow-Up questions:
- For negative feedback you can add up to 9 reasons for why, together with a guiding headline.
- For all feedback types you can add an open ended comment field.
You can delete these questions if you want, and can always add them again, but you can only add one of each. So if you were hoping of adding a complex survey after each ticket, you're out of luck. You still get the same basic *question-why-comment* structure as before, but now in a more customisable way.
## Translations
One issue I encountered when testing this new EAP is Dynamic Content and translations. For now, when you add a Dynamic Content placeholder like `{{dc.good}}` as a label, it gets translated, but somehow still displays the curly braces instead of only the translated string
## API
From an API standpoint this new customisable CSAT is identical to the old CSAT format, so you can safely switch to this new version without needed to change any code if you would, e.g., export your CSAT into an external platform:
```json
{
"ticket": {
"satisfaction_rating": {
"score": "bad",
"id": 15264032482706,
"comment": "sad",
"reason": "The issue was not resolved",
"reason_id": 15264002643474
},
...
}
}
```
And similar, when checking the `reason_id` via API, it still resolves the same way as before:
`https://yourdomain.zendesk.com/api/v2/satisfaction_reasons/15264002643474`
```json
{
"satisfaction_reason": {
"url": "https://d3v-verschoren.zendesk.com/api/v2/satisfaction_reasons/15264002643474.json",
"id": 15264002643474,
"reason_code": 1001,
"value": "The issue was not resolved",
"raw_value": "The issue was not resolved",
"active": false,
"created_at": "2023-11-22T20:42:16Z",
"updated_at": "2023-11-22T20:42:16Z",
"deleted_at": null
}
}
```
## Conclusion
All in all, this is a nice update to the existing CSAT surveys that allows for some nice customisability, without actually breaking any existing features.
One thing I would love to see is some agent visibility into what the customer actually chose. The agent currently only sees a bad/good result, but can't see if the customer choose option 1, 2 or 3 in a 5 step rating:
Similar, sometimes you also want to measure why someone gave a good rating in a similar dropdown as the bad rating. But for now, that is also not possible.
So this feature seems to be exactly what's promised on the label: Zendesk CSAT, but customisable.
I like it.
### Zendesk Custom Objects - Part 3: Users
URL: https://internalnote.com/custom-objects-part-3-users/
Last updated: 2025-09-08T06:41:50.000Z
In this four part series we'll explore the new Custom Objects from an admin, agent and end-user standpoint.
#### Custom Object Series
1. [Introduction to Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/)
2. [Custom Objects and Tickets](https://internalnote.com/custom-objects-part2-tickets/)
3. [Custom Objects and Users](https://internalnote.com/custom-objects-part-3-users/)
4. ➕ [Making Custom Objects available to End-Users](https://internalnote.com/custom-objects-part-4-end-user-and-forms)
The first two articles showed you how to set up custom objects and how agent can interact with them.
This third article will dive into expanding user profiles with Custom Objects and how you can use them to trigger different priorities based on linked records and how to create custom record tied to a specific user.
# Map records to user profile
So far we've covered creating records and linking them to tickets. But we can also use Custom Objects to expand User Profiles in Zendesk.
Traditionally whenever we wanted to expand a customer profile we'd resort to user fields or tags. We tag a customer with `VIP` or set the Customer Type field to `prospect`. With Custom objects we can link rich objects to customers and provide more information than a regular field can.
Going back to our Pokémon example, in that world trainers can earn badges which gives them certain privileges. In our demo we'll use them to do the following:
1. Make it clear what badge a trainer has earned
2. Increase the priority of a trainers' ticket based on their badge
3. Add the Gym Owner in cc to the ticket.
## Creating the Custom Record
As always, whenever you create a record it's best to first map out the required data.
In our case we need the following fields:
- Name: name of the badge
- Gym Owner: Lookup Field linked to a specific `user` in our Zendesk environment
- Gym Type: Lookup Field linked to the Custom Object `Pokémon Type`
- Information: text field with info about the badge.
Looking at the previous articles you should already know how to setup most of these fields, but let's dive into one new type: a lookup field that links to a user.
We've setup a few Light Agents in Zendesk tagged with `gym_owner` . These people are responsible for giving second line support on their customers.
Now, in the Custom Object we're creating, add a new field of type *Lookup Relationship* and named Gym Owner.
Select `User` as the related object. Opposed to all prior steps, this time we do want to setup a filter. We only want to show or link users that are tagged with `gym_owner` since these are the only people we want to be linked to badges.
## Creating the badges
Now that we've created the record type, we can start adding records to our object. We'll create a Cascade Badge, of type Water linked to the Gym Owner *Misty.*
## Linking a custom record to a user
Lookup fields work in two directions. In our case we have a badge with a Lookup field that links a gym owner. This is **one record** that reference the badge and includes metadata like type, owner,...
Each badge can only have one gym owner linked.
We can also add a lookup field to a user profile in Zendesk that links a badge to a user. Each **user** in Zendesk can have **one badge linked**. But the same badge can be linked to multiple users.
To link badges to users we need to add a *User Field* via the *Admin Panel > People > Configuration > User Fields*. Create a new field of type Lookup Relationship. Give it a name like "Current Badge" and set the related object to the Badge object we just created.
Now, when we open a user profile in Zendesk, a new user field will appear on the left side. Click on the Badge dropdown and select the Cascade Badge we just created.
Two things will happen. When we look at a customers' essentials card it will show their current badge. And when we look at the badge in the Object Viewer we'll see Badge listed as one of the current owners.
When we click on the badge name in their profile or in the Essentials Card it'll open the detail view of the Badge and we can view its owner, type and other information stored in the record.
# Interacting with custom objects via triggers
Just being able to view data next to tickets is nice, but it we can also interact with that data to automate processes.
## Set ticket priority based on the current badge
We can use a trigger to give all tickets created by badge holders a higher priority in our instance. This will benefit badge owners because a higher priority often means a faster SLA, thus a quicker assist to their issues.
You can see this trigger in action in the screenshot all the way at the start of this article 😇
Setting this up is done by a trigger with the following conditions:
- Ticket > Ticket is `Created`
- AND Requester > Current Badge is `Present`
Choose the following action:
- Ticket > Priority to `High`
# Make assets unique per user
So far in this tutorial we've linked predefined objects to users and tickets. But there are use cases we're objects have a specific type, and exist multiple times. Think multiple copies of a specific laptop, or versions of a software license, or in our cases: the different between a Pokémon species and a specific captured specimen.
So let's expand our Pokémon object to make a difference between Pikachu as a species, and the specific Pikachu captured by Ash, our requester used in the demo's.
## Setting up object
Just as before, let's first define our object. We have a `Captured Pokémon` owned by a `trainer`. That Pokémon is of a specific `pokemon` species.
Let's set this up. We first create a new Object Type named `captured_pokemon`. We add the following fields:
- Name: Captured Pokémon
- Lookup Field: Pokémon Species, that links to the `pokemon` object we created in the first article of this series
- Lookup Field: Trainer, that links to Users. We do not need a filter since anyone can be a trainer.
Since we want our agents to be able to modify these Pokémon, or even create new ones we didn't know of, we'll change the permissions to give all our agents add, edit and delete rights. (Some Pokémon might escape after all).
## Creating a new Captured Pokémon
Go the the Object Viewer via the navbar and choose the *Captured* *Pokémon* object. Click add and create a new Pokémon. Give them a name, choose a species and select their trainer.
Once added, the Pokémon will show up when looking at the Captured Pokémon object, and will also show up on the Trainers profile in Zendesk in the Related tab.
Naturally, this is only part of the story. Now that we have a specific Pokémon to refer to, we might want update our Ticket Forms to refer to that Pokémon instead of the generic `Pokémon` type.
This is done similar to how we did it earlier for the regular object. Create a new Custom Ticket Field of type *Lookup Relationship*, and select the `Captured Pokemon` as a related object.
Now your agents can link tickets to a specific Pokémon, instead of the generic species.
💡
Or similarly, you can link a ticket to a specific laptop with a specific serial owner by a colleague, instead of the generic laptop type to create a more detailed report on actions taken on that laptop.
## Update to the sidebar app
In the previous article we build a sidebar app that showed a nicer overview of the Pokémon data. If we want to use these Captured Pokémon we need to make an extra API call to show the data.
First we need to get the value of the `Captured Pokémon` ticket field:
```javascript
var captured_pokemon_id = await client.get('ticket.customField:custom_field_14504649143442');
```
We then use that value to retrieve the data of the `Captured Pokémon` record, and more specifically its linked species. This way we capture the `Record ID` of the Pokémon we need.
```javascript
var linked_pokemon = await client.request({
url: '/api/v2/custom_objects/pokemon_captured/records/'+captured_pokemon_id+'.json',
type: 'GET',
dataType: 'json'
}).then(async function(pokemon){
return pokemon.custom_object_record.custom_object_fields.pokemon_species;
});
```
And now we can use that `Record ID` to get the data of the actual species:
```javascript
return await client.request({
url: '/api/v2/custom_objects/pokemon/records/'+linked_pokemon+'.json',
type: 'GET',
dataType: 'json'
});
```
The application in the repository we used in the last article is updated to include this functionality. Just make sure to set the "Use Captured Pokémon" setting in the apps' configuration to true when installing.
[GitHub - verschoren/pokedexContribute to verschoren/pokedex development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/pokedex?ref=internalnote.com)
## Notify trainer
Similar to how we used triggers to set the priority of tickets based to linked records, we can also use other data in a record to update our tickets.
Imagine a scenario where our Captured Pokémon have a linked Gym Owner which is the specialist for this Pokémon.
We can now create a trigger that adds the Gym Owner as a follower whenever the tag `specialist_needed` is added to a ticket. This is done by leveraging the nested data in a Custom Object.
In the action step choose *Add Follower* as a step and select your *Captured Pokémon* ticket field > *Specialist* as the value.
Whenever we now add the tag `specialist_needed` to the ticket (e.g via a macro) we'll add the Specialist linked to that specific Pokémon as a Follower to the ticket.
💡
This flow is useful to have e.g. a responsible for a product, a vendor, a third party supplier or manager pulled into the conversation based not the linked record type. It works for all Lookup Fields linked against tickets.
Sadly, it does not yet work for objects linked against the ticket requester.
# What's next?
So, we've now learned how to create custom objects at the start of this series. We then used the objects to give agents additional context in tickets to help them solve tickets faster. And this article showed how to add additional information to end-user profiles and use that information to change priority or escalate conversations to agents.
To wrap up the series we'll solve a missing feature in Zendesk [in the next article](https://internalnote.com/custom-objects-part-4-end-user-and-forms). We'll show Custom Object data in Help Center forms and make it possible for customers to choose object instead of requiring agents to do so.
### Zendesk Custom Objects - Part 2: Tickets
URL: https://internalnote.com/custom-objects-part2-tickets/
Last updated: 2025-09-08T06:45:11.000Z
In this four part series we'll explore the new Custom Objects from an admin, agent and end-user standpoint.
#### Custom Object Series
1. [Introduction to Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/)
2. [Custom Objects and Tickets](https://internalnote.com/custom-objects-part2-tickets/)
3. [Custom Objects and Users](https://internalnote.com/custom-objects-part-3-users/)
4. ➕ [Making Custom Objects available to End-Users](https://internalnote.com/custom-objects-part-4-end-user-and-forms)
The [first article](https://internalnote.com/custom-objects-part-1-introduction/) showed you how to set up custom objects, how to structure and link the data, and how to import data into Zendesk.
This article will show you how to use Custom Objects in forms, how to show the data to Agents, and will include a custom sidebar app that will show the records in a more visual way.
# Custom Objects for Agents
Agents interact with customers via tickets in the Agent Workspace. The Agent Workspace offers a unified inbox off all active conversations, and provides context with the Intelligence Panel and Custom Fields in Forms.
With Custom objects we can expand those traditional forms and fields with more rich objects. Instead of a customer referencing a specific product or asset in a field, we can have that field be linked to an actual Custom Object record and show the agent all related data like it's vendor, location, purchase date e.a.
Similar to how an agent can see a customers' profile via the [Essentials Card](https://internalnote.com/essentials-card/), we can use the new Record Preview Feature to get a *preview* of a linked record right next to a ticket. Or, if a record needs to be modified we can give agents edit rights to records to they can update a phone number, name or description.
In this article we'll expand our Pokédex Object and allow agents to link Tickets to specific Pokémon, and check their stats.
# Showing custom objects in a form
The first step to link Custom Objects to Tickets is by creating so called Lookup Fields. We've already used them in the previous article to link `types` to `pokemon` but this time we'll use them to link `tickets` to `pokemon`.
Lookup fields in forms work similar to drop downs. They offer agents list of options to choose and they are stored in the ticket metadata for future retrieval and/or reporting.
## Adding a Lookup field
To create a Lookup field go to A*dmin panel > Objects and rules > Tickets > Fields* and add a new Custom Field. Choose *Lookup Relationship* as its type.
Give the object a clear name, e.g. "Choose a Pokémon" and link it to the `Pokémon` object we created in the previous article. You can ignore the filter for now.
Once saved, go to A*dmin panel > Objects and rules > Tickets > Forms* and add the Lookup field to a new or existing form. For this demo I created a Pokécenter Form which includes the Lookup Field, a dropdown with a few options and the default Priority and Type fields.
## Viewing Objects when interacting with Tickets.
Now that we have added our Lookup Field to the form, we can have agents interact with that field.
Take a look at the flow below. A customer has asked a question about a Pokémon. Our agent first uses the Lookup Field to search for the Pokémon in question and adds him to the ticket.
You'll notice that once they add the Pokémon, a new icon appears next to the ticket field. Clicking that icon will open the new *Record Preview*. Since the customers' question is related to the `Pokémon type`, the agent can click on the linked `type` and check out its data easily and provide an answer to the customer.
## Viewing linked objects
It might be useful in some flows to not only be able to see what Pokémon is linked to a ticket, but to find all tickets linked to a specific Pokémon.
This can be done by going to the Object Inspector in the navbar, selecting the Pokemon object and searching for a specific one. When you open its record, you'll find the record data on the left, and the right side of the view shows all linked objects. In our case: all the tickets that refer to our Pikachu.
## Permissions
You'll notice that it was the **agent** who added the Pokémon to the ticket, and not the customer filling a form. This is due to a currently limitation in Custom Objects where we can only make custom objects visible, editable or deletable for Agent in your Zendesk environment.
The fourth article in this series will show a work around for this, but out of the box it's the agents doing the work here.
🔐
The reason for this is the potential of leaking company data. Lookup fields can list users, organisations, objects. Imagine creating a lookup fields that contains all your customers and sharing that with your end-users. All of the sudden the names of all your customers are publicly visible. Until Zendesk finds a way to make this risk visible/manageable when setting up Lookup Fields, they have purposefully restricted public access to the objects.
Setting end-users aside, we do need to validate the correct permissions for our other agents before they can interact with the custom objects. This is done by managing the object in the Admin Panel and navigating to the [*Permissions Tab*](https://support.zendesk.com/hc/en-us/articles/6034260247066?ref=internalnote.com). Here you can set View, Edit, Delete or Add rights to the Object.
Your situation will vary, but I've setup this Object with View rights for all users, and only admins have the ability to add, edit or delete records.
Note that you **do not need** edit rights to be able to link tickets and object records! Edit records only apply to changing data within a record. For example: changing Pikachu to a Grass type.
# Create a custom sidebar app
So far we've covered linking records to tickets and seeing the data in the Agent interface. As you noticed in the flow above, the agent can link a Pokémon, view its data in the Record Preview. By clicking the Type in the preview we can then see a Pokémon's type and discover its strength and weaknesses.
But even though this built-in interface is nice, there are things that could be better. Our object has an `image` field that contains a link to an image of the Pokémon. And our agents need to click a few times to discover a Pokémon's strength and weaknesses.
These items aren't showstoppers, but it would be nice to create a more visual representation of the data for our agents.
As you can see, this Agent Workspace contains a custom sidebar app that shows a photo of the linked Pokémon, clearly shows its type in a visual way, and even shows its strength and weakness in one overview.
You can find a full code sample of the app in the repository below.
[GitHub - verschoren/pokedexContribute to verschoren/pokedex development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/pokedex?ref=internalnote.com)
## Getting a linked object
This app lives next to a ticket and makes use of the ZAT Client to read the ticket data of the current ticket.
We first use `client.get('ticket.customField:custom_field_123457890')` to retrieve the value of the Lookup Field. Note that `1234567890` is the ID of the Ticket Field, which can be found via the Admin Panel.
If a Pokémon was linked, this will return the `Record ID` of the linked Custom Object. We then use the Zendesk API to retrieve the entire record of the linked Pokémon via `/api/v2/custom_objects/pokemon/records/{{record_id}}.json`
```javascript
let pokemon = await getPokemon();
async function getPokemon(){
var linked_pokemon = await client.get('ticket.customField:custom_field_14502103878802').then(async function(custom_field){
return custom_field['ticket.customField:custom_field_14502103878802'];
})
if (linked_pokemon == null){
return null;
} else {
return await client.request({
url: '/api/v2/custom_objects/pokemon/records/'+linked_pokemon+'.json',
type: 'GET',
dataType: 'json'
});
}
}
```
```json
//returned data
{
"custom_object_record":
{
"id": "01HD0W1D30KCRZ8PTVZKT86M5P",
"name": "Pikachu",
"custom_object_key": "pokemon",
"custom_object_fields": {
"image": "https://pokedex.verschoren.dev/images/25.png",
"type": "01HD0T5GE1QDVPB97XWTJV5WT2"
},
"external_id": "25",
...
}
}
}
```
Once we have our `pokemon` we also need to load the data of the linked `type` object. Since Zendesk does not (yet) allow to side load linked records, we'll need to make a second API call to retrieve the `pokemon_type` data.
```javascript
let pokemon_type = await getType(pokemon.custom_object_record.custom_object_fields.type);
async function getType(type_id){
return await client.request({
url: '/api/v2/custom_objects/pokemon_type/records/'+type_id+'.json',
type: 'GET',
dataType: 'json'
});
}
```
```json
//returned data
{
"id": "01HD0T5GE1QDVPB97XWTJV5WT2",
"name": "Electric",
"custom_object_key": "pokemon_type",
"custom_object_fields": {
"color": "yellow",
"strength": "01HD0T5GCHTCSFFS3G3AMGVMRH",
"weakness": "01HD0T5GDZ95H63AKDXK7FFQY9"
},
...
}
```
And finally, since we also need the `strength` and `weakness` we can use `getType()` again to load the data for the two linked `pokemon_types`.
```javascript
let weakness = await getType(pokemon_type.custom_object_record.custom_object_fields.weakness);
let strength = await getType(pokemon_type.custom_object_record.custom_object_fields.strength);
```
## Showing the data
As you can see, interacting with Custom Objects via API is not that complex. The same IDs are uses across custom fields and records, making it easy to retrieve records and their data.
Now that we have gathered all data for our linked record, we can use that data to render a nice preview of the record for our agents that include both strengths and weaknesses and a nice photo of the Pokémon.
## Wrapping it up
There's a few final niceties we can add to our app to wrap it up.
First off, we add a link to the Object Inspector so we can view our Pokémon record in the native full screen viewer.
```javascript
$('#view').attr('href',`https://yourdomain.zendesk.com/agent/custom-objects/pokemon/records/${pokemon.custom_object_record.id}`);
```
Secondly, we want our app to update whenever a Pokémon has been added to the ticket, removing the need of a manual refresh by the agent:
```javascript
client.on('ticket.custom_field_14502103878802.changed', function(e) {
init();
});
```
💡
As noted, you can find the entire app in the [GitHub](https://github.com/verschoren/pokedex?ref=internalnote.com) repository. You'll notice that the app already includes a few other features like showing a requesters' Gym Badge and the option to show Pokémon captured by the requester. These are items we'll explain in the next article of this series!
# What's next?
We've now seen how we can have Agents interact with Custom Objects in the Agent Workspace, and how we can use the build in Record Preview or a custom sidebar app to show rich context to agents while working with tickets and Custom Objects.
The [next article](https://internalnote.com/custom-objects-part-3-users/) in this series will dive into expanding user profiles with Custom Objects. We'll explain how you can use them to trigger different priorities based on linked records, and how to show linked objects to users next to their profile and tickets.
## Sign up for Internal Note Plus
The next articles of these series are Internal Note Plus articles. Subscribe today (or make use of our trial!) to keep learning about Custom Objects!
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Preview of the new Generative AI for the Zendesk Help Center
URL: https://internalnote.com/preview-of-the-new-generative-ai-for-knowledge-in-zendesk/
Last updated: 2024-08-19T20:34:24.000Z
Zendesk's [AI Drop Event](https://internalnote.com/zendesk-ai-drop-keynote/) last month included tons of new Zendesk AI features. Their new Generative AI Bot got all the attention and headlines, but in parallel they also announced their new Generative AI for Help Center content.
This new feature allows Content Managers to write short summaries of articles, and use Generative AI to expand the summary or bullet points into long-form support articles. It's powered by [OpenAI](https://support.zendesk.com/hc/en-us/articles/6059285322522-Overview-of-OpenAI-within-Zendesk-Services?ref=internalnote.com), and positioned as a way to create content faster and more efficient by having your writers focus on the essence of an article, and have AI worry about turning it into an actual article.
[Using generative AI to expand help center content (EAP)What’s my plan? Note: The text expansion feature is currently available in an early access program (EAP). You can sign up for the EAP here. The AI-powered text expansion feature helps you to quic…Zendesk helpElizabeth Williams](https://support.zendesk.com/hc/en-us/articles/6267360002714?ref=internalnote.com)
❌
I was planning on writing about today's [Zendesk What's New](https://www.zendesk.com/whats-new/?ref=internalnote.com), but in the end it wouldn't be a very useful article.
Although the presentation was packed with announcements, almost all of them were already known or released in the last few months: Layout Builder, [Agent Home](https://internalnote.com/agent-home-beta/), Omnichannel Routing, [Generative AI Bot](https://internalnote.com/quick-look-at-the-new-generative-ai-bot-for-zendesk/), [Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/) and most of the stuff covered at the AI Drop Event.
The few items that were actually new, like the 360 Employee View app, will be covered in the Zendesk Roundup later this month.
# Let's put it to the test
I gave my [test article](https://support.internalnote.com/hc/en-us/articles/14904161389202?ref=internalnote.com) the following input:
#### Input
Steps to make coffee
1\. Preheat machine
2\. Weigh beans - 18g
3\. Grind Beans
4\. Add to portafilter
5\. Tamp the ground beans
6\. Insert into machine
7\. Pre-infuse
8\. Pull Coffee (30sec -27g)
9\. Enjoy
And clicking the *magic wand* tool in the article editor turned it into this longer article:
#### Expanded Result
Here is a step-by-step guide on how to make coffee: Preheat the coffee machine to ensure that the water reaches the optimal temperature for brewing. Weigh out 18g of coffee beans. This will ensure that you have the right amount of coffee for a balanced flavor. Grind the coffee beans to the desired consistency. The grind size will depend on the brewing method you are using. For example, a fine grind is typically used for espresso. Add the ground coffee to the portafilter, which is the part of the coffee machine where the coffee is placed for brewing. Tamp the ground coffee evenly and firmly using a tamper. This will help to create a level surface and ensure even extraction of flavors during brewing. Insert the portafilter into the coffee machine. Make sure it is securely in place. Pre-infuse the coffee by allowing a small amount of water to saturate the coffee grounds before the full extraction process begins. This helps to release the flavors and aromas of the coffee. Pull the coffee shot by starting the extraction process. The ideal extraction time is around 30 seconds, and you should aim to extract around 27g of coffee. Once the coffee shot is ready, remove the portafilter from the machine and discard the used coffee grounds. Enjoy your freshly brewed coffee! You can drink it as is or add milk, sugar, or any other desired flavorings.
What's evident is that this article is (A) longer and (B) contains a lot more nuance than my, admittedly, very short series of steps. It even included information on the effect of grind size, and explains why a proper tamping of the coffee puck is important. It offers extra nuance to my notes, things that make the article more informative, and trustworthy, but that's all detail that would have taken me time to look up and write.
The generated article also did not deviate from my summary. It contains my steps, and even took over my mistake, namely that a 18g doze of coffee should measure 36-40g coffee when going for a 1:2 or 1:3 ratio. So even though we count on AI to generate an expanded version of the article, I still stayed in control of the actual (wrong) content.
What's also clear is that this article is not at all usable as is. It's one giant block off text, has no line breaks and omits any sense of steps when reading the article. **(Update: fixed, see below)**
But a quick round of editing later turned my summary into this nice article:
#### Final Article
Here is a step-by-step guide on how to make coffee:
1. Preheat the coffee machine to ensure that the water reaches the optimal temperature for brewing.
2. Weigh out 18g of coffee beans. This will ensure that you have the right amount of coffee for a balanced flavor.
3. Grind the coffee beans to the desired consistency. The grind size will depend on the brewing method you are using. For example, a fine grind is typically used for espresso.
4. Add the ground coffee to the portafilter, which is the part of the coffee machine where the coffee is placed for brewing.
5. Tamp the ground coffee evenly and firmly using a tamper. This will help to create a level surface and ensure even extraction of flavors during brewing.
6. Insert the portafilter into the coffee machine. Make sure it is securely in place.
7. Pre-infuse the coffee by allowing a small amount of water to saturate the coffee grounds before the full extraction process begins. This helps to release the flavors and aromas of the coffee.
8. Pull the coffee shot by starting the extraction process. The ideal extraction time is around 30 seconds, and you should aim to extract around 27g of coffee.
9. Once the coffee shot is ready, remove the portafilter from the machine and discard the used coffee grounds.
Enjoy your freshly brewed coffee! You can drink it as is or add milk, sugar, or any other desired flavorings.
# Expand/Tone-Shift
Aside from just expanding text, Zendesk also added[ a tone-shift feature](https://support.zendesk.com/hc/en-us/articles/6353506096410-Announcing-the-addition-of-tone-shift-to-the-Generative-AI-for-Knowledge-EAP?ref=internalnote.com) into the editor, turning text into a more serious, or more playful tone.
By using this feature, similar to the one already available to agents, you can turn this dry closing sentence:
> Enjoy your freshly brewed coffee! You can drink it as is or add milk, sugar, or any other desired flavorings.
Into this more friendly, slightly longer, paragraph.
> Savor the delightful taste of your freshly brewed coffee! Feel free to enjoy it as it is or enhance the experience by adding a splash of milk, a sprinkle of sugar, or any other delicious flavorings that tickle your taste buds.
# My take
I've written before on the absurdity of some of the AI flows. When combining all the AI features we basically now have someone writing a shot summary of an article, that AI expands to a full article on one side.
And on the flip side we have customers asking questions to the Zendesk Bot, where AI turns long articles into a [summarised answer](https://internalnote.com/quick-look-at-the-new-generative-ai-bot-for-zendesk/). The main nuance lies in the fact that the Customer Answer is generated with their intent en context in mind, but surely somewhere in this game of Chinese whispers, something will or can get lost in translation.
That being said, it is a useful feature to have, cause it turns technical summaries into readable articles with nuances and more colourful text, and in times of crisis (when things go offline, when orders are lost) not spending time on writing detailed article for the sake of it, makes time available to actually help customers or dive into the issue.
## A few missing items
While playing around with the EAP I noticed that it handled all input as plaintext. URLS, lists, bold or italic text were all turned into plain text, and even titles and line breaks were gone once AI did its thing. **(Update: fixed, see below)**
I tested both lists of steps like the example above, as well as feeding the AI paragraphs of text (thanks Wikipedia!) but in both scenarios some of the existing structure or markup got lost.
Especially in the case of lists and URLs I hope these feature will get added by the time this EAP turns final release.
# Update (Nov 9th)
The Product Manager of this EAP reached out to me and let me know that the HTML stripping issue is now resolved and articles should now respect lists, urls and other markup when being processed via OpenAI.
It's an EAP for a reason, so pretty cool to see how fast feedback gets implemented and the tool improves.
### Zendesk Custom Objects - Part 1: Introduction
URL: https://internalnote.com/custom-objects-part-1-introduction/
Last updated: 2025-09-08T06:41:54.000Z
A few years ago Zendesk introduced Custom Objects to break out of the classic pattern of ticket - user - organisation.
With custom objects companies could expand the Zendesk CRM with their own assets: printers, orders, contact information, SLA contract,... it promised a lot, but sadly, the APIs and stability didn't really deliver. The Legacy custom objects – as they're now called – always felt like they lived next to Zendesk, as opposed to integrated with the platform.
This all changes with the new Custom Objects. This new release is a completely reworked approach to Custom Objects, that deeply integrates with the Agent Workspace. It has a UI to create object types, a build-in record viewer, and records can be linked to users, organisations, tickets and other records via the [Lookup Fields](https://internalnote.com/lookup-fields-and-ticket-escalation/) Zendesk introduced last year.
# Custom Objects Series
Showing how the New Custom Objects work in a single article would not do them service since they offer so much capabilities right out of the box. So instead of writing one article, I'm doing a series!
In this four part series we'll explore the new Custom Objects from an admin, agent and end-user standpoint.
#### Custom Object Series
1. [Introduction to Custom Objects](https://internalnote.com/custom-objects-part-1-introduction)
2. [Custom Objects and Tickets](https://internalnote.com/custom-objects-part2-tickets/)
3. [Custom Objects and Users](https://internalnote.com/custom-objects-part-3-users/)
4. ➕ [Making Custom Objects available to End-Users](https://internalnote.com/custom-objects-part-4-end-user-and-forms)
This first article will show you how to set up custom objects, how to structure and link the data, and how to import data into Zendesk.
The second article will show you how to use Custom Objects in forms, how to show the data to Agents, and will include a custom sidebar app that will show the records in a more visual way.
The third article will dive into expanding user profiles with Custom Objects and how you can use them to trigger different priorities based on linked records.
The last article will solve a missing feature in Zendesk by showing Custom Objects in Help Center forms and make the selected record available to agents.
# Pokédex
Last spring shortly after the Relate event, I build a quick proof of concept of a Pokédex in Custom Objects. This series will use that concept to explain all the elements related to Custom Objects in Zendesk
[Preview: Creating a Pokédex with Zendesk Custom ObjectsThe awesome people at Zendesk just made the Custom Objects v2 EAP available.So why not build a Pokédex inside Zendesk to get to know the APIs 😉Internal NoteThomas Verschoren](https://internalnote.com/creating-a-pokedex-with-zendesk-custom-objects/)
Why a Pokédex? When building on top of new technology I tend to use a data set or environment I know, so I can focus on exploring the tech, and not worry about thinking of complex data structures, or generate sample data on my own.
Being a father of a 3 and 6 year old, the world of Pokémon was readily available, and there are a lot of online tables, lists and assets ready to copy and use. But even if the world where this article series lives in is not at all related to your business, the concepts in this series can easily be applied to your own business needs.
# Creating Custom Objects
Before starting with Custom Objects it's important to first write down the structure of your data so you can see how the data is organised.
In our case we're building a Pokédex, so we need at least:
- Index, or the number of the Pokémon in the list
- Name
- Image, to make the data a bit more visual
- Type, which element is linked to the Pokémon
Custom Objects have an `external_id` and `name` as default fields. Those nicely map to our `index` and `name` fields.
Since Custom Objects does not support attachments or binary data, we can't just upload an `image`. So we'll use a text field for this value and store the URL of a hosted image.
And finally, we have `Type`. `Type` is a text value, e.g. Electric or Fire, but is a known list of values. There's around 12 different types our Pokémon can have. We might be tempted to use a Dropdown field for this element. This way we can choose one of the preset values, and can filter on those values.
But, `Type` is more complex than just a string of text. In the lore of the games Fire types are weak against water, and strong against grass. So it might be cool if we can somehow store that information too in our Objects.
So instead of creating a dropdown, we'll first create a separate Object for `Types` which contains that related data. We'll then link those two together.
💡
Similarly, if you store products you might have a `Vendor` field attach to them. That vender has an address, name, contact email,... so instead of storing just the name of the Vendor, you could create a `Vendor` object and link it to your `Products` object.
So that's the basic structure we'll need. Two object types, each containing around three fields.
# Creating the Type Object
## Setting up the object
To create an object, go to the *Admin Panel > Objects and Rules > Custom Objects > Objects* and click on *Create Object.*
Give the object a name 'Pokémon Type' and a plural name 'Pokémon Types'. This last one is used whenever we show multiple records of this type. The object key is used whenever we interact via the API. Give it a short name like `pokemon_type`.
💡
I give my object types similar Object Keys whenever they are related. So for this setup all objects will start with `pokemon_` as a prefix so I can more easily find them when using the API.
Next up, we need to add the fields. We already get `name` so we can use that to store e.g. Water. We then add three fields: two lookup fields, and a dropdown.
The dropdown contains a list of colours which we can use in custom apps to improve the way the object looks. Or, we can use the value to reference the colour the Pokémon type has in the game.
Lookup Fields are used to link multiple objects together. In our case we want to add a `strength` and `weakness` field that stores the types the customer is strong or weak. Since these values are `pokemon_types` too, we can add lookup fields that link to `pokemon_types` to reference them. This way we can jump between the elements when looking at them later in the Custom Object inspector.
When adding a Lookup Field you have to select the Related Object, in our case the `Pokémon Type` . You can filter the returned records, but that's something we'll use in a future article. You can ignore it for now.
## Adding records
Now that we have our object type created, we can start adding data. This is done via the Custom Objects Inspector in the Agent Workspace. You can find it via the nabber by clicking on the *Stacked Blocks* icon.
Select the Pokémon Type object and click *Add* to add a new type. You'll notice that you probably need to first create all elements, and then go back to add their Strength and Weakness since most of those referenced types will not yet exist.
[Custom Objects - Pokémon - typesCustom Objects - Pokémon - types.csv1 KBdownload-circle](https://internalnote.com/content/files/2023/10/Custom-Objects---Pok-mon---types.csv "Download")
# Creating the Pokémon Object
Now that we have our Types setup, we can create our main object, the `Pokémon` object record type.
Similar to how we created our `pokemon_type` object, we'll start by creating a new object type and give it a logical name, plural name and key. (I chose `pokemon` for all)
Next, add the fields. Like we discovered in our mapping exercise we can use the existing `name` and `external_id` fields.
We'll rename the External ID though. Pokédex ID has a nicer ring to it. We'll also add one Text Field for our `Image URL` by clicking the *Add Field* button and choosing a text field type. And finally we'll add a Lookup field that references the `Pokémon Type` we created earlier.
## Adding the Pokémon Records.
Last time I looked, the starter Pokédex for Pokémon contained 151 creatures. In total there are over a 1000 of those. Similarly, if you want to use Custom Objects for your own data you probably don't want to enter all records manually like we did with our dozen Pokémon Types earlier.
To solve this Zendesk has added a Data Importer feature that allows you to import CSV files with Object Records into Zendesk.
I've prepared a file for the first 151 Pokémon here:
[Custom Objects - Pokémon - pokemonCustom Objects - Pokémon - pokemon.csv14 KBdownload-circle](https://internalnote.com/content/files/2023/10/Custom-Objects---Pok-mon---pokemon.csv "Download")
The basic import structure is simple. The column names should match the Field Keys off the Fields you created earlier, and obviously the contents should match its type. (So adding a date to a date field, a number to an integer field and so on)
One special case is the *Lookup Field* column. Here Zendesk expects the ID of the element as stored in Zendesk. This ID is generated by Zendesk whenever an object is created and can be found in the URL of the object when looking at it in Agent Workspace.
For my import file I created a separate Sheet in my spreadsheet with list of Types and their corresponding ID and used a `VLOOKUP` to match the type Name with its ID. The resulting CSV file has two columns, `Type_Name` and `Type` where the latter matches the Field Identifier used in my Pokémon type.
💡
Browsing to `https://yourdomain.zendesk.com/api/v2/custom_objects/pokemon_type/records.json` will also show you a list of all records and their IDs.
Once you prepared the import file, you can go to the *Admin Center > Objects and Rules > Tools > Data Importer* to start the Import process. Select a Target Destination (the Pokémon type) and upload your CSV.
The system will validate the file and show the matches column. You can inspect some sample data to see if it all matches. You'll notice that the importer ignored the `type_name` column but nicely mapped my `types` to their correct linked record in the Pokémon Types object.
Press next to import. You'll see a status bar that shows the progress, and once completed you'll see your Pokémon appear in the Object Inspector!
# Viewing Records
Not that we have created our Pokémon object and imported the data, we can use the Inspector to explore our records.
Click the *stacked blocks* icon in the navbar and select the Pokémon object. You'll see a list of records sorted by Creation Date. You can paginate through them, or search via the search bar on top to filter. You can search for both the `name` field (e.g. Mew) or the `External ID` (e.g 151).
When you click on a record you'll see an inspector that shows the field data of the record on the left. The right side of the screen shows all objects that reference this record. We'll explore this in depth in the next article, but when looking at e.g. a Pokémon Type, you'll see all Types that have it as a Strength or Weakness.
# What's Next?
Now that we have our record types and records setup, it's time to integrate them with the Agent Workspace and make them available to agents.
The [next article](https://internalnote.com/custom-objects-part2-tickets/) in this series will show you how to use Custom Objects in Forms, how agents can interact with them, and we'll build a custom sidebar app that makes use of the API to show the data in a nicer way!
[Subscribe now to receive all articles in this series](https://internalnote.com/olus)
### Announcing the Custom Objects series for Zendesk
URL: https://internalnote.com/announcing-the-custom-objects-series/
Last updated: 2025-07-06T17:58:50.000Z
Zendesk Custom Objects have launched last month. It allows you to expand the traditional *ticket - user - organisation* dataset in Zendesk with your own custom data to give agents and end-users more context, automate flows and provide a richer customer experience.
[New updates to Zendesk platform enable businesses to act on customer data in one central workspaceZendesk today announced no-code tools to help businesses bring custom data into its platform and create differentiated customer experiences (CX). Admins can now leverage the in integrated triggers to easily create tailored, efficient workflows. Additionally, agents can leverage end-to-end context to…](https://zendesk.smh.re/Uz2?ref=internalnote.com)
Showing the new Custom Objects in a single article would not do them service since they offer so much capabilities right out of the box. So instead of writing one article, I'm doing a series!
In this four part series we'll explore the new Custom Objects from an admin, agent and end-user standpoint.
# Custom Objects Series
#### Custom Object Series
1. [Introduction to Custom Objects](https://internalnote.com/custom-objects-part-1-introduction/)
2. [Custom Objects and Tickets](https://internalnote.com/custom-objects-part2-tickets/)
3. [Custom Objects and Users](https://internalnote.com/custom-objects-part-3-users/)
4. ➕ [Making Custom Objects available to End-Users](https://internalnote.com/custom-objects-part-4-end-user-and-forms)
This [first article](https://internalnote.com/custom-objects-part-1-introduction/) will show you how to set up custom objects, how to structure and link the data, and how to import data into Zendesk.
The [second article](https://internalnote.com/custom-objects-part2-tickets/) will show you how to use Custom Objects in forms, how to show the data to Agents, and will include a custom sidebar app that will show the records in a more visual way.
The[ third article](https://internalnote.com/custom-objects-part-3-users/) will dive into expanding user profiles with Custom Objects and how you can use them to trigger different priorities based on linked records.
The [last article](https://internalnote.com/custom-objects-part-4-end-user-and-forms) will solve a missing feature in Zendesk by showing Custom Objects in Help Center forms and make the selected record available to agents.
## Subscribe now
Want to follow along and discover how Custom Objects work? Subscribe today and receive the articles right in your inbox. One article each week!
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Zendesk Roundup for October 2023
URL: https://internalnote.com/roundup-for-october-2023/
Last updated: 2023-10-31T06:56:46.000Z
It's only been a few weeks since Zendesk's AI Drop event, and I'm still discovering new features from that release. From a new Zendesk Bot that generates replies, to the new feature that turn a summary into a full article, and improvements for the summary, related tickets and expand feature, the event was packed with awesome new releases for Zendesk Advanced AI
If you want to discover all announcements, as well as read about the new Advanced Security Add-on, take a look at the overview I published right after the event!
[Zendesk AI Keynote - A full overview of the new releasesDiscover all the new Zendesk releases from the AI Drop event on October 5th: Generative AI Bots, Intent management, new privacy and security tools and new AI features to increase agent efficiency.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-ai-drop-keynote/)
The ink on that article wasn't even dry and Zendesk already announced a [What’s New at Zendesk](https://event.zendesk.com/whatsnewatzendesk2023q4emea/loginpromo?partner%5Fcontact=0036[%E2%80%A6]er%5Femail&utm%5Fsource=Premium%20Plus&utm%5Fcampaign=WhatsnewQ423%20!) for November 9th. The event will probably review the recent Custom Objects and Zendesk AI improvements, but I really hope they'll also announce general availability of the new Layout Builder and Agent Home. As usual, if anything new is announced, I'll write an overview right after the event.
Speaking of Custom Objects, in November I'm publishing a four part series on the new Custom Objects. Make sure to subscribe to not miss any article!
#### Custom Object Series
1. Introduction to Custom Objects
2. Custom Objects and Tickets
3. Custom Objects and Users
4. ➕ Making Custom Objects available to End-Users [(Internal Note Plus)](https://internalnote.com/plus)
## Sign up for Internal Note
A blog about Zendesk with a focus on development and the Sunshine platform.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 🎉 New Releases
## 🤖 AI Powered Conversational Experiences
All releases for AI and Messaging have been listed in the aforementioned AI Drop overview, but if you want the summary:
- Generative AI bot that turns articles into summaries based on the customer question
- The ability to edit intent names and assign intents to custom answers
- Assign a persona to your bot and have the answer feel professional, playful or neutral
- Show related tickets in the intelligence panel
- New HR, IT and Financial intent models
Most of these are in EAP, but you can request access and try out Advanced AI via [this link](https://support.zendesk.com/hc/en-us/articles/5608712782362-Using-generative-AI-to-summarize-and-enhance-ticket-comments-EAP-?ref=internalnote.com).
[Preview of the new Generative AI Bot for ZendeskDiscover the all new Generative Zendesk Bot powered by OpenAI in this article!Internal NoteThomas Verschoren](https://internalnote.com/quick-look-at-the-new-generative-ai-bot-for-zendesk/)
## 🔎 Help Center and Self Service
- The Zendesk Knowledge Base [gained](https://support.zendesk.com/hc/en-us/articles/6144535338778?ref=internalnote.com) the ability to generate articles based on a summary via EAP.
- Content Blocks are [now](https://support.zendesk.com/hc/en-us/articles/6012860666906-Announcing-content-blocks-enabled-by-default-in-all-articles?ref=internalnote.com) enabled by default, meaning you can embed reusable blocks of text in articles during creation, without the need to save the article first and manually enable this option.
- The Guide Media Gallery got the ability to (bulk) [delete images](https://support.zendesk.com/hc/en-us/articles/6245286580634-Announcing-bulk-deleting-of-media-in-Guide-media-library?ref=internalnote.com) and [rename images](https://support.zendesk.com/hc/en-us/articles/6281115689626-Announcing-renaming-media-in-Guide-media-library?ref=internalnote.com). This is especially useful for customers with big knowledge bases who migrated from the old 'images are stored within the article' option that existed before. You can now clean up duplicates, and give your assets clear and recognisable names.
## 🧱 Open and Flexible Platform
### Pinned Apps Shortcuts
This is a tiny change, but one that I really like. Agents can [now](https://support.zendesk.com/hc/en-us/articles/6208359261722-Announcing-app-shortcuts-to-help-agents-work-with-tickets?ref=internalnote.com) pin apps in the right sidebar of the Agent Workspace and make them available with the click of a button.
Instead of a long list of apps to scroll through, you can now have your most used apps availably with a colourful icon in the same menu as the intelligence panel, knowledge tab or user essentials view!
I only hope they'll allow Admins to preset these pins via Contextual Workspaces or the new Layout Builder too someday.
Speaking of [Contextual workspaces](https://support.zendesk.com/hc/en-us/articles/6257322050202-Announcing-Knowledge-search-settings-in-contextual-workspaces-?ref=internalnote.com), you can now set a default filter for the Knowledge tab. You can, for example filter the content shown to your support team to the General FAQ section, but give your Finance team a preset view on only the "Payment and Refunds" section of your Help Center. Or, if you have a Spanish and Italian group, you can create a workspace for each and show the content in their respective countries' locales.
### Talk Partner Ticket Events
Zendesk Talk is Zendesk's native solution for integrating calls into your Zendesk environment. Aside from this built-in option, you an also use third party solutions like Aircall, Ring central or Twilio to integrate your existing phone solution into Zendesk.
These solutions already had to option for a nice popup to handle call acceptance, decline and transfer and the integrations could log their metadata into internal notes in tickets.
Now, with these new functionalities, these platforms can show their data in a much richer way with an expandable view that contains duration, recordings and other data.
[Announcing new Talk Partner Edition functionalitiesAnnounced on Rollout starts Rollout ends August 25, 2023 September 19, 2023 September 19, 2023 Talk Partner Edition (TPE) is a collection of APIs that enables developers to connect Computer…Zendesk helpWidson Reis](https://support.zendesk.com/hc/en-us/articles/5968074861466?ref=internalnote.com)
## Webhooks
The new Webhooks features already allows for [user](https://internalnote.com/zendesk-user-events/), [article](https://internalnote.com/webhooks-for-guide/), groups, organisations and community posts alerts, as well as being notified via triggers or automations.
Zendesk has been gradually expanding the availably channels, and added [agent availability Events](https://support.zendesk.com/hc/en-us/articles/4408839108378?ref=internalnote.com) and [Custom Object Triggers](https://support.zendesk.com/hc/en-us/community/posts/6294555096602?ref=internalnote.com) to the list.
> **Why is Zendesk adding Agent Availability Webhooks?**
> Currently, many Zendesk customers and partners use the [Agent Availability APIs ](https://developer.zendesk.com/api-reference/agent-availability/introduction/?ref=internalnote.com)to monitor real-time agent information across Zendesk channels. Being able to subscribe webhooks to agent availability events eliminates the need for that constant manual monitoring, creating a pathway to more efficient and effective event-based applications and integrations. This new functionality also enables you to bypass API rate limits, making it more scalable for larger businesses.
Custom Object Triggers on the other hand allow you to update data in (external) systems based on interactions with Custom Objects in Zendesk by notifying them via a webhook. Whenever you solve a ticket from a customer with a Support Contract stored in Custom Objects, you could for example update their billing in Freshbooks. Or, like you'll read in my [Custom Objects](https://internalnote.com/plus) series, set ticket priority based on linking Custom Objects to user profiles and triggers.
## 🔐 Trust and security
The biggest announcement this month was the new [Advanced Data Privacy and Protection](https://www.zendesk.com/customer-data-privacy-protection/?ref=internalnote.com) add-on. This new addition to the Zendesk line-up expand on existing security platforms.
- The Audit Log that now shows configuration changes, gets expanded with a new **Access Log** that shows all agent interactions with users and tickets.
- **Ticket redaction** of credit cards, can now be expanded to redact dates, SSNs and any other type of metadata (powered by AI)
- Metadata can be **masked** from specific agents so not everyone can see email addresses, birth dates or other PII.
- Advanced **data retention** will allow you to setup complex rules to automatically delete tickets and users to comply with regional privacy legislation.
The Data Retention feature will also be made available in a more basic version to (all) Zendesk Suite users at a later date, but no specifics are available yet. Maybe in next week's What's New?
### Macro Suggestion Permissions
Enterprise customers who have access to custom roles can now make [macro suggestions](https://support.zendesk.com/hc/en-us/articles/6272150912410-Announcing-a-permissions-update-for-macro-suggestions-for-admins-Enterprise-only-?ref=internalnote.com) available to some or all administrators in their instance.
# 💡Insights
Two great support articles from Zendesk this time.
The first one offers an insight to all available JWT types in Zendesk, used to authenticate your end-users on [Messaging](https://internalnote.com/jwt-messaging/) or Guide.
[What are the different types of JWT proposed by Zendesk?Question I’m having trouble understanding all the JWT options proposed by Zendesk. What is each option used for? Answer Zendesk offers three different types of JWT, that each have a distinct purpos…Zendesk helpRemi](https://support.zendesk.com/hc/en-us/articles/6219127152794?ref=internalnote.com)
On the flip side, a less technical article that shows the operational differences between tickets, chat and messaging:
[Conversational styles in messagingMessaging, unlike live chat, allows you to have persistent conversations with your customers. One of the main benefits of this is that these conversations may adopt different conversational styles,…Zendesk helpRob Stack](https://support.zendesk.com/hc/en-us/articles/6088892450586?ref=internalnote.com)
# 🎥 Videos
#
# ⚠ Major Changes
## Twitter
Byebye birdie... here's the X Corp.
> Zendesk, therefore, is aligning with Twitter's rebrand from Twitter to X Corp. We understand the importance of staying in alignment with our partners' branding guidelines, and this update is a reflection of our dedication to excellence.
## Automatic activation of custom statuses
[Custom Statuses](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/), announced last year, will now be [enabled](https://support.zendesk.com/hc/en-us/articles/6203843149210-Announcing-automatic-activation-of-custom-ticket-statuses?ref=internalnote.com) by default for all users. This means your existing statuses will be converted to Status Categories, and you'll get the option to create variants for new, open, pending, on hold and solved.
If you don't use Custom Statuses, **you won't notice a change**. Zendesk will automatically update your triggers and reporting, and everything will keep working as before, but if you want, you can start adding custom statuses now, or after the feature gets auto-enabled.
# 📝 Articles this month
[Preview of the new Generative AI Bot for ZendeskDiscover the all new Generative Zendesk Bot powered by OpenAI in this article!Internal NoteThomas Verschoren](https://internalnote.com/quick-look-at-the-new-generative-ai-bot-for-zendesk/)
[➕ Automatically link incidents and problems in ZendeskWhen a crisis hits agents should focus on resolving problems and informing customers, not waste time on assigning and categorising incident tickets. This article shows you how to automate this flow!Internal NoteThomas Verschoren](https://internalnote.com/automatically-link-incidents-and-problems-in-zendesk/)
[Zendesk AI Keynote - A full overview of the new releasesDiscover all the new Zendesk releases from the AI Drop event on October 5th: Generative AI Bots, Intent management, new privacy and security tools and new AI features to increase agent efficiency.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-ai-drop-keynote/)
[➕How to build a VIP Alert app in Zendesk from scratchThis article contains a full tutorial on how to build a small Zendesk Support app that alerts your agents whenever they’re looking at a VIP ticket.Internal NoteThomas Verschoren](https://internalnote.com/vip-alert-app/)
# And finally...
Who says working with Zendesk can't be fun? If you want to find a good excuse to test the Zendesk Bot and Messaging on Mobile, you can now download a fun demo game built with the Unity SDK that can be linked to your Zendesk instance to test Bots and Messaging right on your phone!
[Zendesk SDK for Unity now has a demo gameAnnounced on Rollout starts October 17, 2023 October 17, 2023 We are excited to announce a demo game for you to see the Zendesk SDK for Unity in a real-life game and to be able to test your ow…Zendesk helpGofran Shakair](https://support.zendesk.com/hc/en-us/articles/6249260905754?ref=internalnote.com)
If you want to test it out yourself:
1. Install the app on your iPhone or Android Phone
2. Go to the Admin Panel > Channels > Messaging and add a new Channel
3. Choose the Android SDK and give it a useful name, like Unity or Demo Game
4. Go to the Installation Tab and copy the Channel ID key.
5. Paste the key in the configuration screen of the Demo App.
6. Play the game and/or test out your Bot.
### How to build a VIP Alert app in Zendesk from scratch
URL: https://internalnote.com/vip-alert-app/
Last updated: 2025-08-01T12:06:30.000Z
This article was inspired by a Community Post where a Zendesk customer asked about the best way to alert agents when a ticket requester is a VIP user.
People suggested tweaking the organisation name with an alert (🚨🚨🚨 Acme.Inc), change the subject to start with the word VIP via a trigger, dedicated views,.. there were many suggested solutions.
[Identifying a VIP Organization On The TicketWe have a need that I can’t imagine is unique to us: We need our agents to be able to quickly identify that an incoming ticket is from a VIP customer. Tagging the VIP orgnizations is an easy thing…Zendesk helpJay McCormack](https://support.zendesk.com/hc/en-us/community/posts/6137207769498/comments/6181921058074?ref=internalnote.com)
My suggestion was to leverage SLA by giving VIP users a higher priority or use the new [Essentials Card](https://internalnote.com/essentials-card/) ti highlight the VIP tag. A more advanced solution was building a dedicated app in Zendesk for this:
> ... and finally I would write a simple Zendesk app that looks at the current user and if they contain the tag VIP we show a popup top right.
>
> I wrote a quick version of the app, available here as a private app. You can find the latest version in the TMP folder.
> [https://github.com/verschoren/vip\_alerts](https://github.com/verschoren/vip%5Falerts?ref=internalnote.com)
My comment linked directly to the app, this article will explain how to build it.
You can download the application via this GitHub repository, or you can keep reading and build your own! (I would live to publish these apps directly on the marketplace as free apps, but my current employer contract limits this for understandable reasons)
[GitHub - verschoren/vip\_alertsContribute to verschoren/vip\_alerts development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/vip%5Falerts?ref=internalnote.com)
➕
This article is exclusive to [Internal Note Plus](https://internalnote.com/plus) subscribers.
# Requirements
Zendesk apps are written in HTML and Javascript and can be build and packaged via the ZCLI. To get started, you [need to install ZCLI](https://developer.zendesk.com/documentation/apps/getting-started/using-zcli/?ref=internalnote.com).
Once done, create a new folder for this project on your computer and open the folder in your terminal and run:
```bash
zcli apps:new
```
The script will ask for some information:
# The Code
After running the script you'll end up with these files:
## manifest.json
Our app can only be installed if we have a valid Manifest file in the bundle.
There's a few things we need to add to the default `manifest.json` file.
### Autohide
We need to add an `"autoHide": true` parameter to hide our app from agents and run it in the background.
```json
"location": {
"support": {
"ticket_sidebar": {
"url": "assets/iframe.html",
"flexible": true,
"autoHide": true
}
}
}
```
Second we need to add a few settings via a parameter array. We'll use these later in the apps' logic.
```json
"parameters": [
{
"name": "vip_tag",
"type": "text",
"required": true,
"default": "vip"
},
{
"name": "message",
"type": "text",
"required": true,
"default": "This user is a VIP!"
},
{
"name": "sticky",
"type": "checkbox"
}
]
```
## assets/iframe.html
As defined in the `manifest.json` file our will be a background app and won't be visible for agents, except for notifications, so the UI of the app will be non-existent and this file will be mostly empty.
Replace the content of `iframe.html` with the code below. Note that we link to `js/main.js` to store our logic.
```html
Oops
{{#is error 'unauthorized'}}
{{link 'sign_in'}}
{{/is}}
{{#is error 'forbidden'}}
{{t 'not_authorized'}}
{{/is}}
{{#is error 'not_found'}}
{{t 'nonexistent_page'}}
{{t 'mistyped_address_or_moved_page'}}
{{/is}}
{{search scoped=settings.scoped_kb_search submit=false}}
{{#if promoted_articles}}
{{t 'promoted_articles'}}
{{#each promoted_articles}}
{{title}}
{{/each}}
{{/if}}
```
This does however not resolve the main issue: if a user clicks on a link for article X, and article X got deleted or replace with article Y, they still reach dead end without context or an easy way to find the actual content.
# Redirect Rules
Luckily, this issue is now resolved 🥳.
Zendesk Guide now has an API endpoint for redirect rules. In short this allows you to send people who open article X to article Y automatically.
When is this useful?
- You have an existing popular article that got replace by a newer article.
- You have deprecated or replaced a service and want to redirect visitors of those articles to the page explaining the removal of the service
- You have released a brand new version and want to send all visitors of the old product to the new product documentation instead
- You merged to articles and want visitors of one article to be send to another
## How does it work?
Currently the API endpoint is fairly limited, but does offer the basics we need. We can redirect articles, sections and community topics to any other resource.
Examples of help center URLs that can be redirected:
```
/hc/en-us/articles/1138
/hc/en-us/community/topics/1977
/hc/en-us/sections/1984
```
But you can also redirect — so called — vanity urls. This is useful if you want to prevent dead links after a migration to Zendesk, and replicate your legacy faq paths, or want nice urls to redirect to a specific article.
```
/faq/how-to-update
/help/1138
/announcements
```
You can find the full documentation in [this Help Center article](https://developer.zendesk.com/api-reference/help%5Fcenter/help-center-api/redirect%5Frules/?ref=internalnote.com) but in essence you can set a `from`, `to` and `status` value for each redirect you configure, and optionally a `brand` value for environments with multiple brands configured.
The API supports the following HTTP Statuses, from which 301 and 302 will probably be the most used ones.
| HTTP Status | Meaning |
| ----------- | ----------------------------------------------------------------------------------------------- |
| 301 | Moved Permanently - The requested resource has been permanently moved to a new location. |
| 302 | Found - The requested resource has been temporarily moved to a different location. |
| 303 | See Other - The response to the request can be found under a different URL. |
| 307 | Temporary Redirect - The requested resource has been temporarily moved to a different location. |
| 308 | Permanent Redirect - The requested resource has been permanently moved to a different location. |
## Creating or Updating a Redirect Rule
To create a Redirect Rule you need to `POST` to `/api/v2/guide/redirect_rules`.
```json
{
"redirect_rule": {
"redirect_from": "/hc/en-us/articles/123",
"redirect_status": 301,
"redirect_to": "https://support.example.com/hc/en-us/456"
}
}
```
The `redirect_from` is the URL your customers are visiting. It can be either an article `/hc/en-us/articles/1138`, community topic `/hc/en-us/community/topics/1977` or help center section `/hc/en-us/sections/1984`.
💡
Note that the URL only accepts the ID of an article. If the URL contains a slug (e.g (1234-article-title) you should only use 1234.
The `redirect_to` can be any URL as long as it is a HTTP(s) url or starts with a `/` . The latter will use the Help Center base URL as a prefix.
Note that if there already is a rule with the same `redirect_from` URL that the existing one will be overwritten and your `POST` will act similar as an update.
Similarly, to remove a rule, you should first list all rules via a `GET` command, then retrieve the `ID` of the rule you want to delete, and then to a `DELETE /api/v2/guide/redirect_rules/{redirect_rule_id}`. (Remember that deleting a 301 permanent redirect can do weird things with Google indexing)
## Example
I've use to have an article that explained how to subscribe to this blog at an path [https://support.internalnote.com/hc/en-us/articles/12677034076434](https://support.internalnote.com/hc/en-us/articles/12677034076434?ref=internalnote.com)
That article has since been migrated to a native page on the blog, but I reference that article in multiple places. So by adding a redirect rule with the following payload, I can make sure people end up on the right page.
```json
{
"redirect_rule": {
"redirect_from": "/hc/en-us/articles/12677034076434",
"redirect_status": 301,
"redirect_to": "https://internalnote.com/#/portal/signup/63f0d0f4034c3d004d5ef75a/yearly"
}
}
```
Which means that, whenever someone visits [the article](https://support.internalnote.com/hc/en-us/articles/12677034076434?ref=internalnote.com), they will get redirected to the [signup page]({ "redirect%5Frule": { "redirect%5Ffrom": "/hc/en-us/articles/12677034076434", "redirect%5Fstatus": 301, "redirect%5Fto": "https://internalnote.com/#/portal/signup/63ff05d99deb9e003d619e70/yearly" } }) for my blog.
> Which reminds me: did you subscribe yet? It's free, or optionally paid, and really helps this website to grow!
# Advanced Redirects and wish list
Another fun setup is something like the following
```json
{
"redirect_rule": {
"redirect_from": "/jurassicpark",
"redirect_status": 302,
"redirect_to": "https://support.internalnote.com/hc/en-us/search?utf8=✓&query=jurassic+park"
}
}
```
This rule gives you a shorthand for searching the Help Center for specific articles.
But what I'd really love to see is a way to have regex support for the redirect rules like e.g. [Next.js has](https://nextjs.org/docs/pages/api-reference/next-config-js/redirects?ref=internalnote.com#regex-path-matching) so we can use one rule to redirect a bunch of matching articles.
But for now, if you've recently migrated from a competitor and you want to bulk redirect all old Freshdesk URLs like `https://support.domain.com/en/support/solutions/folders/7000008400` to matching articles on Zendesk, this API certainly comes in handy, even though it’s going to require a lot of API calls to setup each redirect one by one.
As a last feature request, it would be nice to have some reporting on this so we know which redirects are (un)used for SEO optimization.
## Conclusion
I love how Zendesk keeps improving the Help Center website experience even though the focus of most customercare is moving to chatbots and conversational design.
But (seemingly) simple things like adding redirects makes Guide so much more powerful.
What will you use it for?
## Sign up for Internal Note
Turning Zendesk into practice. – A newsletter about Zendesk written by Thomas Verschoren.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
### Pre-filling custom fields in the Zendesk Bot with Messaging Metadata
URL: https://internalnote.com/messaging-metadata/
Last updated: 2025-07-21T10:01:55.000Z
Last week's article [showed](https://internalnote.com/show-restricted-articles-in-zendesk-bots/) how Zendesk now allows you to show restricted Help Center articles in Guide, removing one of the final missing items for feature parity between the Classic Widget and Messaging from the list.
Less than a week later they released another update to Messaging that resolves maybe the biggest open item on that list:
[Announcement: Introduction to Messaging MetadataIntroduction We are very excited to announce the introduction of Metadata APIs to be used with our flagship Messaging product. This capability will allow our customers to send metadata relating to…Zendesk helpMick O’Donnell](https://support.zendesk.com/hc/en-us/articles/5868905484442?ref=internalnote.com)
In short, it is now possible to fill in custom fields in the Messaging flows via a bit of code and preselecting dropdown, pre-filling text fields or checking checkboxes without user interaction.
This not only enables now flows in your Conversational Bot, but also removes manual actions that used to be done by your users: you can enter order numbers, select a venue, tag conversations with VIP, or set values you can use in API calls (retrieve an order) or branched conditions in your flows.
This is all done with one fairly simple line of code:
```javascript
zE('messenger:set', 'conversationFields', [{ id: 'id', value: 'string'}]);
```
Let's dive in!
# Pre-filling Ticket Fields
To demo this new feature I've build three demo environments:
- [Movie Picker](https://demo.internalnote.com/prefill?ref=internalnote.com) \- a demo on how to pre-fill a text field and use the input to make an API call.
- [Star Tours](https://demo.internalnote.com/startours?ref=internalnote.com) \- a demo on how to choose a dropdown option and handle the choice via conditional branching.
- [VIP Flow](https://demo.internalnote.com/proactive-contact.html?ref=internalnote.com#vip) \- set a VIP tag whenever a VIP customer is logged in and contacts you via the bot.
## 🥳 Thanks for reading Internal Note
If you like this kind of content, please consider ****subscribing** via email or ****share** the article to your colleagues.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# Demo Flow 1 - Pre-fill a text field
This demo builds on top of the [Ask for Details](https://internalnote.com/flow-builder-ask-for-details/) article that I published earlier this year. It runs as follows:
- A user enters a movie in the bot flow
- We make an API call to retrieve info about the movie
- We show the returned info (title, director, release year) as a result.
Now that we can pre-fill data in the Messaging Flow we can make this flow a bit nicer. We can e.g. show a list of movies on our website, and whenever the customer selects a movie we fill in the movie field with its title, and open the widget.
## Setup
The first item you need for this flow is a ticket field to capture the Movie Title. Note down the `Field ID` which is part of the URL while editing the ticket field in the Admin Center: [https://domain.zendesk.com/admin/objects-rules/tickets/ticket-fields/**7662882404114**](https://d3v-verschoren.zendesk.com/admin/objects-rules/tickets/ticket-fields/7662882404114?ref=internalnote.com)
## Bot
To see how to build this bot, please refer to this earlier article.
[Flow Builder - Ask for detailsThe new Ask For Details option in Flow Builder allows you to pull in contextual information via API into your Zendesk Chat Bot.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-ask-for-details/)
One extra I added to make the flow easier to discover is create a proactive message that preselects the **🎬 Movie Info** intent in the Admin Center. This is an optional step but makes it easier for customers to discover the demo.
However, if we pre-fill the fields they will also show up as filled in without selecting an intent in advance.
## Website
The final part is the actual new part of this release.
We're going to pre-fill our custom field with the value of a selected movie.
[https://demo.internalnote.com/prefill](https://demo.internalnote.com/prefill?ref=internalnote.com)
The first step is creating an array of movie posters to click on. Note that we add a `data-target` value to each poster so we know which poster was clicked on.
```html
```
Next we react to a click on a `.movie` and set the custom field with the `Field ID` we copied earlier to the `data-target` of the selected item.
And finally we open the widget to start the conversation.
```javascript
$(document).ready(function() {
$('.movie').click(function(){
var movie = $(this).data('target');
zE('messenger:set', 'conversationFields', [
{ id: '7662882404114', value: movie}
]);
zE('messenger', 'open');
})
});
```
Note that even if we preselect a movie, the customer will still be able to see and change the value.
## Use Cases
You can adapt this flow in a myriad of ways:
- Whenever a customer is looking at an order in their order history you can fill in the Order Number field in your bot flows so that any conversation starts references that order
- If a customer is looking at a product page you can fill-in the SKU number so you know what item they're interested in
- Since we can not yet pass user metadata other than email or name, you can fill in a customer id or membership number field with the value of their profile
- If a customer needs help with changing a booking you can launch the widget, pass the conversation to an agent and already pre-fill in their booking number.
# Demo Flow 2 - Pre-fill a dropdown
This next demo doesn't use a text field but makes use of drop-down with a preset list of options.
In this demo we show 3 holiday locations in a galaxy far, far away. When the user selects a destination we pre-fill the location in the messaging flow and when the customers interacts with the bot we'll use this value to branch to the right flow.
## Setup
Step one is creating a dropdown with a list of locations in the Admin Center. Note down the `Field ID.`
.
## Bot
Next up we create a new intent for our bot. For a full breakdown on how to build complex Conversational Bot flows, take a look at this article:
[Learn how to build a full-featured Flow Builder Bot for Zendesk.In this article we will build a full-featured Flow Builder Bot for Zendesk. We’ll use every step type, use API calls and variables and show you how to create a bot yourself in a full length video tutorial.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-dinosaurs/)
Our intent functions as follows:
1. We add an *Ask for Details* step where we ask to select a location by adding the `Holiday Location` field to the Ask for Details step. We'll call this: Pick a Location.
2. We add a *Branch by Condition* step with 3 branches.
1. `Pick a location is Hoth`
2. `Pick a location is Tatooine`
3. Else (the last option)
3. For each branch we add some text, an image and a *Was this helpful?* step
And finally we add a new proactive flow. This is optional, but by using this flow we can select the *Intent* and immediately jump to it when the page loads.
## Website
The logic for this step is similar to the first demo.
[https://demo.internalnote.com/startours](https://demo.internalnote.com/startours?ref=internalnote.com)
First we create a list of locations. Each location has a `data-target` with a location name that matches the `tag` of the dropdown we created earlier.
```html
```
Next we react to a click event and pre-fill our field by setting the `id` to the `Field ID` and the value to the chosen location.
```javascript
$(document).ready(function() {
$('.location').click(function(){
var location = $(this).data('target');
zE('messenger:set', 'conversationFields', [
{ id: '12367640319506', value: location}
]);
zE('messenger', 'open');
});
});
```
## Use Cases
Based on the above there's a few scenario's where this could be useful:
- You're a hotel chain with multiple locations. If a user looks at a specific hotel you can preset the *Hotel* dropdown so you can route the requests to the right team
- You've got a request type field that's used to route between Support issues and Sales inquiries. Based on the page the customer is looking at (Support page or Product page) you can preselect the request type.
- For logged in customers you can set the "Are you already customer" field to yes, and no for others.
## Demo 3: Setting tags
You can set Tags. These tags will only be applied to the created ticket once a conversation has started, and cannot (yet) be used on conditions within the bot flow.
```javascript
zE("messenger:set", "conversationTags", ["vip", "gold"])
```
For example, on [this](https://proactive.internalnote.com/contact.html?ref=internalnote.com#vip) demo page whenever our VIP logs in, we set a tag `VIP` to the conversation, and we wipe the tag for all other users:
```javascript
if (vip) {
Login({
name: 'Vito Corleone',
email:'vito+'+random+'@corleone.example',
external_id: random
});
zE("messenger:set", "conversationTags", ["vip"])
} else {
Login({
name: 'Maximus Decimus Meridius',
email:'maximus+'+random+'@example.com',
external_id: random
});
zE("messenger:set", "conversationTags", [])
}
```
# Final notes on Messaging Metadata
There's a few caveats when using the new feature.
Fields are always visible to the end-user and are you can't make them read only like you could with the [Classic Widget](https://developer.zendesk.com/api-reference/widget/settings/?ref=internalnote.com#fields).
One final "issue" is the fact that once the customer has started a conversation you can't update the values anymore.
Also, you can set multiple fields at once by adding more items to the array
```javascript
zE('messenger:set', 'conversationFields', [
{ id: 1234, value: 'text_string'},
{ id: 5678, value: true},
{ id: 1977, value: 'dropdown_option'},
{ id: 1138, value: 42}
]);
```
To wrap this up, this release is one that's been on my [Wishlist](https://internalnote.com/zendesk-messaging-feedback/) for a long time and makes more complex flows possible. Are there things I still want? Sure, hiding pre-filled fields, making them read only, selecting an intent via the same API, ... are just a few of them.
What do you think of this new feature? Which flows does it unlock for you?
### Show restricted articles in Zendesk Bots
URL: https://internalnote.com/show-restricted-articles-in-zendesk-bots/
Last updated: 2024-08-19T20:39:20.000Z
One of the core tenants of a good customer care setup is a way to allow customers to self service. This not only deflects tickets and lowers the workload of your agents, but also allows customers to resolve issues faster, leading to a higher CSAT.
For this reason the more Help Center articles are made publicly available to all of your customers (and Google..), the better your self service approach will work. Customers will search your Help Center, interact with your Bot, or get auto replies with these articles as your first line of defence.
But sometimes you want articles to be available to only a subset of users. Maybe you use Zendesk for an Internal Help Desk, and you want your IT guidelines only visible to your employees. Or you are a hotel and have special amenities only available to your VIP guests. Or you have support articles that should only be visible to customers who buy a specific product or support SLA.
This is where user segments and restricted articles in Zendesk Guide come in. You can create segments of users based on tags or organisations, and then make articles only available to those segments. These articles are then invisible to guest users, or logged in users outside of the segment, but will be searchable and readable to logged in users that do belong to the segment.
💡
Did you know you can make (parts) of your restricted articles visible to everyone to improve discovery? [Read the article](https://internalnote.com/zendesk-guide-membersonly/)!
Up till now this would not work for conversations started via Zendesk Messaging on the Web. The Zendesk Bot did not care about segments, and even if you authenticated your user, it would only show publicly available articles. But this has now changed!
[Announcing new updates for Zendesk bots that improves employee experiencesAnnounced on Rollout starts Rollout ends July 4, 2023 July 3, 2023 July 6, 2022 We are excited to announce the following releases for Zendesk bots Conversation bots can now be deployed on Sl…Zendesk helpLisa Tam](https://support.zendesk.com/hc/en-us/articles/5879606612506?ref=internalnote.com#h%5F01H45EYPVJ3FK55SNRAV8FVAZ9)
Any user that interacts with your Bot within the Zendesk Web Widget can now search for and see restricted articles, provided they are authenticated. Flow Builder even allows you to add restricted articles as part of the Show Articles step in your flows.
When an anonymous user uses the widget restricted articles will not be part of the suggested articles via search. For authenticated users they will appear if they match their segment.
Similarly, if you add a restricted article to a "Show Articles step", they'll show up as a generic card without exposing any titles or content. For authenticated users the system will show the restricted articles' title and a snippet of the content.
If you want to you can jump into the Bots' settings and make the metadata of suggested restricted articles visible to everyone regardless of logged in status. But watch out, the snippet might expose article content you don't want to be visible to the world!
## **What do you need to get this to work?**
Since you are interested in using this feature in Messaging we can assume you already have segments on your Help Center, allow users to login and restrict some articles to those segments.
So, to get started with restricted articles in Zendesk Bot, you basically only need to enable Authentication for Messaging and you're done. Luckily, I've got an entire article on this topic:
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging/)
# **Example Flow**
As explained in the overview above, by default Zendesk will show restricted articles as placeholders to unauthenticated users.
In this article we will build a bot flow that will **only** show restricted articles to logged in users if applicable, and will fall back to only show public articles to everyone else. A bit more complex of a setup.
You can test it out by going to our demo page.
[Internal Note - Zendesk Messaging AuthenticationDemo page to showcase the JWT Authentication for Zendesk MessagingZendesk Messaging AuthenticationInternal Note](https://jwt.internalnote.com/?ref=internalnote.com)
If you log in as `vito@corleone.example` you'll get the VIP experience. If you log in as `john@example.com` you'll get the regular flow.
# Bot Setup
We'll build a dedicated bot flow to show this new functionality. It flows as follows:
## 🥳 Thanks for reading Internal Note
If you like this kind of content, please consider ****subscribing** via email or ****share** the article to your colleagues.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
1. We check via `conditions` if the user is authenticated or not.
2. If they are authenticated we use an `API Step` to retrieve their data from the Zendesk instance.
3. We check the returned user data for a `VIP tag` and if they are a VIP we show the segmented article to them.
4. In all other cases we show a more generic reply with public articles.
## Assumptions
1. You have a working Messaging Authentication flow.
2. You have created a [User Field](https://support.zendesk.com/hc/en-us/articles/4408822051866-Adding-custom-fields-to-users?ref=internalnote.com) of type `checkbox` that sets a `vip` tag to a user
3. You have created a [User Segment](https://support.zendesk.com/hc/en-us/articles/4408837707290-Creating-user-segments-for-Guide-user-permissions?ref=internalnote.com) in Guide that contains all users with a `vip` tag
4. You have an article that is restricted to that [segment](https://support.zendesk.com/hc/en-us/articles/4408824005914?ref=internalnote.com).
## Check if the user is authenticated
Since restricted articles only show up for logged in users if they match the segment applied to the ticket, being logged in is a first requirement for showing these articles and displaying a title and snippet.
All you need to do is use a `Branch Condition` step and check for `Authentication status is true` in the conditions.
## Retrieve customer data via API
Now that we know if the user is logged in, we need to check if they are a VIP user or not. In our scenario we created Guide segment for users that are tagged with `vip`. This tag is set by a User Field of type checkbox called VIP.
Since Zendesk does not allow to add any tags to the authentication payload, and since we can only use email and name as native variables inside Bot Builder, we need to retrieve this data from Zendesk itself via the API.
We use a `Make API call` step to call the Zendesk search API to search for users that match our authenticated users' email address. It's not the most elegant way, but it's what's available for now.
### Connections
Before you can make an API call you need to go to *Admin Center > API and Integrations > Connections* and add a **two** secure connection.
1. **zendesklocal**
Use `Basic Authentication` as the type and add an admin email address followed by `/token` and a Zendesk API token.
Add `https://*.zendesk.com` as the allowed domain
2. **zendeskwidget**
Use `Basic Authentication` as the type and add an admin email address followed by `/token` and a Zendesk API token.
Add `https://*.yourhelpcenterdomain.com` as the allowed domain
💡
We need to run a test API call in the next step to set the required variables. Since the test runs from your Zendesk admin panel, and the widget will run from your Help Center we need to (temporarily) create two connections. One for setup, one for production.
### API Call
Once we have the connections setup we can add the `API call` step to our bot flow. Set it as a step in the `Is Authenticated` branch
1. Endpoint URL
Set it to [https://d3v-verschoren.zendesk.com/api/v2/search?query=type%3Auser%20email%3A](https://d3v-verschoren.zendesk.com/api/v2/search?query=type%3Auser%20email%3A{{system.user.email}}&ref=internalnote.com) followed by pressing the {+} button and add the *Messaging Metadata > provided email* as a variable
2. Authentication
Choose the `zendesklocal` authentication method
Next press Make API Call to do a testrun. The system will fetch a random user. Scroll down the *Variables* list and expand until you see the `user_fields` and click on the *Save button* next to the VIP field to store it as a variable called `vip`. In our example, since it's a checkbox it has either a true or false value.
3\. One final step before wrapping up and saving this step is to swap the API Authentication method from `zendesklocal` to `zendeskwidget`.
Even though we can assume the user exists (they can log in) they might not yet have any interaction with Zendesk yet, so there might not be a user profile. So this API call can fail.
## Check for VIP Tag
Now that we have an authenticated user and we have stored their VIP status in a variable we can **finally** check if our user is a VIP user.
We'll use another `Branch by Condition` step for this one. Add the step in the `API Call Successful` and set the IF condition to `vip = true`
## Show Segmented articles
Now that we know we have an authenticated user that is a VIP we can safely show them restricted articles for their section and we can be sure that the user will see the title, snippet and will be able to see the entire ticket if they click the link.
1. Add a `Show Help Center Articles` step in the is VIP branch
2. Search through your list of article and select the (restricted) article you want to show
💡
For all other branches (Not authenticated, not found, not a VIP) you can create other flows. Since all three should probably get very similar flows, you can create a flow for one branch and copy it to all other branches.
This can be done by right clicking the top step of a branch, and choose "Copy this step and all below". Right click on an empty branch and choose "Paste steps"
# Wrap up
I think it's fair to say that *finally* is a word to use in this context. Showing restricted articles has been an existing feature in the Classic Widget, and the lack of support in Messaging made it impossible for most Business-to-Employee or Internal use cases to move to Messaging and the new Zendesk Bot.
This new flow elegantly builds on the work done for Zendesk Messaging authentication and works without any complex custom work for most environments. The flow above is a bit more complex by design because I wanted to **only** show restricted articles to users that can access them, but in most other scenario's it will just work.
🫣
There is [still some weirdness](https://internalnote.com/deepdive-into-messaging-profiles/) going on with authentication and mapping messaging authenticated users to existing end users that never interacted with messaging. Until this is fixed (and Zendesk maps based on email addresses) your logged in users might not be part of the segments their email counterparts belong too..
##
##
### Zendesk Roundup for June 2023
URL: https://internalnote.com/zendesk-roundup-for-june/
Last updated: 2023-10-12T14:08:02.000Z
June was Zendesk Showcase month with Zendesk holding a bunch of mini-Relate events across the world. I got invited to the local Showcase in Utrecht where we spend the afternoon watching presentations about all the new releases.
The event opened with a keynotefrom Zendesk EMEA CTO Matthias Göhler going over all the new features launched at Relate. Most of it was already known if you watched the main event in May, but I really liked his focus on our EMEA market and especially the way OpenAI, Privacy and GDPR interact.
Paraphrasing Göhler:
> The Zendesk Suggested Reply feature is in EAP and will remain so until OpenAI is GDPR compliant. So use it at your own risk.
Up next was an awesome product deep dive from one of Zendesk's EMEA consultants. He took everything Zendesk announced in the last year and turned it into a nice story that really showed how these items could impact actual workflows.
To wrap things up we had a round-table with Sarah Al-Hussaini (COO @ [Ultimate](https://ultimate.ai/?ref=internalnote.com)) and a few other Zendesk people and customers on AI and the impact on CX.
And now, the updates for June!
## 🥳 Thanks for reading Internal Note
If you like this kind of content, please consider ****subscribing** via email or ****share** the article to your colleagues.
Subscribe
Email sent! Check your inbox to complete your signup.
No spam. Unsubscribe anytime.
# 🎉 New Releases
## Zendesk Guide Updates
- Zendesk updated their Guide Templates to a new v3 release with a focus on adding improved rich text editing and accessibility features. The Copenhagen theme has been updated automatically, custom themes will get the features once new versions are released. (like the ones my team at [Premium Plus](https://www.zendesk.com/marketplace/themes/?query=premium+plus&ref=internalnote.com) built)
[Announcing Templating API v3 with improved rich text editorAnnounced on Rollout starts Rollout ends June 1, 2023 June 1, 2023 June 8, 2023 We’re releasing v3 of the help center Templating API, which introduces a better rich text editor (RTE) and fix…Zendesk helpGorka Cardona-Lauridsen](https://support.zendesk.com/hc/en-us/articles/5762406613530?ref=internalnote.com)
- You can now configure Guide security settings to allow restricted content to appear in the body of notification emails sent to section followers. This used to be bug where those emails would not show/load images, and this has now been fixed.
[Announcing an update to new article in section e-mail notificationAnnounced on Rollout starts Rollout ends June 22, 2023 June 22, 2023 June 22, 2022 We are excited to announce that you can now configure the way in which article notification emails are sent…Zendesk helpKatarzyna Karpinska](https://support.zendesk.com/hc/en-us/articles/5838337923866-Announcing-an-update-to-new-article-in-section-e-mail-notification?ref=internalnote.com)
## Side Conversations
Side Conversations got a bit of love this month. The new [context panel location](https://support.zendesk.com/hc/en-us/articles/4486696180378-Announcing-side-conversations-in-the-context-panel?ref=internalnote.com) is still in EAP, but in the meanwhile the Microsoft Teams feature is now available for all.
If you want to configure this new channel, note that Side Conversations have [moved](https://support.zendesk.com/hc/en-us/articles/5852242092186-Announcing-a-change-to-side-conversations-settings?ref=internalnote.com) to a new location in Settings.
[Using Microsoft Teams in side conversationsWhat’s my plan? Side conversations are spaces in a ticket where agents can have a conversation with a specific group of people, or discuss a specific area of concern or…Zendesk helpColleen Hall](https://support.zendesk.com/hc/en-us/articles/5191537451290?ref=internalnote.com)
## Messaging Voice API
Launched late last month, Zendesk now allows you to enable calls via the Zendesk Messaging widget, powered by their Digital Talk lines. It's a half-launch since it's not yet integrated with Bot Builder or the widget, but with a bit of custom code you can easily enable this feature anywhere on your website or Help Center.
[Voice API powered by Zendesk MessagingYou can now let your customers place calls via the Messaging Widget. This article shows you how, and explains how to integrate Voice with your Zendesk Bot.Internal NoteThomas Verschoren](https://internalnote.com/voice-api-for-zendesk/)
## Capacity Management for Messaging
Omnichannel Routing is a big new feature in Zendesk that will underpin most of the new developments in Zendesk like Agent Home, Intelligent Triage and Agent Status. Zendesk has been adding features to their platform for over a year now to build up Omnichannel Routing to a powerful engine that will replace existing triggers and views.
This month they added support for routing Messaging Tickets based on their (in)activity.
[Announcing assignment and capacity management for messaging in omnichannel routingAnnounced on Rollout on June 15, 2023 June 15, 2023 Zendesk is pleased to announce a new setting to help you manage omnichannel routing capacity rules for messaging tickets. What’s changing?…Zendesk helpPrakruti Hindia](https://support.zendesk.com/hc/en-us/articles/5785684459418-Announcing-assignment-and-capacity-management-for-messaging-in-omnichannel-routing?ref=internalnote.com)
## Also announced
- **Total Resolution Time** was [added](https://support.zendesk.com/hc/en-us/articles/5785457764634-Announcing-a-new-SLA-metric-Total-resolution-time?ref=internalnote.com) as a new SLA metric which measures the total lifetime of a ticket across all statuses.
- **When you resize the composer window in a ticket** in the Agent Workspace, the window size is saved across multiple tickets. It stays the same until you resize it again. See [Composing messages in the Zendesk Agent Workspace](https://support.zendesk.com/hc/en-us/articles/4408831849882?ref=internalnote.com#topic%5Fs55%5Fzzg%5Frmb).
- Zendesk **fixed** an issue where tickets got **bad CSAT scores** due to link expanders in email clients accidentally clicking those links. Zendesk now updated their systems to prevent these kind of automatic clicks.
# ⚒️ EAPs
## Agent Home
The new Agent Home has been made available in beta for all interested customers. It replaces the classic Agent Dashboard and deeply integrates with Omnichannel Routing to give your agents a clear overview of work to be done.
[➕ Overview of the new Agent HomeAgent Home is a new way for your Zendesk agents to interact with tickets and conversations that focusses on work to be done in a nice overview. Read the article to read my first impressions!Internal NoteThomas Verschoren](https://internalnote.com/agent-home-beta/)
## A powerful enhancement for the Layout Builder EAP
Layout Builder is a new feature in Zendesk that allows you to create custom interfaces in Zendesk. You can move elements like the conversation view, ticket fields, sidebar apps or intelligent panel anywhere in the UI, you can hide elements, or make elements bigger and smaller.
The EAP has been available for a while now, but it got a major update by integrating with Contextual Workspaces. This new enhancement allows you to dynamically trigger your custom layout by brand, group, ticket status, form, and much more.
One of the next updates *should* also allow for pinned custom apps in the sidebar so they are even easier to reach for your agents.
[\[Release announcement\] Layouts + Contextual Workspaces is now live for EAP customers. You can now apply custom layouts based on brand, group, form, and more!Hi everyone, thanks for using Layout Builder and providing feedback through the various channels. One major feature request that has bubbled up is the ability to apply a custom layout to a particul…Zendesk helpPaul Von](https://support.zendesk.com/hc/en-us/community/posts/5706915959578-Upcoming-summer-release-of-Layouts-Contextual-Workspaces-will-allow-application-of-multiple-layouts?ref=internalnote.com)
# 💡Insights
Zendesk's documentation team has been on a roll recently and wrote a couple of handy flows to setup in Messaging:
- [Bypassing the 24-hour rule with macro's](https://support.zendesk.com/hc/en-us/articles/5869718332954?ref=internalnote.com)
- [Notify customers that a department is offline](https://support.zendesk.com/hc/en-us/articles/5746602384794?ref=internalnote.com)
- [Check agent availability within a bot conversation](https://support.zendesk.com/hc/en-us/articles/5706660392602?ref=internalnote.com)
I really like this renewed focus on recipes and how-to's in Zendesk's Help Center resources. Explaining practical flows that build on top of Zendesk is one of the reasons I started this blog, and Zendesk itself also writing this kind of stuff is a pretty good motivation to keep this blog going.
# ⚠ Major Changes
Agent Workspace, Messaging and Omnichannel routing are powerful additions to the Zendesk platform that have seen a rather slow pickup since it has a big impact on how agents work.
The end result is better experiences for both customers and agents, but logically, most Zendesk users have been a bit hesitant in deploying big changes to their environments.
However, recently Zendesk started to aggressively push activations of the above features to more and customers for [messaging](https://support.zendesk.com/hc/en-us/articles/5637090228250-Announcing-assisted-messaging-activations?ref=internalnote.com) and [omnichannel routing](https://support.zendesk.com/hc/en-us/articles/5716181238938-Announcing-automatic-activation-of-omnichannel-routing-?ref=internalnote.com).
You'll get an email from them once you're eligible, but you might want to make use of the summer slowdown to look into proactively moving to these features at your own pace.
# 🎥 Videos
[Zendesk has acquired TymeshiftThis morning, we announced Zendesk’s acquisition of Tymeshift! Many of you in the Zendesk Community already know and use Tymeshift, but for those of you who don’t, Tymeshift is an AI-powered soluti…Zendesk helpNicole Saunders](https://support.zendesk.com/hc/en-us/community/posts/5846605714842-Zendesk-has-acquired-Tymeshift?ref=internalnote.com)
# 📝 Articles this month
- [Zendesk Roundup for May 2023](https://internalnote.com/monthly-note-for-may/)
- [Voice API for Messaging](https://internalnote.com/voice-api-for-zendesk/)
- [Sending Automated Messages via WhatsApp with Sunshine Conversations](https://internalnote.com/sunshine-conversation-automations/)
- [Overview of the new Agent Home](https://internalnote.com/agent-home-beta/)
# And finally...
Did you know Zendesk has a build-in feature to check for missing attachments? 😱
[Why do I receive a missing attachment notification when submitting a ticket response?Question When I submit a ticket response, a Missing attachment? notification displays, but my reply doesn’t include ticket attachments. Why do I receive this message? Answer This message is promp…Zendesk helpNatassja Jordan](https://support.zendesk.com/hc/en-us/articles/5710417631258?ref=internalnote.com)
##
### Preview of the new Agent Home
URL: https://internalnote.com/agent-home-beta/
Last updated: 2024-08-19T20:36:03.000Z
Let's start this article with a quick game: one of these screenshots was taken in 2012\. The other was taken today. Can you spot the differences?
When comparing Zendesk 10 years ago and Zendesk today a **lot** has changed. Even Zendesk two years ago is almost incomparable with the version you use today.
Agent Workspace got a full redesign with a lot less visual clutter and integrated tickets, messaging and context. Similarly, the Admin Center completely reworked the Zendesk admin experience with a reworked structure, search and cleaner settings views. All reporting got moved to Explore, and even the Chat and Web Widget experience got a full make-over.
But there's one spot in Zendesk that's always been a big head-scratcher for me: the Agent Dashboard. It's the first element in the Agent Workspace navigation and it's a place I seldom use. The Dashboard contains a random list of priority sorted tickets, and a list of updates and wraps up with a list of stats.
I've seen customers using this view as their main entry point, and discover that there are dozens, if not hundreds of unattended tickets in their views once they click through. And guessing from quick poll I ran on our internal team slack at the office, I guess I'm not the only one not using this part of the Agent Workspace.
# Agent Home
A few weeks ago one of the product managers at Zendesk reached out to me to show a new concept they were working on and ask for some initial feedback. That concept turned out to be the new Agent Home. I had already seen some earlier versions of this new idea in slide decks, but apparently the new Agent Home was not only almost ready, but the beta was launched a few weeks later.
[Announcing Agent Home (beta)Announced on Beta rollout June 13, 2023 June 13, 2023 Zendesk is introducing an updated page in Support, called Agent Home. Agent Home is the one-stop-shop for agents to manage all their work,…Zendesk helpZac Garcia](https://support.zendesk.com/hc/en-us/articles/5785063561242?ref=internalnote.com)
So, what is *Agent Home*? Agent Home is a new dashboard for Agents. It shows Agents an overview of work to be done, as well as a list of recent updated tickets, and easy links to their followed or cc'd tickets.
Normally an Agent would work from the Views tab and most instances have one or more *Action Needed* views that list open or active tickets and work through the list of tickets handling the most urgent, oldest or, if cherry picking, easiest ticket first.
With the arrival of Omnichannel Routing and Messaging we see most instances move away from this shared list of tickets to tickets and conversations getting assigned to agents based on skills and availability.
Instead of a shared list of tickets, agents get tickets assigned to them on a *one-by-one* basis, and reopened or replied-to tickets get reassigned based on context and status. But even though this flow works with a view-based system, it's not ideal.
Enter Agent Home. It's a new overview for agent that takes inspiration from the [Asana Home Page](https://asana.com/guide/help/fundamentals/navigating-asana?ref=internalnote.com) or other task based dashboards. It'll show agent their assigned/routed tickets, active messaging conversations and other tickets that require their attention. Which means that, if you've got your Omnichannel routing (see below) setup correctly, they're gonna work to one view and one view only.
# What about Views?
Now that Agent Home is available, there are now three places in Zendesk to get an overview and insights in tickets: Agent Home, Views, Explore. Their functionality overlaps, but I see them having three distinct use cases and users in mind:
- Agent Home is by nature for agents, it shows **active tickets** and helps them to get their **work done**
- Views are now mostly for Team Leads; they allow to get an **overview of all backlog** in one plays sorted by intent, status, sla or whatever fits your needs. It allows for quick decisions and reassignments based on current workload
- Explore is for Managers. It gives insights and long term data on the status of the entire CX team and allows for longterm decisions.
# Omnichannel Routing
Agent Home is available for all Zendesk environments with Agent Workspace, but Zendesk advices it mostly for users on Messaging and Omnichannel Routing due to its focus on ticket routing and assignment.
Omnichannel Routing is rather new and relatively complex to setup, and I'm planning on a big article about the topic later this summer. So [subscribe](https://internalnote.com/#/portal/signup) if you want to receive the article!
[About omnichannel routing with unified agent statusWhat’s my plan? Omnichannel routing with unified agent status allows you to direct tickets from email (including web form, side conversations, and API), calls, a…Zendesk helpJacquelyn Brewer](https://support.zendesk.com/hc/en-us/articles/4409149119514?ref=internalnote.com)
# Wrap Up
Any change to the old Agent Dashboard would have been an improvement, and even removing that tab all together would have made the product better overal.
But this new Agent Home is pretty amazing. Even with only a week of access I've shifted to this new tab as my main view to work on third line tickets that get escalated to me.
It is a beta though and some features like live refresh, active messaging conversations and filters other than status are not available yet. That being said, it's a solid and stable tool that, in my opinion, can be used in active environments without much change management or training **if you already use omnichannel based routing or have SLA and priority enabled on your instance.**
Do I have a wish list of features? Sure!
- Allow to create filters on intents and store them as quick links in the sidebar
- Allow developers to create new widgets to be shown in the sidebar (Google Calendar? Asana? [Message Board](https://www.zendesk.com/marketplace/apps/support/738366/message-board/?ref=internalnote.com)? Weather?
- Show a colleagues dashboard so I can quickly see workload of my team members as a teamleader.
- Show my requested tickets, show side conversations I'm a part of as filters below the existing Cc'd and Follower views.
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### ➕ Sending Automated Messages via WhatsApp with Sunshine Conversations and Zendesk
URL: https://internalnote.com/sunshine-conversation-automations/
Last updated: 2025-09-08T06:42:44.000Z
In a [previous article](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/) I wrote about how you can now use the new Sunshine Conversations in Zendesk Suite to work around the 24h limit for WhatsApp conversations.
There are however a lot more (complex) use cases possible via the Conversations API. Think about sending confirmations or reminders for bookings. Send password reset codes, confirm subscriptions or changes in order status,.. the list goes on.
All this is possible by leveraging two kinds of conversation types: messages, and notifications, whoch both allow you to send messages over social channels like WhatsApp, Instagram, Twitter,...
# Conversation Types
There are two types of conversation updates you can send to your customers.
The first are **comments**. These are replies to existing conversations and can be in the form of Agent replies (via Agent Workspace), free text or templates (both via API). The important part here is that as long as you are in an existing conversation with a customer over e.g. WhatsApp or Instagram DM, you're allowed to send anything to them.
The other type are **notifications**. These allow you to send a message to anyone without the requirement of an existing conversation. You can however only send approved message templates and these notifications come at a higher cost (you're also limited to the amount you can send depending on your verifications state)
**Templates** for e.g. WhatsApp are created via the WhatsApp business portal, other platforms have similar requirements. Check out this [article](https://docs.smooch.io/guide/key-concepts/?ref=internalnote.com#message) for more info.
# Sending messages
There's a few ways we can send out messages:
1. Automated via Zendesk triggers/automations, similar to how we solved the WhatsApp issue, this is ideal for reminders, confirmation messages,...
2. Automated via external platforms like your webshop, booking tool,...
3. Manually by agents in Agent Workspace who can use e.g. a button to trigger a template.
Options (1) and (2) require a webhook and some worker or script that handles the message. (3) can be done via a sidebar app in Zendesk. In this article we'll explore both options.
Let's dive in!
# Messages via webhooks
The easiest way to make a scalable way to interact with Sunshine Conversations is to setup a script *somewhere* that takes messages and users as input, and handles the flow to the Sunshine Conversation endpoints.
For this tutorial I'll assume you have access to Cloudflare Workers (free!) and will handle the WhatsApp channel.
You can copy the entire workflow from the example [GitHub](https://github.com/verschoren/outbound%5Fmessaging?ref=internalnote.com) repository and paste it into a blank Cloudflare Worker via your browser. Just remember to create the required variables as defined below.
[GitHub - verschoren/outbound\_messagingContribute to verschoren/outbound\_messaging development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/outbound%5Fmessaging?ref=internalnote.com)
### Requirements
- **Sunshine Conversations API Credentials**
This can be done via the Admin Center [https://yourdomain.zendesk.com/admin/apps-integrations/apis/conversations-api](https://d3v-verschoren.zendesk.com/admin/apps-integrations/apis/conversations-api?ref=internalnote.com). This gives you an `app_id`, `secret_key` and `key_id`
- **Integration ID for WhatsApp**
You need to know your `integration_id` for WhatsApp. You can find this in Admin Center by selecting your WhatsApp channel and copying the last part of the URL: [https://yourdomain.zendesk.com/admin/channels/messaging\_and\_social/channels\_list/edit/**646b411abc0c536ba1a98263**](https://d3v-verschoren.zendesk.com/admin/channels/messaging%5Fand%5Fsocial/channels%5Flist/edit/646b411abc0c536ba1a98263?ref=internalnote.com)
- Zendesk API Token
You will need a Zendesk [API Token](https://www.google.com/search?client=safari&rls=en&q=zendesk+api+token&ie=UTF-8&oe=UTF-8&ref=internalnote.com), an admin email address, and your Zendesk subdomain.
- You need (at least) one [approved WhatsApp template](https://www.google.com/search?q=create+whatsapp+template&client=safari&rls=en&sxsrf=APwXEdchRzrKGY06VApVE1TPOVf-8VbqFw%3A1685543701718&ei=FVt3ZKW9K-eM9u8P1LyfmAg&ved=0ahUKEwjlm-DM45%5F%5FAhVnhv0HHVTeB4MQ4dUDCA4&uact=5&oq=create+whatsapp+template&gs%5Flcp=Cgxnd3Mtd2l6LXNlcnAQAzIFCAAQgAQyBQgAEIAEMgUIABCABDIFCAAQgAQyCAgAEAUQBxAeMgYIABAIEB4yBggAEAgQHjoKCAAQRxDWBBCwAzoKCAAQigUQsAMQQzoQCC4QigUQxwEQ0QMQsAMQQzoHCCMQsAIQJzoGCAAQBxAeOgcIABANEIAEOggIABAIEAcQHjoICAAQigUQhgNKBAhBGABQ-AJYwQhgwgpoAXABeACAAViIAZMEkgEBN5gBAKABAcABAcgBCg&sclient=gws-wiz-serp&ref=internalnote.com)
### Creating variables
I stored all these inside Cloudflare Workers as Environment Variables:
## Script Overview
The basis script we build has two endpoints: `/notifications` and `/messages` to handle the two use cases mentioned above. Notifications can be send to anyone, but require a template. Messages can be send to active conversations and can contain messages or templates.
1. We check the url pathname for /notifications or /messages
2. We get the input `await request.json()`
3. We handle any errors that occur.
## Sending Notifications
Notifications require two inputs: a user identifier (phone number) and a template name from an approved WhatsApp template.
```json
//POST https://domain.workers.dev/notifications
{
"phone":"1234567890",
"template":"template_name"
}
```
All logic for sending conversations is contained in the `sendNotification()` function. It first creates a `payload` that references the phone, template and integration ID, and then contacts the Smooch API endpoint.
```javascript
async function sendNotification(phone, template, env){
var payload = {
"destination": {
"integrationId": env.sunco_integration_id,
"destinationId": phone
},
"author": {
"role": "appMaker"
},
"messageSchema": "whatsapp",
"message": {
"type": "template",
"template": {
"namespace": "XXXXXXXX_XXXX_XXXX_XXXX_XXXXXXXXXXXX",
"name": template,
"language": {
"policy": "deterministic",
"code": "en"
}
}
}
}
const api_endpoint = `https://api.smooch.io/v1.1/apps/${env.sunco_app_id}/notifications`
const sunco_key = btoa(env.sunco_key_id + ":" + env.sunco_secret_key);
const init = {
body: JSON.stringify(payload),
method: "POST",
headers: {
"content-type": "application/json",
"authorization": "Basic " + sunco_key
},
};
const response = await fetch(api_endpoint, init);
const results = await response.json();
return results;
}
```
## Sending Comments
Where sending notifications is pretty straightforward, sending comments is a bit more complicated. Assuming you start from a known Zendesk user:
1. You need to find the Sunshine Conversations User ID for your user
2. You then need to find the active conversation for that user.
3. You can then send your message.
The comments endpoint expects the following JSON payload. Requester is the user ID of a Zendesk user and message can be any freeform text we want.
```json
// POST https://domains.workers.dev/messages
{
"requester":"1234567890",
"message":"message"
}
```
🔍
If you only know the email of a user, you can search for a user based on the email via `https://domain.zendesk.com/api/v2/search.json?query=type:user email:name@domain.com"` and return `results[0].id`
### User Identity
Assuming you have a user ID from a Zendesk user, you can use that to return the identities for that user. Most users have only one Messaging identity, namely from the channel they first contacted you. If you have a tendency to merge your users you might need to expand this function to filter out the right one based on the type.
We need to target the Zendesk API for this, and specifically the `/users/{user_id}/identities` endpoint.
```javascript
async function getIdentities(requester_id, env){
const api_endpoint = `https://${env.zendesk_domain}.zendesk.com/api/v2/users/${requester_id}/identities.json`
const zendesk_key = btoa(env.zendesk_admin_email + "/token:" + env.zendesk_token);
const init = {
method: "GET",
headers: {
"content-type": "application/json",
"authorization": "Basic " + zendesk_key
},
};
const response = await fetch(api_endpoint, init);
const results = await response.json();
for (let index = 0; index < results.identities.length; index++) {
const identity = results.identities[index];
if (identity.type == "messaging"){
var messaging_id = identity.value;
return messaging_id;
}
}
return '';
}
```
### Conversation ID
Once we have our user id in messaging, we can use that to find the active conversation for that user. Remember, comments can only be send to active conversations, so we need to target the right one.
This happens via the Smooch API, which is the old brandname for what's now called Sunshine Conversations (SunCo). By default SunCo only has a single conversation per user, and WhatsApp only allows for one anyway, so we can safely return the first conversation we find.
```javascript
async function getConversationId(messaging_id, env){
const api_endpoint = `https://api.smooch.io/v2/apps/${env.sunco_app_id}/conversations?filter[userId]=${messaging_id}`
const sunco_key = btoa(env.sunco_key_id + ":" + env.sunco_secret_key);
const init = {
method: "GET",
headers: {
"content-type": "application/json",
"authorization": "Basic " + sunco_key
},
};
const response = await fetch(api_endpoint, init);
const results = await response.json();
return results.conversations[0].id;
}
```
### Send Message
Now that we know which conversation to target we can finally send the message. This happens by configuring a new `payload` that contains the `message` from our original POST payload.
Note that we use `zd:agentWorkspace` as a source. This way any reply will surely arrive back to your agents in Zendesk!
```javascript
async function sendMessage(conversation_id, message, env){
const api_endpoint = `https://api.smooch.io/v2/apps/${env.sunco_app_id}/conversations/${conversation_id}/messages`
const sunco_key = btoa(env.sunco_key_id + ":" + env.sunco_secret_key);
var message_payload = {
"content": {
"type": "text",
"text": message
},
"author": {
"type": "business"
},
"source": {
"type": "zd:agentWorkspace"
}
}
const init = {
body: JSON.stringify(message_payload),
method: "POST",
headers: {
"content-type": "application/json",
"authorization": "Basic " + sunco_key
},
};
const response = await fetch(api_endpoint, init);
const results = await response.json();
return results;
}
```
## How to use this.
Now that we have a deployed script that can handle any WhatsApp message, you can use this to `POST` any kind of message to your customers, and this flow is ideal for automated API flows.
If you want an easy way to integrate you could invoke this script via e.g. Zapier as a webhook target and sent a reminder to a user based on a Google calendar with appointments to remind customers of an upcoming event.
# Sidebar App
The previous flow described how you could create a script you could invoke via a POST command to trigger messages and notifications to customer.
There is however another way you can use Sunshine Conversations and that is leveraging sidebar apps in Agent Workspace. In the repository below you'll find a fully functional sidebar app for Zendesk that allows you to:
- Send a message to a user from the ticket sidebar.
- Send a template to a user from both the ticket sidebar as well as a user profile.
The settings of the app allow you to enter two approved templates and require you to enter the same `app_id, secret_key, key_id, integration_id` as you entered for the worker script above.
You can download the app via the GitHub repository below. It's a fully functional private app you can modify or install in your instance.
[GitHub - verschoren/outbound\_messagingContribute to verschoren/outbound\_messaging development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/outbound%5Fmessaging?ref=internalnote.com)
## Code Highlights
This being a Zendesk app makes the app flow a bit different than our script.
### Get user info
Since our app runs in both the `ticket_sidebar` and the `user_sidebar` we can use the `client.context();` function to get the location of our app, and we use `client.metadata();` to securely retrieve our API credentials we store in the apps' settings.
```javascript
var client = ZAFClient.init();
var metadata,context;
var recipient,phone;
var conversation_id,key,integration_id
metadata = await client.metadata();
context = await client.context();
```
Once we have our location we need to get the user information we want to sent a message to.
```javascript
if (context.location == 'ticket_sidebar'){
recipient = await client.get('ticket.requester')
} else {
recipient = await client.get('ticket.requester')
}
phone = await getUser(recipient);
async function getUser(requester){
return await client.request({
url: '/api/v2/users/' + requester.id,
type: 'GET',
dataType: 'json'
}).then(function(data) {
$('#phone').html(data.user.phone.replaceAll(' ', ''));
return data.user.phone.replaceAll(' ', '');
});
}
```
And now that we have the user, similar to how we did it in our script, we can get the messaging identity of the user from our `recipient.identities` object. The rest of the app flows similar to how we handled our earlier script.
## How to use the app
Agents can reply to conversations and send random messages to customers via the Message Field, although, honestly, in this case using the Agent Workspace is more logical. They can also send predefined templates via the two buttons underneath the text field.
You could imagine expanding this app and integrating it with your CRM, booking system or others, to allow agent to send specific templates with relevant placeholders to their end users.
# Conclusion
So, there you have it, two different ways we can interact with Sunshine Conversations, one ideal for automations, and one to make it possible for agents to interact with Sunshine Conversations via a sidebar app.
### Digital Voice API powered by Zendesk Messaging
URL: https://internalnote.com/voice-api-for-zendesk/
Last updated: 2024-08-19T20:37:46.000Z
The Classic Zendesk widget was a powerful tool. It allowed for [full customization](https://internalnote.com/customize-and-brand-your-zendesk-widget/), offered help center search, chat, webforms, authentication, pre-filling values and ... Zendesk Talk support out of the box.
The new Messaging widget has gradually catching up with its older sibling. The customisation API has been expanding with a lot more native features available in the admin panel, you can [authenticate](https://internalnote.com/jwt-messaging/) Zendesk Messaging, and, arguably, the new Zendesk Bot experience is way better than the old model search and forms flow.
This week brings the latest addition to its feature set: [Voice integration](https://support.zendesk.com/hc/en-us/articles/5531102360090-Announcing-the-new-Voice-API-powered-by-Zendesk-Messaging?ref=internalnote.com) that allows customers to call your CX team right from within the browser. No phone or other software required.
It's build on top of Zendesk Talk's [Digital Lines](https://www.google.com/search?client=safari&rls=en&q=digital+line+zendes&ie=UTF-8&oe=UTF-8&ref=internalnote.com) feature where you can get Talk enabled in your environment without having a real number.
# Feature overview
The new Voice feature works as follows:
- You add a *Call Us* button on your website
- When clicked you invoke a specific API endpoint for the new Messaging widget
- The widget shows a Call now (or we're offline) view allowing customers to call you.
- Calls appear in Agent Workspace and agents can handle them just like any other ticket.
You can test out the flow below:
[Internal Note - Voice API for MessagingDemo page to showcase the JWT Authentication for Zendesk MessagingVoice API for MessagingInternal Note](https://widget.internalnote.com/voice?ref=internalnote.com)
# Enabling Voice via Messaging
## Get a digital line
First you need to add a Digital Line to your Zendesk instance. Zendesk has a pretty good article explaining it:
[Adding a Talk digital lineWhat’s my plan? In Understanding Talk Embedded voice, you learned the basics about how Talk digital lines can help you improve the experience for your customers, wherever…Zendesk helpRob Stack](https://support.zendesk.com/hc/en-us/articles/1260805715389-Adding-a-Talk-digital-line?ref=internalnote.com)
Next you will need to find the Digital Line ID of your line. You can do this by going to [https://yourdomain.zendesk.com/api/v2/channels/voice/lines](https://d3v-verschoren.zendesk.com/api/v2/channels/voice/lines?ref=internalnote.com) in your browser while being logged in as an admin to your Zendesk instance.
In the returned data, search for the `line_id` of the line with `line_type:digital`
```json
{
"lines": [
{
"line_type": "digital",
"line_id": "6c0e70e6ef77d7a711a073f249ab6304",
...
```
Once you have your `line_id` you can invoke the Voice widget via:
```javascript
zE('messenger:open', 'voice', '{{line_id}}')
//replace {{line_id}} with your value
```
## Triggering the widget
Sadly there no native way (yet) to trigger the voice channel directly from the widget. However, there are a few ways you can show the option to your customers:
### Call us button
The easiest way to trigger the feature is buy adding a button somewhere on your website that opens the widget.
```html
Call Us
```
[Call Us](https://internalnote.com/voice-api-for-zendesk?voice)
### URL-based triggers
Sometime you might want to send customers directly to the widget without them needing to click on anything.
An example scenario might be a *Call Us* link in an email you send to your customers. For example, if I ask you to click on [https://widget.internalnote.com?voice=true](https://widget.internalnote.com/?voice=true&ref=internalnote.com) you will land on my demo page and the call widget will be immediately available to you.
By using the code below, you can automatically trigger a call based on the url parameter `voice=true`
```javascript
$(function() {
//check for URL parameter called "voice" and if it is set, open the voice channel
var urlParams = new URLSearchParams(window.location.search);
if (urlParams.has('voice')) {
zE('messenger:open', 'voice', '6c0e70e6ef77d7a711a073f249ab6304')
}
});
```
### Triggering a call via the Zendesk Bot
One of the limitations of the Voice via Messaging feature is its lack of integration with the Zendesk Bot. Where the [Classic Widget](https://developer.zendesk.com/api-reference/widget/core/?ref=internalnote.com#contactoptions) had a `contactOptions` feature which listed chat, forms and voice as options in a list, the current bot misses this feature.
If you do want to offer a similar feature, there's a small workaround you can use to make this work:
1. Add the URL based trigger code above to your website
2. Insert a Bot Message to your flow and add a button to the message
3. Set the link of the button to *https://domain.com?voice=true* and give it a useful label
4. Whenever a customer clicks the *call us* button it will open a new window with the call function enabled.
I sure hope this will be a native feature in the future.
# Conclusion
Voice via Messaging is one of those feature additions that seem really cool, but in practice lack in functionality. Hacking this into the Zendesk Widget with an API and neglecting to enable any native way to trigger it makes me really wonder why this feature had to go live now.
I can understand adding it to the Bot Builder might require additional resources from different teams, but it feels like a natural step to add to that list of options.
Or, if that was not possible in the near future, even adding a *Continue via phone* option to this dropdown would have been nice.
That being said, it is a nice feature addition to Zendesk's omnichannel set of features that brings the Messaging widget yet one step closer to feature parity with the classic widget.
### Zendesk Roundup for May 2023
URL: https://internalnote.com/monthly-note-for-may/
Last updated: 2023-10-12T14:07:49.000Z
A few weeks ago Zendesk held there big Relate conference which [introduced](https://internalnote.com/relate-2023/) Zendesk AI, Adaptable Agent Workspace and a lot of security and privacy improvements.
For those thinking that was enough for one month, well think again, cause this roundup is going to be a long one 😅
> *Our climate commitment: Zendesk signs its first off-take agreement to scale carbon removal technology.*
The above is a quote from a recent [press release](https://www.zendesk.com/newsroom/articles/our-climate-commitment-zendesk-signs-its-first-offtake-agreement-to-scale-carbon-removal-technology?ref=internalnote.com) by Zendesk. Closely [followed](https://www.zendesk.com/newsroom/articles/zendesk-usatoday-climate-leaders/?utm%5Fcampaign=61d659e08d1774000193466d&utm%5Fcontent=646f97876f007d00015a6e18&utm%5Fmedium=smarpshare&utm%5Fsource=linkedin) by "**Zendesk ranks in top 10 of USA Today’s Climate Leaders".**
There's just something about companies that put people and climate above profit..
🥳
Also, a small personal milestone: Internal Note reached its hundred subscriber this month. Six months in and I can finally start calling this website a real blog and no longer a proof of concept. So for all of you [subscribing](https://internalnote.com/#/portal/signup) (free or paid), thanks! 🥳
And if you're not yet subscribed and just reading this: I compile an overview of Zendesk News every month. If you don't want to miss it, please [subscribe](https://internalnote.com/#/portal/signup/free) to the blog. It's free. (Or optionally paid if you really like it)
Every new subscriber motivates me to keep putting in the effort.
Thomas
Now, on to the updates!
# 🎉 New Releases
## Help Center Updates
Zendesk Guide keeps getting better month after month. This month marks the release of *Semantic search in Guide,* a completely revamped search engine for Guide that moves away from exact keyword matches to using the meaning of a query to find matches. It's enabled automatically for all Zendesk users.
Additionally, Zendesk Guide now gives the option to feature up to six articles in the search dropdown. Similar to Promoted Articles, these articles allow you to intercept customers and draw attention to the articles you want them to read first.
You can now [update](https://support.zendesk.com/hc/en-us/articles/5401983216794.html?ref=internalnote.com) translations via API for articles that contain Content Blocks, and the [Audit Log](https://support.zendesk.com/hc/en-us/articles/5685642913050-Announcing-Help-center-events-available-in-the-Audit-log?ref=internalnote.com) now shows configuration and security changes in Guide too!
And to round up this overview, Zendesk just announced the availability of the new account level image gallery for Guide. Similar to the [Media Gallery](https://wordpress.com/support/media/?ref=internalnote.com) in Wordpress this will give you a single place to get an overview of images in your Help Center and easily share images across Content Blocks and Articles.
[Announcing semantic search in GuideAnnounced on Rollout starts May 1, 2023 April 28, 2023 We’re thrilled to announce the newest leap Guide is taking towards intelligent help centers, semantic search. Unlike search methods that…Zendesk helpOrsolya Forster](https://support.zendesk.com/hc/en-us/articles/5641000272922-Announcing-semantic-search-in-Guide-?ref=internalnote.com)
[Announcing featured articles in Guide searchAnnounced on Rollout starts Rollout ends May 15, 2023 May 15, 2023 May 22, 2023 We’re excited to announce the new article recommendation tool in Guide. Featured articles allows Guide admins t…Zendesk helpOrsolya Forster](https://support.zendesk.com/hc/en-us/articles/5685608132890?ref=internalnote.com)
[Announcing the new account level image gallery for articlesAnnounced on Rollout starts Rollout ends May 24, 2023 May 24, 2023 June 9, 2023 We are excited to announce the new account level image gallery for articles, which provides you with new tools…Zendesk helpKatarzyna Karpinska](https://support.zendesk.com/hc/en-us/articles/5737099213978-Announcing-the-new-account-level-image-gallery-for-articles?ref=internalnote.com)
## Group SLA
One of the biggest requests I hear from customers when talking about Service Level Agreements is a way to measure internal reply times, e.g. when escalating tickets to other teams or departments. As shown [in this article](https://internalnote.com/sla-policies/) from earlier this month, that's exactly what the new Group SLA policies allow you to do: report on and show expected handling time to agents in Zendesk for tickets escalated between teams.
[Announcing group service level agreementsAnnounced on Rollout starts Rollout ends May 10, 2023 May 15, 2023 May 19, 2023 Zendesk is excited to announce the introduction of group service level agreement (SLA) policies! Commonly known…Zendesk helpScott Allison](https://support.zendesk.com/hc/en-us/articles/5617720797338?ref=internalnote.com)
## Sunshine Conversations API is now part of Suite
The relationship between Zendesk Suite and Sunshine Conversations (SunCo) is a complex one. One one hand Zendesk Messaging delivered a lot of the capabilities of SunCo out of the box, on the other hand, integrating Conversational bots like [Ultimate](https://ultimate.ai/?ref=internalnote.com), or enabling [proactive messaging](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk), was only possible by buying the, arguably, expensive SunCo add-on which often, especially for smaller companies, costs more than their Zendesk setup.
Luckily this is now fixed thanks to **Sunshine conversations now being included** with Zendesk Suite Professional or Enterprise plans!
Natively available within the Admin Panel this integration now allows powerful capabilities like sending marketing messages via WhatsApp, enabling third party Chatbots, routing messages from Zendesk to Slack,...
📥
If you're interested in this new platform capability, and you loved the "[Breaking the 24-hour Whatsapp rule](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk)" from last week, please subscribe. There's lots more coming!
## Draft Mode in Composer
Often when writing a long or complex reply agents will turn to the internal note (😎) mode in the Comments Composer to prevent accidentally submitting an incomplete answer. With the move to Agent Workspace switching from public comment to internal note became a modal shift and the written text gets hidden. Now, with the new draft mode in the ticket composer you can write a draft first, then send it when you’re ready to submit the update, preventing you from accidentally submitting a reply.
[Announcing draft mode in the composerPhased rollout - Announced on May 16, 2023 Rollout starts Rollout ends Phase 1: All channels that support public replies, except chat and messaging May 16, 2023 May 19, 2023 Phase 2: Chat and m…Zendesk helpAmisha Sharma](https://support.zendesk.com/hc/en-us/articles/5637822550938?ref=internalnote.com)
## Conversational Bots
Speaking of draft modes, Zendesk has enabled a new option in bot builder tool that allows Admins to test new flows without needing to publish the changes first.
[Announcing new testing capability for conversation botsRollout start Rollout end May 19, 2023 May 23, 2023 We’re excited to announce that we have released a new bot testing capability for conversation bots. This announcement includes the following…Zendesk helpLisa Tam](https://support.zendesk.com/hc/en-us/articles/5727388526874-Announcing-new-testing-capability-for-conversation-bots?ref=internalnote.com)
## Other small changes
- The new [Developer Portal Change Log](https://developer.zendesk.com/api-reference/changelog/changelog/?ref=internalnote.com) captures significant changes, additions, fixes, deprecations, removals, and betas and EAPs related to Zendesk APIs and SDKs, and the list is filterable by event type.
- Admis can now [see](https://support.zendesk.com/hc/en-us/articles/5711217754394-Announcing-visibility-of-all-macros-to-admins-in-the-Macros-API?ref=internalnote.com) all macros in an instance, even personal ones both via the Admin Panel as well as via API.Visibility of all macros to admins.
- The Group View got [redesigned](https://support.zendesk.com/hc/en-us/articles/5547375026714-Announcing-an-updated-Groups-page?ref=internalnote.com) and got faster search to match the new UI for Agents, Views or Triggers.
- User profiles now show a nice time offset for customers in different timezones.
# 💡Insights
## Using Generative AI in Customer Care
The same week Zendesk announced Zendesk AI, Ultimate posted a nice overview of ways Generative AI can be used for customer care.
> Generative AI is a powerful tool that (when built into a broader automation or CX strategy) can help companies to deliver faster, better support — from offering a more conversational experience for customers, to assisting agents and supporting bot builders. So let’s take a look at the top 6 use cases of gen AI for customer service. - Reetu Kainulainen | Ultimate
[How to Intelligently Use Generative AI in Customer ServiceHere’s a deep dive into what gen AI is and 6 leading examples of generative AI and how it can be used in the support space.ultimate.aiReetu Kainulainen](https://www.ultimate.ai/blog/collaborations/gen-ai?ref=internalnote.com)
## Conversations with Zendesk
Brand new podcast hosted by [Zendesk Community](https://support.zendesk.com/hc/en-us/community/topics?ref=internalnote.com) Manager Nicole Saunders. The first episode was releases a little while ago so go listen!
[Conversations with Zendesk - Interviews about Customer Service, Support, and Customer Experience on Apple PodcastsBusiness · 2023Apple Podcasts](https://podcasts.apple.com/be/podcast/conversations-with-zendesk-the-podcast/id1685847701?ref=internalnote.com)
# ⚠ Major Changes
## Zendesk moves to Monthly Active Users
Zendesk has announced a few changes to the way it works with conversations and bots.
First off, Answer Bot and Flow Builder are gone. From now on there will be Zendesk Bots: The Conversational Bot powered by Bot Builder, and Autoreply to handle your email and webform automations. And similar to Answer Bot in the past, customers will now get a set of Monthly Active Users (MAU) as part of their Zendesk Suite, and can buy additional MAUs if they go over that limit.
There's a lot of details to go over, but in a nutshell, every unique user interact with your Conversational Bot (Free-text entry, Clicking on a quick-reply option or Submitting a detail collection form) it's counted as a MAU. You can lower this amount by enabling [Authentication for Messaging](https://internalnote.com/jwt-messaging/). Similarly, if a customer gets suggested articles via Autoreply, they will also counts an an MAU.
Note, a user that contacts you multiple times in the same month only counts as one MAU.
[About monthly active users for Zendesk botsNote: We have been actively listening to our customers about this change (thank you) and have decided to revisit our plan for bot usage pricing. This shift is in light of your helpful feedback and…Zendesk helpAimee Spanier](https://support.zendesk.com/hc/en-us/articles/5352026794010-About-Monthly-Active-Users-for-Zendesk-bots?ref=internalnote.com#topic%5Fgww%5F13x%5F4wb)
Similarly, users leveraging the new Sunshine Conversation APIs in Zendesk Suite also get an alloted quota (see Above). For users who need more proactive messages, or more users, you can buy additional quota via a new[ add-ons](https://support.zendesk.com/hc/en-us/articles/4408834152730?ref=internalnote.com).
# 🎥 Videos
# 📝 Articles this month
[Setting up SLA Policies and enabling Group Policies.In this article we’ll explore the new Group SLA policies in Zendesk, lay out an approach to setup a scalable SLA policy across your instance and make sure agents take care of the right ticket first.Internal NoteThomas Verschoren](https://internalnote.com/sla-policies/)
[What’s new at Zendesk Relate 2023Interested in what Zendesk announced at Relate? Want to know more about Zendesk AI, Intelligent Triage, Adaptable Agent Workspace and Conversational Commerce? Read my “What’s new at Relate 2023” overview and discovered all the new announcements!Internal NoteThomas Verschoren](https://internalnote.com/relate-2023/)
[➕ An in-depth overview of Proactive Messages for ZendeskDiscover the new Zendesk proactive messaging. Learn about the main features and advanced flows, including the ability to show proactive messages based on specific marketing campaigns. Target customers based on their locale, or offer a premier experience to VIP users.Internal NoteThomas Verschoren](https://internalnote.com/proactive-ticketing-for-messaging/)
[➕ Bypassing the 24-hour rule for WhatsApp in ZendeskIn this article we’ll show you how to bypass the 24-hour WhatsApp limit in Zendesk by leveraging the included Sunshine Conversations API.Internal NoteThomas Verschoren](https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk)
# And finally...
I stumbled on this pretty handy article listing all feature deprecations in Zendesk. Might be useful to plan for future obsolesce.
Speaking of.. have you moved to Agent Workspace and Messaging already?
[What is being removed \[updated: September 2023\]Please refer to the following articles to learn more about Zendesk feature removal: How much notice does Zendesk give for feature removal How Zendesk communicates feature removal Why Zendesk remo…Zendesk helpJennifer Rowe](https://support.zendesk.com/hc/en-us/articles/4408843026714-What-is-being-removed-updated-April-2023-%202021-10-16T04:00:19Z%20-%202023-05-01T16:08:21Z?ref=internalnote.com)
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### ➕ Bypassing the 24-hour rule for WhatsApp in Zendesk
URL: https://internalnote.com/breaking-the-24h-rule-for-whatsapp-in-zendesk/
Last updated: 2025-09-08T06:42:39.000Z
One of the benefits of Zendesk's Omnichannel approach is that you can connect with customers across a myriad of channels from one centralised tool. Where Zendesk used to be focussed mainly on email, and later Facebook and Twitter, Zendesk Messaging now allows you to interact with customers across web, Facebook, Instagram, Twitter and WhatsApp easily.
Even though Zendesk provides a unified experience for your agents, and a native experience for your end-users on whatever platform they decide to use, some channels have specific limitations defined by their vendor to prevent spam or other bad behaviour on their platform.
WhatsApp for example has a 24-hour rule which prevents business integrations from replying to customers that haven't interacted with your business for over 24 hours. Example: If a customer sends a ticket at 10AM, you need to make sure your agent replies AND you get a reply from the customer before 10AM the next day. If those 24h pass without a customer reply to your agents' remark, you can't send any reminder or more info to that customer anymore.
[Working with WhatsApp tickets and the 24-hour ruleWhat’s my plan? If your WhatsApp account has been added to Zendesk Support, you can communicate with end users about support requests using…Zendesk helpAmy Malka Edited April 11, 2023 20:53 Zendesk Documentation Team](https://support.zendesk.com/hc/en-us/articles/4408829291162-Working-with-WhatsApp-tickets-and-the-24-hour-rule?ref=internalnote.com#:~:text=This%20is%20a%20WhatsApp%20rule,%2C%20you%20can%27t%20reply.)
# Bypassing the 24-hour rule
Luckily, there is a way to work around this limitation by leveraging the WhatsApp Business API and Notification feature. If you use the WhatsApp Business API you can send outbound messages to customers regardless of the expired time.
If a customer replies, their messages gets added to your existing conversation, and your agents can reply to them within a new 24h window.
Sending Outbound notifications to WhatsApp users requires you to have a Sunshine Conversations license. This used to be an additional add-on you needed to buy for your Zendesk environment,
# Sunshine Conversations in Zendesk Suite
Starting this month every Zendesk Professional or Enterprise Suite user gets access to Sunshine Conversations. You get full Platform and API access with an included limit of 1,000 Monthly Active Users (MAU) and 1,000 outbound notifications.
For larger use cases you can purchase additional users and notifications.
[About Sunshine Conversations platform access and supportWhat’s my plan? Sunshine Conversations is a messaging platform that unlocks the conversations API for customers who want advanced customizations. It is designed to help businesses build interactiv…Zendesk helpGary Beichler](https://support.zendesk.com/hc/en-us/articles/5514407356954?ref=internalnote.com)
What's nice about this is that, where previously there was no way to work around the 24-hour limit, now every Zendesk customer can solve this limitation for free and native within Zendesk.
# How does this work?
The setup is rather easy:
1. A customer contacts you over WhatsApp
2. Their conversation gets picked up by a Zendesk Conversation Bot to allow for self service and ticket deflection
3. If needed, the Bot forwards the message to an Agent
4. The Agent replies to the ticket and puts the ticket on Pending once the customer stops interacting
5. An automation monitors the WhatsApp conversations and sends out a reminder to customers after 24 hours
6. If the customer reacts before the 24-hour limit, or reacts to the outbound message, the conversation re-opens and we have a new 24h window to handle the conversation.
# Requirements
In order to enable automatic notifications to your customers there are a few requirements. Most of these are probably already setup when you first enabled WhatsApp for Zendesk Messaging:
- You need a verified Facebook Business Account - [link](https://www.facebook.com/business/help/2058515294227817?id=180505742745347)
- It's best to also verify your WhatsApp for Business account - [link](https://developers.facebook.com/docs/whatsapp/overview/business-accounts/?ref=internalnote.com)
- You need to have WhatsApp linked to Zendesk Messaging - [link](https://support.zendesk.com/hc/en-us/articles/4408842821786-Adding-WhatsApp-channels-to-the-Zendesk-Agent-Workspace?ref=internalnote.com)
- Your templates (see later) need to be validated by Meta. This takes \~2-3 days.
## A word about cost
Sending out outbound messages over Sunshine Conversations has two costs involved:
1. You need to pay for MAU in Sunshine Conversations. You get a certain amount (1,000 Monthly Active Users (MAU) and 1,000 outbound notifications) included for free. If you go over you need to buy a [Sunshine Conversations add-on](https://support.zendesk.com/hc/en-us/articles/5514407356954-About-Sunshine-Conversations-platform-access-and-support?ref=internalnote.com).
2. Outbound WhatsApp messages have a cost charged by [Meta](https://www.facebook.com/business/help/2225184664363779?id=2129163877102343). You get a 250-1000 depending on your verification status
# Setup
To setup this flow we need to do four things:
1. Create a Sunshine Conversations API integration
2. Create a Template in Facebook Business Manager
3. Setup a webhook in Zendesk
4. Setup an automation in Zendesk
The next part of this article will show you how.
# Create a Sunshine Conversations API integration
The first step in this setup requires the creation of a new Conversation API key.
Go to the Admin Center > Apps and Integrations > APIs > Conversations API and click *Create API Key*. Give it a name and copy the 3 values Zendesk shows somewhere safe. You'll need this App ID, Key ID and Secret Key in the next steps.
🤔
Zendesk has both a **Conversation Integrations* and **Conversations API* option in the Admin Panel. The former is used for ChatBots (like [Ultimate](https://www.ultimate.ai/?ref=internalnote.com), the latter for these kind of custom integrations.
# Create a Template in Facebook Business Manager
You need to go to your [Facebook Business Manager](https://business.facebook.com/wa/manage/message-templates?ref=internalnote.com) to create a new Template for your WhatsApp account. Go to WhatsApp Manager >> Account Tools >> Templates.
⁉️
Meta is quite strict about using the right type for the right message. ****Marketing** Messages can be send to more people but are a lot more expensive and are ideal for outbound sales campaigns.
Since we only need to alert an existing user about an existing interaction, we can use the ****Utility** category.
1. Create a new *Utility*template for your message.
2. Give the template a clear name like *24h\_reminder*
3. Leave the language to English for now.
4. Add a body text like the one below.
5. I added a Quick Reply button with a "*Keep ticket open"* option, so customers can reply even faster.
6. Submit your template
> Hey,
> Our agent are awaiting your reply. Please respond to keep your ticket open, or ignore to automatically close your ticket within 24 hours
The review process will take 2-3 days and you will get an email once your template is approved. Only when the template is approved will you be able to send out notifications.
# Webhook
Create a trigger or automation based Webhook URL.
The *Endpoint URL* should be to the following URL. Note the `app_id` value needs to be replaced with your App ID.
```
https://api.smooch.io/v1.1/apps/app_id/notifications
```
Leave the *Type* to `POST` and the *Request Type* to `JSON`.
Choose *Basic Authentication* with: Username: `key_id` and Password: `secret_key`
# Setting up the Automation
## Integration ID
You will need an Integration ID to get the automation to work. This will tell Sunshine Conversations via which channel they should send out the notification.
You can find your Integration ID by going to Admin Panel > Messaging > WhatsApp and copying the last part of the URL shown.
```
https://subdomain.com/admin/channels/messaging_and_social/channels_list/edit/646b411abc0c536ba1a98263
```
Replace `[integration_id]` in the payload below with the ID of your WhatsApp integration.
## Conditions
We only want to send out a notification for tickets that match the following conditions:
- `Channel is WhatsApp`
- `Ticket Status is Pending`
- `Tags does not contain 24h_reminder`
This makes sure we only send out the notification once.
- `Ticket hours since update is more than 10`
⁉️
Notifications allow you to send a message to the customer without keeping the 24h rule in mind. But to keep conversations fresh I prefer sending out a messages rather quickly.
## Actions
As a first action choose `Add tag: 24h_reminder` to let the automation know it has run.
Now choose `Notifications - Webhook: Sunshine Conversations Notifications` as the action, and add the following in the JSON body field. Make sure to replace the `[integration_id]` with yours.
🌎
The payload below contains a `"code":"en"` value. If your template has no English translation, make sure to replace `en` with your language. Eg. `fr` or `nl`
```json
{
"destination": {
"integrationId": "[integration_id]",
"destinationId": "{{ticket.requester.phone}}"
},
"author": {
"role": "appMaker"
},
"messageSchema": "whatsapp",
"message": {
"type": "template",
"template": {
"namespace": "XXXXXXXX_XXXX_XXXX_XXXX_XXXXXXXXXXXX",
"name": "24h_reminder",
"language": {
"policy": "deterministic",
"code": "en"
}
}
}
}
```
# The Result
As shown in the screenshot at the start of this article, each time a ticket passes the 24h limit while we're waiting for the customer, that customer will get an automated message, and if they reply, we can extend the 24h limit.
This is only one use case where the new Sunshine Conversations API inside of Zendes Suite becomes useful. Following a similar flow like the one above we could easily use this API to send out appointment confirmations, activation messages, shipping updates or any other kind of utility or marketing based notification to customers.
These interactions don't even need to be based on an existing conversation. If you send out a reminder for a reservation, and the customer replies to your message to make a change, this will create a new conversation with your CX team.
You can find a full overview of the API capabilities here:
[Outbound MessagingSmooch Documentation](https://docs.smooch.io/guide/outbound-messaging/?ref=internalnote.com#notification-api)
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### My approach to setting up Zendesk SLA Policies and enabling Group Policies.
URL: https://internalnote.com/sla-policies/
Last updated: 2025-09-08T06:44:12.000Z
Earlier this month Zendesk announced the availability of Group SLA policies as an extension of the regular SLA policies that existed for customer facing tickets.
SLAs, or Service Level Agreements are a way to measure reply and handling time on tickets and to make sure tickets get replied to in a realistic timing.
[About SLA policies and how they workWhat’s my plan? Fastpath: Admin Center > Objects and rules > Business rules > Service level agreements A service level agreement, or SLA, is a policy yo…Zendesk helpColleen Hall](https://support.zendesk.com/hc/en-us/articles/5600997516058?ref=internalnote.com)
All SLA policies in Zendesk are are based on four elements:
- Priority, ranging from low to urgent as a way to differentiate different tickets
- Business schedule, or differentiating between counting only actual working hours or also continue counting when your agents are offline
- Measurement type, ranging from first reply, next reply to total working time.
- A filter to define which tickets an SLA policy applies too.
# My approach
I like to keep my Zendesk environments rather sparse and easily readable. So when I go and setup SLA policies I've got a few basic steps I take first:
1. Split conversational and ticketing requests
2. Set priority on all tickets based on category or customer type
3. Map all tickets to a specific group
4. Have a schedule for each group
My SLA policies then follow a clear order. For a given group, e.g. Support I have the following two policies:
- SLA Support - Tickets
- SLA Support - Messaging
Within each SLA Policy I filter based on the group and channels, and I always set the operational schedule to Business Hours, cause we can't account for time we're not at work.
And once those policies are made I fill in only First Reply Time (create a sense of urgency) and Next Reply Time (create room for informed replies)
# Preparation
## Define Channels
Each business has their own needs, and each channel has to be treated differently, but in general we can define two types of channels: conversational and ticketing
**Conversational** channels, like chat or Facebook Messenger, tend to be focused on short messages sent in burst. Issues can be resolved quickly, and neither party expects the conversation to take long.
**Ticketing** channels like email are often used for longer, more complex interactions. An email can contain a lot more nuance than a message, and we often even escalate conversations to email ticketing if the issue is too complex or needs to be handled by different teams (e.g. Finance or IT)
For that reason I like to make a difference in SLA between the two types of interactions.
📞
Zendesk Talk or other voice channels fall back to the Ticketing type for me. Calls are to be taken immediately and can't be asynchronous, but the follow-up almost always happens via email or a callback anyway
## Set Priority
Each SLA Policy can contain up to four different targets for each of the metrics, ordered by Priority. I used to create a lot of different SLA policies to handle different ticket types. An SLA for returns, an SLA for complaints, another one for downtime, ... This leads to a lot confusion on which SLA applies when.
To remove this confusion I now always fall back to four triggers that set the Priority for each incoming ticket, with a default of "normal".
That way, each ticket that comes into my Zendesk environment via a specific channel is first prioritised, then assigned to a group and then gets one SLA applied based on the combination of these 3 elements. For regular SLA policies you might be tempted to put these metrics in the conditions of the policy, but for Group SLA policies that isn't possible, and you can only work with Priority. So turned ticket conditions into priorities early one via triggers is worth it and makes working with SLA a lot easier.
⁉️
Did you know you can remove some complexity by editing the Priority Ticket fields? When editing the field you can choose to use four values (low, normal, high and urgent) or just two (normal and high)
I start by creating 4 triggers
The first three triggers are similar. They each contain a list of conditions in the *`Meet ANY of the following conditions`* step that should match that priority.
This can be any combination of elements, like Subject contains urgent, or customer is VIP, or category is GDPR, or any combination of elements. The goal here is to have a few triggers that narrow down your inquiries to a single item: priority.
Once you've got your conditions setup, you add a single action: proirity is urgent (or high, or low).
If the triggers get too complex, feel free to create multiple triggers for specific scenario's.
Once you've created the first three triggers, you can add the fourth one: Priority is Normal. This is our fallback trigger that makes sure if none of the above apply, your ticket gets treated as normal. Cause remember: no priority means no SLA!
Note that I did not add any conditions like *`ticket is updated`* or *`created`* to the triggers. I want these triggers to run on every ticket and update to make sure the SLAs get updated when needed. If you want agents to be able to manually update their priority too, add a *`Priority is not changed`* condition to the triggers.
## Schedules
You should make sure to setup at least one schedule for your Zendesk instance, since SLA's can (and should) be based on Business Hours.
Note that if you have multiple schedules running, you should create triggers that assigns a schedule to a ticket. Otherwise tickets get created without a schedule and the SLA does not get calculated.
# Creating the SLA Policies
Once the above steps are completed, creating the SLA's is pretty straightforward:
1. Go to the Admin Panel and create a new SLA policy.
2. Give it a logical name, eg *`SLA Messaging - Support` or `SLA Ticketing - Support`* to make it clear that this is an SLA that applies to tickets for the Support group and the Messaging Channels (or not)
3. Start by adding one condition to each SLA Policy: `*Channel* is *Messaging*` or `*Channel* is not *Messaging*`. Since *Messaging* is a parameter that contains all Messaging Channels (Twitter DM, Facebook Messenger, WhatsApp, Web Widget,...), and SLA Policies are applies top-to-bottom, with the first applicable rule applying, this will make sure all Messaging tickets get the first SLA, with the rest the other one.
4. Add a second condition: `*Group* is *Support*`, to scope the SLA to that group.
5. Set the `*Hours of operation* to *Business Hours*` so we only count the time agents can really have an effect on tickets.
## Metrics
Now comes the crucial part. Setting up the metrics. I like to keep the data clear for Agent so they know how to interpret the numbers shown.
For Ticketing channels I often only fill in First Reply Time and Next Reply Time. -
- First reply time should be kept short. The faster we acknowledge a problem, the happier the customer is. They know we're working on it.
- Next reply time can be longer. Giving agents time to look into requests in detail results in better responses and less chance of getting many back and forts.
If you're worried that longer next reply times will result in backlog I would recommend resolving this issue at the beginning and put more effort in enabling self service and ticket deflection.
For Messaging Channels these two metrics [don't exist](https://support.zendesk.com/hc/en-us/articles/4408822351642-Limitations-in-messaging-functionality?ref=internalnote.com) (yet). Here the Periodic update option can offer relief. It measures the time between Agent actions, and since messaging tends to be short and quick, we can use this as a good replacement.
Fill in the columns for each priority with the required value. Or, if you have no idea what to fill in, you can go for a 2-4-8-16 metric for first reply time, where urgent ones see a reply in 2 business hours, and low priority tickets see a maximum of 2 working days (measured as 9-5, 8h per business day). Let these SLAs run, and check back in a month to see if they are realistic.
For next reply I often go for a basic 4h for high/urgent and 8h for normal/high before a follow up reply is due.
## The result
If you've followed the above steps you'll end up with a view not too dissimilar from this one, a clear overview of policies ordered by channel and responsible agent group.
What will happen with your tickets is also clear:
1. A VIP customer emails about a lost package
2. We have a trigger that sets these kinds of requests to High
3. And assign these tickets to our Support group
Since this is a request raised via email, and assigned to the group Support, our *`Ticket SLA - Support SLA`* should start. And since the request is Urgent, we need to reply within 4 hours.
Similar, a user sends a message over WhatsApp requesting info on getting a discount on a product. We tag discount requests as Low, and assign it to the Sales team. The *`Messaging SLA - Sales`* starts and we need to reply within 16h, or 2 days.
If a new type of requests requires a specific urgency in replies, you can add that type of requests to the trigger that sets priority as a condition and the SLA will automatically apply. Or if urgent now means "Within 30minutes" for your third level IT group, you can update the *`SLA Ticket - 3th level IT`* policy and update the metric for next reply time for that team.
# SLA based views for a fair order
SLA Policies are only useful when we give Agents a good way to interact with them. One way to make it more intuitive for agents to handle tickets based on priority and make sure they handle the most urgent ones first is to work with SLA based views.
Let's say I have an SLA setup with First Reply times of Urgent: 2h and Normal: 4h
1. A customer emails me with an normal question at 9AM
2. A customer emails me with a urgent question at 10AM
3. A customer emails me with a normal question at 10AM
4. A customer emails me with an urgent question at 11 AM
If we use an ID sorted view, or first in first out based, and I look at this list at 12PM, I might see the tickets 1-2-3-4 sorted as follows:
But how do we handle these tickets fairly once priority comes into play? Does ticket 4 come before or after ticket 1? You might be tempted to sort by priority. But that means the more urgent tickets are created, the lower that first ticket will drop. Not exactly fair for that user, right?
This is where SLA based views are useful. In these views we sort tickets by SLA in an ascending order, meaning the first ticket to breach comes first.
Let's take our four tickets as an example. Taking our priority and SLA policies into account:
1. Needs to be replied to by 1PM (9AM + 4h)
2. Needs to be replied to by 12PM (10AM + 2h)
3. Needs to be replied to by 2PM ( 10AM + 4h)
4. Needs to be replied to by 1PM (11AM + 2h)
If we sort tickets based on their SLA expiration, the tickets will appear in the following order for your agents: 2143 at 1PM.
Urgent tickets still are handled before less urgent ones, but we make sure we handle all tickets taking their SLA into account, preventing tickets getting breach by always picking the first or most urgent one first
Once we also take Next Reply Time into account this logic becomes a bit more complex, but the same rules apply. Tickets where the Next Reply time expired will be intermixed with new tickets but everyone gets threated fairly.
# Group SLA Policies
The inspiration for this article came from the recently announces Group SLAs. These SLAs are, in contrast to the regular SLA policies, not focussing on reply times to the customer, but measure ownership by a specific group of agents.
The setup has a lot less options, but builds upon similar concepts.
- **Conditions** are group based. So you can only measure data based on which group is the current owner of a ticket.
- **Priority** is the same priority as the regular SLA policies look at. Here you see another reason why having triggers assigning priorities and working only with those in SLA policies is worthwhile. Group SLA conditions can't use anything else to set metrics.
- **Hours of operations** are once again the choice between Business and Calendar hours. To keep things consistent, keep it the same as the regular SLAs.
So, going back to our VIP customer who lost their package above: imagine the customer requested a refund for the lost parcel. We could initiate a side conversation via Child Ticket with the Finance Team to get approval for a refund, or assign the ticket to the Finance Team directly. Depending on the Group SLA setup, that team would then see a Group Ownership SLA counter, with a timing depending on the Priority of the ticket.
## Views
When I first setup Group SLA policies I assumed they would intermix with the regular SLAs in views. But, apparently it's a second column you can enable. Which means, if you sort tickets by SLA breach, you have to pick one of them.
## Restrictions
### Assignment vs Escalation
There is one weird behaviour I found with these new SLAs though: there's no way to differentiate between being the primary owner of a ticket, and getting a ticket escalated in these SLA policies.
In a flow where your Support team is a first line support team, and escalates to the Finance team like we showed above, we see that that Finance team gets 30 minutes to ownership time and reassign it to the Support team again.
But what happens if you set a Group SLA for your first line support team?
Imagine most tickets take two days to get resolved. But you set the Ownership to a day. How does that work? Cause the only way first line can fulfill that SLA is by solving in less than a day, and/or escalating. I wonder what Zendesk wants us to do here. Do we don't set a Group SLA for a first line team? Do we set the SLA to the expected resolution time of a ticket?
The same can be said for environments where the IT team is both a first line for colleagues and employees, but second line for customer request related to the website. Setting a Group SLA is logical for escalated tickets. Your CX team wants replies back fast. But illogical for requests from colleagues directly, if IT has a supporting role there the ticket lifetime is probably longer than the ownership time for escalations.
Group SLA Policies are fairly new though, so I can only assume these quirks get ironed out. But if you have any insight in this, feel free to let me know.
One workaround I though of was creating two groups: IT, and IT Escalated. One for first line IT support, and a second group for escalated tickets that takes a Group SLA into account.
Another solution could be custom statusses. Group + Custom Status = SLA. Tickets assigned as `Open` could then be measure one way, `Open via Escalation` could be measured differently. But that would require an expansion of the feature.
### How to handle external escalations?
A second part of escalations that can't be measured is anything that escalates outside of Zendesk.
Imagine your IT team working in Jira and you use the [Zendesk to Jira integration](https://support.zendesk.com/hc/en-us/articles/4408837969946-Setting-up-the-Zendesk-Support-for-Jira-integration?ref=internalnote.com). Or you escalate to your Marketing in Slack, or an external vendor via Side Conversations. I'd assumed that we could measure these kind of escalations too since they fall under how Zendesk describes this feature:
> Tickets often pass between multiple departments or teams on the path to resolution. Group SLAs allow admins to set target times for those groups and separately track resolution times between departments.
However, with groups as the only available condition, this seems impossible to do. You could work with placeholder groups for assignment + escalation and then re-assign to the main group if the ticket gets updates, but that would mean end-user responses are invisible until we re-assign to the main group, which then resets the other groups' SLA. Weird. But then again, this is a first release.
# Additional Notes
## What about specific SLAs for a customer?
Let's say you followed the above steps and configured a similar SLA policy for your instance. But have a few customers whose SLA contract differs from the default. They have unique SLA requirements that apply across all your agents and departments. Or their SLA metrics really differ from your general numbers.
You can still apply the above for all your tickets and create a few specific SLA policies for these specific customers. Just scope them with e.g. `Organisation is Acme Inc` or `Ticket tags contain Gold SLA` .
Just make sure they stay above the other policies so they get applied first, since the first applicable policy gets activated.
All other customers don't apply these policies and will use the normal ones we just build.
## What about customers that have a paid SLA, and other customers?
The same flow can still apply. You could split up the policies to differ between those with a Paid SLA, and the others. It might seem redundant, but since all policies follow the same logic, it's a lot easier to maintain.
- Premier SLA - Messaging - Support
- Default - Messaging - Support
- Premier SLA - Email - Support
- Default - Email - Support
- ...
It's useful to use the clone button in these scenario's to speed things up.
# What's next?
The new Group SLA policies were the incentive to write this article. I like the fact that Zendesk is expanding their Service Level Agreements to be compatible with more scenario's, but I would've loved to have seen a bit more customisability to the Group SLA feature.
Why can't we differ between tickets escalated to a team and owner by a team? How can we measure SLA's for departments working outside of Zendesk we reach via Teams, Slack or the Jira integration? Where's the support for reply-based SLA's for Messaging? I hope (and I'm sure) Zendesk will keep in improving this feature.
If you've got any feedback on my approach, I would love to hear how you handle SLA policies and which scenario's I haven't thought of! Leave a comment below, or feel free to send me an email with feedback!
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### What's new at Zendesk Relate 2023
URL: https://internalnote.com/relate-2023/
Last updated: 2023-10-12T14:07:41.000Z
What WWDC is for Apple, [Relate](https://mc.zendesk.com/mc/zendesk-relate/?ref=internalnote.com) is for Zendesk. A keynote presentation that not only shows what the company has been working on the past year(s), but also sets the tone for the future vision of the company.
And similar to this years' [Google I/O](https://techcrunch.com/2023/05/10/heres-everything-google-has-announced-at-i-o-so-far/?ref=internalnote.com), this keynote can be summarized in three words: AI, privacy and productivity. With the former taking the lead.
Let's dive in!
📥
I compile an overview of Zendesk News every month. If you don't want to miss it, please [subscribe](https://internalnote.com/#/portal/signup/free) to the blog. It's free. (Or optionally paid if you really like it)
Every new subscriber motivates me to keep putting in the effort.
Thanks,
Thomas
The presentation started with an intro by CEO Tom Eggemeier. Not only did he announce he's dropping the Interim from his title (congratulations!), he also set the tone for the presentation and announced Zendesk's new tagline.
> ... it’s also a commitment to deliver on the promise of intelligent CX. So, in addition to bringing great products like Zendesk AI and Conversational Commerce to market, we will transform the way we work.
>
> Today I committed to the following:
> 1\. Renewing our innovation focus
> 2\. Increasing our focus on our own customer experience and doubling down on our commitment to delivering on being “customer first”
> 3\. Providing more expertise in CX with a point-of-view on how customers can improve key customer satisfaction metrics, specifically the ones that impact loyalty while simultaneously improving costs.
## The intelligent heart of the customer experience
Gone is ***Champions of customer service*.** It's replaced with a new commitment that sums it all up:
- *Intelligent*, or driven by AI, ML and data
- *Heart*, sentiment and human connections at its core
- *Customer Experience*. No marketing, no sales, no dev-ops, but customers and agents.
I like the new slogan, just like the new branding it shows a shift from a company focussed on growth and expansion, to a company that's clearly at the top of its game and want to flex its muscles and show its prowess.
Source: Zendesk Newsroom
> In rolling out the new brand identity, we said goodbye to kale, one of our core brand colours, the ‘relationshapes,’ and metaphorical imagery we’d used in the past. The new look focuses on the future and the company we represent, with a few key elements that honour our history. Our new primary colours are espresso and cream, and secondary colours are berry and matcha.
It feels like a more mature but still familiar version of the current branding. Quieter colours, a more serious font, but still the same great Z-logo.
Really curious on how this will impact the product UI. Some marketing screenshots (see below) show a different take on Zendesk's [Garden](https://garden.zendesk.com/?ref=internalnote.com) design, so really curious how the product design will evolve.
# Announcements
Zendesk made three announcements during the presentation, with a few other smaller niceties woven throughout the presentation:
1. Zendesk AI
2. A renewed focus on privacy
3. An adaptable Agent Workspace
# 🤖 Zendesk AI
> What does the customer want and how do they feel?
Naturally, the main focus of the event was Zendesk AI: a new suite of AI and ML capabilities woven throughout the entire product suite.
Zendesk CTO Adrian McDermott positioned AI as a trusted partner, empowering conversational experiences and built on top of your data in a way that combines efficiency and empathy.
He also focused on their new partnership with OpenAI, integrating their generative LLM into Zendesk's core. This allows them to not only use Zendesk AI to detect *intent*, *language* and *sentiment* , but also generate text and summaries based on agent input.
Zendesk trained their AI model on 18 billion tickets and 15 years of ticket history, but then gives each customer a model that's (re)trained on their specific ticket set. This allows them to basically jump-start any customer to get valuable data and content out of Zendesk AI, but will also allow them to guarantee security and privacy. You work with your own data, not others'.
## Intelligence in the Panel
Source: Relate Livestream screenshot
The first big new item was an improvement to the context panel that's shown to the right of your tickets. Zendesk AI will power a new Intelligence Panel.
The panel will show *customer intent, sentiment and language* and allow the agent to also generate a *summary* of the conversation to quickly get up to speed.
That same panel will also suggest a list of macros to the agent, accompagnied by a confidence level to show your agents how sure the system is about using macro.
Last month my team at Premium Plus launched a basic summary app on the Zendesk Marketplace. But I'm really glad app is now a native feature. Not only will native be faster, it's also contained within Zendesk and doesn't need external API's, which is a pro for privacy, and the fact that it's AI is trained on a customers' specific data will make the summary even more precise.
I also think this summary feature will make escalations to second line so much easier. No need to read up on half a dozen comments, but a succinct overview of the current state of the ticket.
## Enhance Ticket Comments
The comment field gets a new button which allows agents to enhance ticket comments . It'll powered by OpenAI's generative AI and allows agents to quickly jot down a solution or response, and then use Zendesk AI to turn it into a good reply for their customer. (in theory)
The Enhance ticket comment feature offers three options:
- **Expand**: turn a short sentence in a longer message with additional context taken from the conversation.
- **Make more friendly**: Turn a factual reply into a more casual or conversational message.
- **Make more formal**: Turn a reply into a more professional tone.
[Using AI to summarize and enhance ticket comments (EAP)Note: The summarization and enhancing features are currently available in an early access program (EAP). You can sign up for the EAP here. You must have Agent Workspace activated to use the feature…Zendesk helpErin O’Callaghan](https://support.zendesk.com/hc/en-us/articles/5608712782362?ref=internalnote.com)
The whole idea of generated emails like [Google](https://www.theverge.com/2023/5/13/23719115/google-ai-help-me-write-communications-email?ref=internalnote.com) does, or now Zendesk, still feels a bit weird to me. I keep thinking we'll end up in an absurd scenario where:
1. *Customer has an issue and writes a three sentence email and uses ChatGPT to expand that into three paragraphs of text.*
2. *Zendesk receives email and an agent uses Zendesk AI to turn email into three sentence summary*
3. Agent *replies in three lines and uses Zendesk AI to turn it into three paragraphs of text*
4. *Customer receives reply and uses ChatGPT to summarise the reply. into three short lines again.*
With a lot of wasted bits in the middle and some nuances getting lost in translation. But I understand the benefits too. We all wrote that short email at 4:55PM before we leave the office which could have been a nicer written email when you reread it the day after.
## Intelligent Triage
Source: Relate Livestream screenshot
> Intelligent triage enables the Zendesk ticket fields for intent, language, and sentiment predictions and confidence, which you can then use for setting up views for teams, automating with triggers or automations, and enhancing Explore reports.
The same values for sentiment, intent and language that are shown to the agent in the context panel, are also available for use in triggers or views:
- **Language** works similar to the existing language filters (although I wonder if it's the same dropdown but with better detection or a full replacement?).
- **Sentiment** is a good, bad, ugly kinda categorisation where you can then sort, assign, or set priority based on the customers' sentiment as detected in the comments' received.
- **Intent** is a bit more tricky. It basically replaces your typical Category dropdown with a set of \~100 predefined intents for specific industries. You can then use these intents to create specific views, or route tickets to specific agents. (If you're outside of these specific industries, you'll get sentiment and language, but no intents if I understand it correctly.). The intents are: Retail and e-commerce, Software platforms or applications and Finance services.
Whenever I set up a Zendesk instance I'm a big proponent of a *Work to be done* approach to views. Have one view with open tickets sorted by SLA with the most urgent one is at the top, and train agents to get that *inbox* to zero by hopefully solving inquiries, or replying and moving tickets to pending.
Intent based views are a completely different approach. Agents work in a categorised based inbox to handle tickets, instead of one sorted by priority. But I'm sure that, by combining both approaches, this new intent training will sure lead to benefits. Can't wait to play around with this one.
## Advanced bots with pre-trained intents
Source: Zendesk Help
> If you give a bot another name, does it still reply the same?
Zendesk used this Relate to [clean up its ticket deflection offerings](https://support.zendesk.com/hc/en-us/articles/5514903812762-Announcing-updates-to-Zendesk-bots-simplified-naming-admin-experience-pricing-and-new-enhancements-?ref=internalnote.com). All Zendesk customers get access to the s*tandard* Zendesk bot that includes a conversation bot (the old Flow Builder) and Autoreply (what used to be Answer Bot).
Customers that have the Advanced AI add-on get access the the a*dvanced* Zendesk bot that includes pre-trained intents and intent tagging.
These pre-defined intents allow you to create dedicated autoreply triggers for specific intents. E.g. if Zendesk AI detects an email about "How to I return a product", you can create a specific email with information that gets send as a reply to those customers. All other customers still get the three suggested articles reply we know from Answer Bot.
[Using pre-trained intents in answersPre-trained intents are a set of industry-specific topics that have been developed by Zendesk using our own anonymized customer service data. By using these pre-trained intents in your bot answers,…Zendesk helpAimee Spanier](https://support.zendesk.com/hc/en-us/articles/5537827011994?ref=internalnote.com)
The advanced bot announcement has two main parts to it;
Part one, the rename, seems logical. The more bots are there on the market, the more having a platform-owned bot that's named Zendesk Bot makes sense. It's similar to Google Maps, Apple TV or Facebook Messenger. If your primary product is also your company name, it's free branding and removes confusion. It's the same reason why, personally, I don't have an issue with the [optional](https://support.zendesk.com/hc/en-us/articles/4408887194778-Hiding-the-Powered-by-Zendesk-logo?ref=internalnote.com) *Powered by Zendesk* footer in the Messaging widget.
The second part is a bit more complex. Where other companies like [Ultimate](https://www.ultimate.ai/?ref=internalnote.com) have intent training and reporting as a core part of the product, Zendesk makes it an add-on. Naturally, Zendesk AI will improve their bots' capabilities, but one might wonder why the standard bot doesn't come with this pre-trained set of intents.
## Conversational Commerce
Source: Zendesk Labs
Jonathan Aniano, SVP Product, started this section of the presentation showcasing a nice extension of the conversation bot which integrates Shopify into Messaging.
Whenever a customer inquires about a product, Zendesk AI interprets the conversation and, when applicable, will show a carousel of products to the customer instead of e.g. a list of Help Center articles.
Or, based on the customer sentiment, they can also automatically offer a discount, or make sure that, if an agent is needed, they can route the conversation to the right agent. That same agent then has the Shopify Zendesk integration running and can quickly update, refund or discount a customers' order.
As a technology demo of what should be possible today, this seemed like a pretty cool flow to have running inside your retail organisation.
However, just like other Messaging demo's we saw at Relate Events these last few years which showcased delivery services, shoe stores and other Sunshine Conversation integrations, I do wonder: how much of this works out of the box, and how much is complex code and how fast is this deployable?
The fact that Zendesk setup[ a new Labs](https://zendesklabs.zendesk.com/hc/en-us?ref=internalnote.com) experience for companies to book workshops makes me think this involves a *bit* more code than the demo let on. However, once it works, this seems really cool and nicely combines Messaging, bots and Agent Workspace.
## Add-ons and Pricing
Source: Relate Livestream screenshot
While writing Internal Note, I try to stay away from pricing and sales, since I'm a developer at heart, and I rather discuss value and impact, than dollars and margins.
But it seems a bit contradictory that, at the same time Zendesk shifted the focus of Zendesk as a product towards being *the intelligent heart of the customer experience*, in reality, everything announced related to Zendesk AI is gated behind a new add-on: *Advanced AI*which cost 50$ per agent/month. And this is **on top** of the just announced [price increase](https://support.zendesk.com/hc/en-us/articles/5555300573850-Announcing-the-Zendesk-2023-pricing-update-What-you-need-to-know?ref=internalnote.com) across all license types.
If have one piece of critique about last week's presentation, it's the fact that **price** was never mentioned, and the word **add-on** was only mentioned once, and even then, just barely.
Giving **every customer their own AI model,** that is continuously trained, gets expensive fast and probably not every customer needs these advanced AI capabilities. So a price increase across the board to cover these costs and give every customer these Advanced AI features might not be feasible or even necessary.
But it's a bit of a shame that "*available today*" means "as an add-on" and not "core part of our offering", and pricing or add-ons were not transparently mentioned in the presentation. One slide with the following would have been nice, and a bit more transparant I think:
Source: Zendesk Pricing
Regardless of pricing or the approach on how they market it, the new offering is pretty straightforward: Zendesk offers two levels of AI functionality to meet your business needs:
- **Zendesk AI**, included with Suite Professional and Enterprise plans
- **Zendesk Advanced AI**, available as an add-on for Suite Professional and Enterprise plans
### Zendesk AI
Nothing new to see here, these existing features are basically rebranded into Zendesk AI.
- Standard bots (formerly known as Answer Bot and Flow Builder)
- Suggested macros for agents
- Content Cues for content managers
### Zendesk Advanced AI
All the new goodies:
- Advanced bots with pre-trained intents
- Intelligence in the context panel
- Enhanced ticket comments
- Intelligent Triage
- Macro suggestions for Admins
# 🔒 Privacy
This slide was the best part of the entire presentation for me.
Source: Relate Livestream screenshot
> Shared commitment to lead the way with new AI trust and privacy standards based on choice, transparency and AI that enhances protection.
## Building Responsible AI
Zendesk outlines its approach to responsible AI in the article below. I'm glad a company takes the time to write about impact, instead of only hyping features.
[https://www.zendesk.com/newsroom/articles/building-responsible-ai/](https://www.zendesk.com/newsroom/articles/building-responsible-ai/?ref=internalnote.com)
What's especially cool is that, if I understand it correctly, the Zendesk AI models are partially powered by OpenAI's models, but run entirely within Zendesk's infrastructure and are (re)trained on a each customers' own dataset. Customers that don't want to contribute to the main models can opt-out, and Zendesk commits to being transparant on how and when it gathers anonymised data to train its AI.
## Advanced Data Protection
In a nutshell: if your company requires their own encryption keys (BYOK), there's now an Advanced Encryption add-on available in EAP that allows you to encrypt (part of) your Zendesk environment with your own keys and broker.
[https://www.zendesk.com/blog/advanced-encryption-zendesk/](https://www.zendesk.com/blog/advanced-encryption-zendesk/?ref=internalnote.com)
# ⚙️ Adaptable Agent Workspace
> What does the customer want? How do they feel?
> Who can help them the fastest?
This brings us to the last part of the presentation: Adaptable Agent Workspace.
In this section we were shown improvements in Agent Workspace that basically turned a set of existing capabilities to 11, empowered by AI.
## Intelligent Triage and Realtime Capacity
Source: Relate Livestream screenshot
The first announcement in this section combines all the development Zendesk has done these last years with regard to skill based routing, agent availability and -omnichannel routing and makes them more powerful thanks to Zendesk AI.
This new set of business rules will allow you to take a conversation and route it to the best agent to handle the case, based on a combination of intent/language/sentiment, SLA, user identity and routing rules.
Implementing this feature, or the current omnichannel routing for that matter, does require a mind shift for your CX team. Where FIFO or longest waiting customer, and assignment to "French team", or "Second Line" are how most teams are setup, this new flow breaks down these big flows and turns each conversation into a unique scenario that routes to agent based on weights and availability instead of groups and generic rules.
## Macro Suggestions for Admins
Suggested macros for agents. Macro suggestions for admins. Two sides of the same coin, and if you weren't confused by reading the sentence once, you can try again with Zendesk's own explainer below. Naming things is hard 😅
> **Tip:** Macro suggestions and suggested macros are different, though related, features:
> **Macro suggestions are suggestions made to admins** about new macros that could be created based on repeated content from all agent replies in your account.
> **Suggested macros are suggestions made to agents** about which existing macro to apply to a ticket based on the content of that specific ticket.
All kidding aside, this is a cool feature. Zendesk AI indexes all your tickets, and based on Intents that occur regularly, and agent comments on those tickets, Zendesk will suggest new macro's to your admins with pre-generated text (thanks OpenAI!) and they can approve and deploy these new macros with a click of a button.
It's not clear if these macro's are just text, or if they also include status or other custom fields.
## Sneak Peak: Admin Home
Source: Relate Livestream screenshot
To conclude the announcements for Agent Workspace we got a sneak peak at the new Admin Home and the new Spike Alert feature.
In a nutshell, whenever Zendesk AI detects a spike in tickets related to an issue (say: Shipping delays or server downtime), the Admin Home will show a list of proposed things the Admin can implement to deflect tickets, help customers and lower the load on your agents:
- **Al-generated workforce management**
Zendesk will automatically assign agents to specific channels and intents based on their skills so the best agents are assigned to those tickets to handle these tickets efficiently.
- **Al-generated knowledge base article**
Based on Agent replies on those tickets the system will not only suggest macro's but will also write a Help Center article to publish, powered by OpenAI
- **AI Generated Layout builder**
The Agent Workspace layout will adapt its UI to give agents looking at these kind of tickets the necessary apps and context. If you get a spike in order inquiries, the shopify app will be front and center. When user authentication goes down, your app to validate user profiles will be visible,...
It's clear that, at least for customers buying Zendesk Advanced AI, it's clear we're moving away from the old triggers and views based Zendesk to a more automated system that basically does for Admins and Agent with managing tickets, what Zendesk Guide and Self Service does for Customers: proactively offer the right information and tools to prevent and/or easily resolve issues before they escalate. I'm really curious where they take this.
Also, this UI looks and feels entirely different from what we know and recognise as Zendesk today. It's probably part of the new style and branding, and I'm curious how this will trickle down to the Agent Workspace and ticket interface.
## Sunshine Platform
Source: Relate Livestream screenshot
Bandar El-Eita, Senior Director of Product Marketing, started the Adaptable Agent Workspace section of the presentation with this overview of the Zendesk Platform.
I might be reading things that are not there, but I sure hope the fourth ring shown on the slide means that, sooner or later, we can developer against `https://subdomain.zendesk.com/api/ai` and use Zendesk' AI to build integrations and custom apps powered by a customers' own data and without sharing any of their data outside of Zendesk.
One can hope...
# 🥳 Conclusion
Three thousand and something words later, and I've barely scratched the surface of what Zendesk announced this Relate. As a fan of the product, I can't wait to play around with these new features. And as CTO of a [Zendesk Premier Partner](https://premiumplus.io/?ref=internalnote.com), I've already started mapping features to customers and see who could benefit from what.
The (lack of) clarity about price and value in the presentation will surely turn a few initial smiles into frowns. Nobody likes to be shown nice things without up front risks and costs, but once the dust settles and Zendesk AI becomes the new normal, these ML powered capabilities will change the way customers and agents interact and work.
Secondly, I do wonder how Zendesk as a product will evolve from here on after. It's clear that Zendesk is no longer the company of tickets and self service, but is now a company focussed on conversations and automations at a human level, deeply integrated with AI. But similarly to how an iPhone user has arguably a worse experience without an Apple ID, or without paying for iCloud storage, I wonder if Zendesk AI is the start of a similar schism. Customers with AI enhanced environments, and legacy customers' without. Clearly, Zendesk's argument here is that the former will deliver the better CX experience of the two.
But I wonder if, in the future, those Advanced AI customers will get a better Zendesk product experience too. How much product R&D focus will go to the old way of views, triggers and custom fields, and how much product attention will go to Zendesk AI? Only time will tell, and Zendesk AI is only a first big step into the larger world of Intelligent CX.
Really curious where they take it from here.
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### An in-depth overview of Proactive Messages for Zendesk
URL: https://internalnote.com/proactive-ticketing-for-messaging/
Last updated: 2024-08-19T20:36:26.000Z
Last month Zendesk [launched](https://internalnote.com/monthly-note-for-april-2023/) their Proactive messages for Zendesk Messaging. This new feature allows you to reach out to customers by showing a small popup on top of your Zendesk widget. If the customer clicks on the message or widget, you can then immediately route them to the right [intent](https://internalnote.com/flow-builder-dinosaurs/) and try to proactively offer them solutions to any questions they might have.
[Announcing Proactive Messages in MessagingRollout start Rollout end March 27, 2023 March 30, 2023 What’s changing and why? We are thrilled to announce, launch of Proactive messages for our messaging customers. It’s our latest additi…Zendesk helpArpan Nagdeve](https://support.zendesk.com/hc/en-us/articles/5540317327770-Announcing-Proactive-Messages-in-Messaging-?ref=internalnote.com)
Proactive messages basically offers the same functionality as the old Zendesk Chat [triggers](https://support.zendesk.com/hc/en-us/articles/4408842880282-Zendesk-Chat-triggers-conditions-and-actions-reference?ref=internalnote.com), albeit with a lot less options. The new feature misses visitor location, time stamps, page view count, tags or reacting to a customer initiating the chat.
What we lose in functionality, we win in ease of use. The new Admin UI is a breeze to use with a modern UI, a better flow to setup the messages and build-in reporting.
Proactive messages are currently available for the [Web Widget](https://support.zendesk.com/hc/en-us/articles/5511266103834?ref=internalnote.com) and [Mobile SDKs](https://support.zendesk.com/hc/en-us/articles/5511216991898?ref=internalnote.com).
In this article we'll dive into the main feature set, and then go into more advanced flows. One more advanced flow allows you to show proactive messages to customers that match a specific marketing campaign, based on UTM Campaigns tags.
We build a specific flow based on a customers' locale, and combine Messaging authentication and tags to offer a premier experience you your VIP users.
Let's dive in.
# Feature Preview
You can test out the flows described below on this page.
[https://demo.internalnote.com/proactive](https://demo.internalnote.com/proactive?ref=internalnote.com)
Some flows use specific code to make them work, especially the logged in user scenario. Take a look at the repository to see how it works:
[GitHub - verschoren/zendesk\_widget: Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand.Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand. - GitHub - verschoren/zendesk\_widget:…GitHubverschoren](https://github.com/verschoren/zendesk%5Fwidget?ref=internalnote.com)
# Let's start with a basic Setup
We'll start with a pretty straightforward example. A customer visits our sample page and looks at the homepage for 30 seconds.
They're greeting with a welcome message that then shows info about this website.
The way it's setup is as follows:
## Compose the message
1. We first create a new proactive trigger and assign it to the right brand and channel.
2. We then Set a proactive greeting and select the `Bot` option.
3. We already build an Intent "About Internal Note" which we can use for this trigger.
## Triggers
Next up we select a trigger. Triggers are basically the rules that define when a trigger will fire. We can select both Visitor based options (title, url, repeat visitor), or Device based rules (Language).
For this basic example we'll set a rule as follows: `Visitor.Page URL is https://proactive.internalnote.com` . This will match our homepage exactly, and will make sure it won't fire on subpages. If you change the trigger to `contains at least one of` the trigger will fire on any page.
## Frequency and Timing
Once we setup that trigger we can then choose when we want the message to appear.
- We can choose to show the message upon payload or after a delay
- We can choose between always and outside or inside business hours
- We can choose to show the message once per user, once per session or each time.
For this basic demo I chose to show the widget after 30 seconds, ignore business hours and show it once per session. You can test it out on [https://demo.internalnote.com/proactive](https://demo.internalnote.com/proactive?ref=internalnote.com)
# Advanced Flows
In this next section we'll build three advanced flows:
1. A way to target customers based on marketing campaigns.
2. A premier experience for your VIP customers build on top of Authentication.
3. Handle customer locale for better assignment and routing.
# Target based on URL Parameters
Sometimes just targeting based on the page path is a bit to generic. Luckily the `Visitor.Page URL` rule is flexible enough to even target specific tags and parameters in the URL. Which opens a world of possibilities:
- You can a marketing campaign and want to target users with a specific `?utm_source` or `?utm_campaign` tag.
- You run a webshop and want to target visitors that are looking at a very specific variant `?variant_sku=1977`
## Example
I have two example product pages on my demo environment:
- A regular Product page that most customers visit - [link](https://demo.internalnote.com/proactive-product.html?ref=internalnote.com)
- The same product page but tagged with an `utm_campaign=trex` tag. This tag is for example added when customers arrive at the page from a specific ad. - [link](https://demo.internalnote.com/proactive-product.html?utm%5Fcampaign=trex&ref=internalnote.com)
The regular page shows the proactive message on the left, whereas the customers matching my campaign get a specific promo message.
## Setup
If you only wish to target the `campaign` users, you can create a proactive flow that has the following rules:
- A rule to match the campaign tag: `Visitor.Page URL contains at least one of utm_campaign`
- A rule to restrict only to product pages: `Visitor.Page URL contains at least one of product`
If you want to create proactive message per campaign you can create multiple proactive messages which each exactly mapping to one campaign tag:
- Proactive Message #1: `Visitor.Page URL contains none of utm_campaign=trex`
- Proactive Message #2: `Visitor.Page URL contains none of utm_campaign=raptor`
In my demo I created a generic trigger for product pages that don't contain a campaign tag, and a custom one for the T-Rex campaign.
# Flow 3: Offer a premier experience to VIP Users
The next flow expands on the custom URLs flow but combines it with [Messaging authentication](https://internalnote.com/jwt-messaging/).
## Example
We have a website where customers can login with their account.
Once logged in we have two types of users:
1. Our VIP users - [link](https://demo.internalnote.com/proactive-contact?ref=internalnote.com#vip)
2. Our regular users - [link](https://demo.internalnote.com/proactive-contact?ref=internalnote.com)
Our webmaster made it so that whenever a VIP user logs in we get a unique hash added to the URL `#vip`. We can now setup a rule that checks for the `hash` in the URL and immediately shows a proactive message that will contact the VIP Customer to an Agent, instead of the Chatbot.
## Setup
This flow is a bit more complex though than just another URL rule.
The page also makes use of the [Authentication Flow](https://internalnote.com/jwt-messaging/) of Messaging. This makes it so that the logged in VIP users not only get a notification we're there to help them, we also:
- Add a tag to those tickets `VIP` so we recognise them in Support
- We authenticate so the user is not asked for a name or email
- If any conversation was already active, they can just resume their session.
⁉️
There's currently a bug where, if you select the `Agent` flow for a Proactive message, and have an authenticated user, the message will still asks for details but not show the fields for the user to fill in
Regular users have the benefit of being logged in too on this page, but aren't prompted for support and will have to pass the Chatbot first.
⁉️
It would be awesome if we could somehow have "Logged In" as a rule condition for Proactive Messages, this would remove the need for the `#VIP` tag.
# Flow 4: Assignment based on locale
This final flow will trigger a specific pro-active message for french speaking customers and will use a `Device.Language` trigger condition.
See [this page](https://demo.internalnote.com/fr?ref=internalnote.com) to test out this flow
## Landing Page
When you have a webpage that serves multiple countries you probably also have a language switcher somewhere on your website that sets the locale of your website. Most often this happens with a `https://domain.com/fr/` path.
Whenever you setup such a flow, it's also useful to force the Zendesk Widget to follow that locale too by adding to your website just below the widget embed code.
```javascript
zE('messenger:set', 'locale','fr')
```
## Setup the Proactive message
Once we have a website that is localised, and (optionally but advices) a widget that follows that translation, we can create a rule set that uses the `Device` conditions instead of the `Visitor` conditions to trigger a proactive message.
In this scenario we filter on `Device.Language is fr` to react to pages for the french market. Note that we also add a tag `Messaging_French` to these conversations.
## Setup an assignment Trigger
Thanks to the trigger, all conversations that start in french will now be tagged with a specific tag, and we can use a trigger to route them to a dedicated team.
🇫🇷
You could also ignore the proactive message tag and just route based on the requester language.
# Reporting and impact
The Proactive messaging feature in Zendesk has a nice overview in the Admin Panel to monitor the effectiveness of your campaigns. For each proactive message you can measure the sends, opened and replies in a nice overview.
Personally I find it a bit weird that this data is not (yet) a part of Explore, but thanks to the tags used on the tickets created, you can still pull in data into Explore and get some insights in your campaigns there:
# Limitations
The examples above show that there's quite a lot of complex flows that can already be build with Proactive messages.
However, while playing with the tool, some [limitations](https://internalnote.com/zendesk-messaging-feedback/) did come up.
⁉️
When using the [Custom Launcher for Messaging](https://internalnote.com/custom-launcher-for-zendesk-messaging-and-unread-counts/), proactive messages are ****not supported!** See [this](https://support.zendesk.com/hc/en-us/articles/5381304334234/comments/5575083752986?ref=internalnote.com) comment.
## Rules and Conditions
One item I would love is to get access to a users' logged-in state to trigger different messages. This would make the VIP flow above a lot nicer, or would prevent us from annoying logged in customers, while still targeting visitors. similarly, being able to reference the name of the logged in user would be nice too.
Alternatively, rules would benefit nicely from allowing regex matches for more complex parsing and the ability to read localStorage cookies to have more nuanced triggers.
## API
Another omission I find curious is the complete lack of any API or developer toolkits for this.
I would love to be able to trigger a proactive message or intent client side via e.g. `zE('messenger:proactive', 'message', 'type:intent','tag')`
Similarly, a way to create, update, (dis)able or delete proactive messages via API would make it possible for tools like Hubspot to schedule and deploy campaigns automatically.
Similarly, I'd love to be able to schedule messages. E.g. launch our proactive campaign next Monday.
## Tags
While exploring flows for this article one of the ideas I had was to create a few Proactive Messages tagged with distinct tags (e.g. apple, pear, orange).
My idea was to combine those tags with the new Conditional Step in Flow Builder to automatically select the right path based on the assigned tag.
Sadly, Tags are only applied once a conversation starts, so there is no way for pro-active messages to filter intents.
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### Zendesk Roundup for April 2023
URL: https://internalnote.com/monthly-note-for-april-2023/
Last updated: 2023-10-12T14:07:30.000Z
> Offline is just online with extreme latency.
This quote does a better job explaining Messaging and Conversations than any other quote ever could. I found it in someone's LinkedIn post a while back and it really resonated.
With social channels specifically, but similar to email or webforms, it doesn't really matter if your company is active 9/5, 24/7 or if it's a public holiday. Customers have questions 365 years a day, and with the combination of Self Service, Answer Bot and Intelligent Triage you can make sure that customers can find some answers always, and that *if* an agent is active, they get the right kind of tickets assigned.
Similarly, with the new Teams and Slack integrations, escalating to team members becomes async too. You can be sure someone sees and will reply to your question, but until they do, that ticket can be safely in an *On Hold* status. Out of sight, out of mind.
Anyhow, on to the updates!
📥
I send out this overview of Zendesk Related News every month. If you don't want to miss it, please [subscribe](https://internalnote.com/#/portal/signup/free) to the blog. It's free. (Or optionally paid if you really like it)
Every new subscriber motivates me to keep putting in the effort.
Thanks,
Thomas
# 🎉 New Releases
## Proactive Messages for Messaging
Already mentioned in [What's New March 2023](https://internalnote.com/link-list-for-march-2023/), and now available: proactive conversations. I've got the featured enabled for all [Flow Builder](https://internalnote.com/tag/flow-builder-and-chatbot/) articles on this website. When you look at an article, Messaging will offer you the intent the article describes as a suggested conversation.
The fact that the feature combines both a dashboard to get insights, the ability to select specific intent and tag the conversations makes targeting the right customers and assigning to the right teams a breeze.
[Announcing Proactive Messages in MessagingRollout start Rollout end March 27, 2023 March 30, 2023 What’s changing and why? We are thrilled to announce, launch of Proactive messages for our messaging customers. It’s our latest additi…Zendesk helpArpan Nagdeve](https://support.zendesk.com/hc/en-us/articles/5540317327770-Announcing-Proactive-Messages-in-Messaging-?ref=internalnote.com)
## Custom Headers in Webhooks
Webhooks now support custom headers and API key authentication. This makes it [possible](https://internalnote.com/custom-authentication-for-webhooks-update/) to integrate Zendesk with a lot more platforms, like Zendesk's own Sell, Asana, Cloudflare or other APIs that require very specific headers.
[Announcing API key authentication and custom headers for webhooksAnnounced on Rollout starts March 30, 2023 March 30, 2023 Zendesk is excited to announce that you can now use API key authentication and define additional headers when configuring webhooks.…Zendesk helpZach Anthony](https://support.zendesk.com/hc/en-us/articles/5532092885658-Announcing-API-key-authentication-and-custom-headers-for-webhooks?ref=internalnote.com)
## Webhooks for Guide
Last year Zendesk announced *event based triggers* for Webhooks that focussed on users, organisations and groups. Now they've expanded these events to also include Help Center and Community items like article published, comment posted,..
This allows for some cool use cases like [announcing new articles on Slack](https://internalnote.com/webhooks-for-guide/).
[Announcing webhooks for help center and community eventsAnnounced on Rollout on April 26, 2023 April 26, 2023 After launching the ability to set up webhooks that can receive Zendesk events last year, we’re excited to release support for initiating…Zendesk helpZach Anthony](https://support.zendesk.com/hc/en-us/articles/5590028739738-Announcing-webhooks-for-help-center-and-community-events?ref=internalnote.com)
## Lookup relationship fields available in Explore
[Lookup relationship fields](https://internalnote.com/lookup-fields-and-ticket-escalation/) were introduced last year as a way to link Zendesk Tickets, Users and Organisations. You could assign an Account Manager to an organisation, link customers to a Sales Rep, or link a Vendor to a specific ticket.
These mappings are now available in Explore so you can create reports based on these mappings.
[Announcing general availability of the Agent Availability APIsAnnounced on Rollout on July 21, 2023 July 21, 2023 Zendesk is excited to announce general availability of the Agent Availability API. What is changing? Previously, agent statuses could…Zendesk helpVolkan Akdugan](https://support.zendesk.com/hc/en-us/articles/5518158586522-Announcing-general-availability-of-the-Agent-Availability-APIs?ref=internalnote.com)
## Custom Objects EAP
This might be the biggest news this month. The new Custom Objects Early Access Program (EAP) is active, and if you've signed up, chances are you have access.
The new Custom Objects are basically Lookup Fields on steroids and allow for some very cool and extensive custom flows right within your Zendesk. I played around with earlier and managed to [build a full Pokédex](https://internalnote.com/creating-a-pokedex-with-zendesk-custom-objects/) inside of my Zendesk account. Take a look at the video if you've missed it!
[What is the new Custom Objects EAP?Zendesk provides many types of native data objects for storing and managing your customer data, including tickets, users, organizations, and more. We call these standard objects. However, standard…Zendesk helpAshwin Raju](https://support.zendesk.com/hc/en-us/community/posts/5359269969178-What-is-the-new-Custom-Objects-EAP-?ref=internalnote.com)
Since the EAP is still in active development, I've started compiling a list of remarks. Most of them also link to a Zendesk Community Post so if you agree with some of these, please add a comment or upvote the post!
[Custom Objects EAP FeedbackBelow is a list of feature request and remarks for the Custom Objects EAP.Internal NoteThomas Verschoren](https://internalnote.com/custom-objects-eap/)
## All the small things
- You can now search and filter on images in the account level gallery for content blocks. - [link](https://www.google.com/search?client=safari&rls=en&q=zendesk+guide+filter+saarch+images&ie=UTF-8&oe=UTF-8&ref=internalnote.com)
- Views introduced a [new pagination experience](https://support.zendesk.com/hc/en-us/articles/4420490020890?ref=internalnote.com)
- The Developer Portal now has a [Beta and EAP](https://developer.zendesk.com/api-reference/betas/introduction/?ref=internalnote.com) section.
# 💡Insights
Zendesk posted a nice insight in [Build vs Buy](https://www.zendesk.com/blog/build-vs-buy/?ref=internalnote.com) on their website. As CTO and Developer I also find this a difficult balance to strike. Do I let one of my developers develop an internal tool? Or do we buy of the shelve. It's a balance between long term maintenance cost vs short term investment.
> **Zendesk adds OpenAI integration to expand AI-powered customer experiences**
> Combining the industry leading capabilities of the Zendesk Suite with the power of OpenAl helps businesses deliver a more intelligent customer experience while saving time and money - [Link](https://www.zendesk.com/newsroom/articles/zendesk-openai/?ref=internalnote.com)
Next month it's [Zendesk Relate](https://event.zendesk.com/zendeskrelate2023broadcast1?partner%5Fcontact=0038000002FnSkNAAV&partner%5Faccount=0018000001RVSYfAAP&utm%5Fmedium=partner%5Femail&utm%5Fsource=Verschoren&utm%5Fcampaign=internalnote) time, and I feel Zendesk has shifted even more of their communication about the event towards AI. And what I find really curious is the way they're leaning into OpenAI. They already had a lot of ML knowhow in house thanks to the purchase of [Cleverly](https://techcrunch.com/2021/08/26/zendesk-acquires-ai-automation-startup-cleverly-to-advance-customer-service/?guccounter=1&guce%5Freferrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce%5Freferrer%5Fsig=AQAAAIeIAgFSxaWe00swk-r9zjCwY0Qp3Q0iPLfMyn7-uuHFfm4a4MclNCVpNZ4unlaYa-O8hLaoAOwq483KRGzXu53d3fFVFzqpzxEkLcq090QgtNGEUUC%5Fom9c-Rx1qVXjoS1Ux5qIt698vgetuqCVDc-YvKYDjI1AOcrpGVnusMFo&ref=internalnote.com) last year, so I feel the mention of OpenAI is more a way to spark interest in the event and to catch the attention of everyone following the AI Hype, than an actual "O-o we better start doing something with this too"
I'm really curious about the full picture of new capabilities they will announce at Relate, but from what I've seen it's going to be pretty cool. 😉
# ⚠ Major Changes
## Messaging API connections
Starting May 31, 2023, you can only use API connections to authenticate REST API calls from a messaging bot. This article walks you through the process of updating your bots to use API connections. See [Updating messaging bots to use secure API connections for API calls](https://support.zendesk.com/hc/en-us/articles/5469553859866?ref=internalnote.com).
## Announcing the removal of legacy Slack integration notifications
If you’re using the existing Slack channel notifications, you’ll need to [migrate](https://support.zendesk.com/hc/en-us/articles/5494222584346-Announcing-the-removal-of-legacy-Slack-integration-notifications?ref=internalnote.com) these to triggers by September 18, 2023\. After that date, these notifications will cease to notify Slack channels.
## Announcing plan-based product limits for images in Guide
The introduction of [plan-based product limits](https://support.zendesk.com/hc/en-us/articles/5603632694042-Announcing-plan-based-product-limits-for-images-in-Guide?ref=internalnote.com) for images in Guide means that the number of images you can upload and use in Guide now depends on the Guide plan you are on:
# 🎥 Videos
# 🗺️ Ecosystem
Not only Zendesk (and the entire world) is playing around with ChatGPT these days. Ultimate, a major player in the Ticket Automation space, showed off a demo of their new *UltimateGPT* Chatbot.
[Meet UltimateGPT: The LLM-Powered Bot to Revolutionize Your SupportIntroducing UltimateGPT, a groundbreaking new product that integrates the power of ChatGPT into your support center and simplifies support automation.ultimate.aiUltimate](https://www.ultimate.ai/blog/ultimate-life/meet-ultimategpt-the-llm-powered-bot-to-revolutionize-your-support?ref=internalnote.com)
# 📝 Articles this month
- [Deep-dive into issues related to Messaging Authentication and JWT User Mapping](https://internalnote.com/deepdive-into-messaging-profiles/)
- [Learn how to build a full-featured Flow Builder Bot for Zendesk.](https://internalnote.com/flow-builder-dinosaurs/)
- [Preview: Creating a Pokédex with Zendesk Custom Objects](https://internalnote.com/creating-a-pokedex-with-zendesk-custom-objects/) (Online only, not mailed)
- [Improve Your Content Discoverability with a Linked List in Zendesk Guide](https://internalnote.com/linked-list-for-guide/)
- [Expanded API Support via Custom Authentication for Zendesk Webhooks](https://internalnote.com/custom-authentication-for-webhooks-update/)
- [Publish new Articles to Slack via the new Webhooks for Zendesk Help Center](https://internalnote.com/webhooks-for-guide/)
# And finally...
Do you know the power of `:` in Zendesk?
If you use it in a Macro title, it'll break up the title into a Category and Title and Macros will show up as a menu with subsections in your Agent Workspace. Like this: `Ticket Closings::Glad it worked`
Similarly, doing the same in a Dynamic Content will activate a hidden `type` filter in the UI. And Typing `:` in any comment field will trigger an emoji dropdown! 🤯
🥳
Thanks for reading this article and the blog. If you liked this content, please consider [****subscribing**](https://internalnote.com/#/portal/signup) and ****share** this article to your colleagues.
### Publish new Articles to Slack via the new Webhooks for Zendesk Help Center
URL: https://internalnote.com/webhooks-for-guide/
Last updated: 2025-09-08T06:42:53.000Z
Zendesk announced webhooks for their Help Center this week. A pretty cool new expansion on the existing Zendesk Event Webhooks feature introduced last year. Building on top of the existing users, agent, groups and organisation events, you can now get notified for changes in your articles (or community).
[Announcing webhooks for help center and community eventsAnnounced on Rollout on April 26, 2023 April 26, 2023 After launching the ability to set up webhooks that can receive Zendesk events last year, we’re excited to release support for initiating…Zendesk helpZach Anthony](https://support.zendesk.com/hc/en-us/articles/5590028739738-Announcing-webhooks-for-help-center-and-community-events?ref=internalnote.com)
You can get notified on the following Help Center events:
- Any help center article events
- Article published and/or unpublished
- Article subscription created
- Article vote created/changed/removed
- Article comment published/unpublished/created/changed
And for each event you get a POST request to a webhook endpoint of your choice:
```json
{
"account_id": 10168721,
"detail": {
"brand_id": "224348602",
"id": "7005414006398"
},
"event": {
"author_id": "225859532",
"category_id": "7005437869694",
"locale": "en-us",
"section_id": "7005437870078",
"title": "James Bond Gadgets"
},
"id": "01GX4P86AC7T0AFF86QGAHTEFR",
"subject": "zen:article:7005414006398",
"time": "2023-04-03T23:11:49.571545199Z",
"type": "zen:event-type:article.published",
"zendesk_event_version": "2022-11-06"
}
```
# What's possible now?
I used to follow all sections on my Help Centers with a dedicated end-user and then monitor that users' mailbox for changes. I wuld then use Zapier or Cloudflare Workers for Email to automate flows based on received emails to monitor and then automate based on changes in Guide.
These new webhooks allow me to tear down that Rube Goldberg Machine of an automation flow and use cleaner and more efficient code.
Examples:
- Notify your teams via Slack about article publications
- Post new articles to Twitter, Facebook or other socials via Zapier
- Notify your Marketing team so they can add it to release notes
- ...
In this article we'll build a quick example that pushes any new article to Slack to notify your internal teams.
☕
If you like these kind of implementation demos of new Zendesk features, please consider [subscribing](https://internalnote.com/#/portal/signup) to this newsletter. It's free. (Or optionally paid if you really like it)
Every new subscriber motivates me to keep putting in the effort.
Thanks,
Thomas
# Webhook JSON Payload
The payload you receive contains a few elements:
- `"type": "zen:event-type:article.published"` defines the type of event pushed.
- `"event": {...}` contains the author, category, section, locale and title of the article.
- `"detail": {...}` contains the brand (useful for multibrand environments) and article ID
# Post new articles to Slack
## Setup a new Slack app
Any integration that posts messages to Slack requires a Slack App.
[An introduction to the Slack platformThe Slack platform allows you to you extend and automate your workspaces to cultivate conversation, inspire action, and synergize services.Slack APISlack](https://api.slack.com/start/overview?ref=internalnote.com#creating)
Start by going to [https://api.slack.com/apps](https://api.slack.com/apps?ref=internalnote.com) and create a new app.
- You need to configure the **Webhook** Feature and create a new Webhook that points to a channel of your choice.
- In the **Basic Information** settings you can give the app a Name, Logo and description.
Once you setup a Webhook you'll get an URL you can use in the next step:
```
https://hooks.slack.com/services/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
```
## Create Message Payload
To create a nicely formatted message you can use Slack's [Block Builder](https://app.slack.com/block-kit-builder/?ref=internalnote.com) to build a nice message with buttons and text.
```json
{
"blocks": [
{
"type": "section",
"text": {
"type": "mrkdwn",
"text": "New article published on our Help Center"
}
},
{
"type": "section",
"text": {
"type": "mrkdwn",
"text": "Title"
},
"accessory": {
"type": "button",
"text": {
"type": "plain_text",
"text": "View Article",
"emoji": true
},
"value": "article",
"url": https://internalnote.com,
"action_id": "button-action"
}
}
]
}
```
## Setup a Cloudflare Worker
You could handle the incoming webhook in a variety of platforms, but I prefer to use Cloudflare Workers since they are free, fast and easy to use (and can be run in a [Carbon Neutral Green mode](https://internalnote.com/webhooks-for-guide)).
You can find the full code in the Repository below, or follow along to learn how it works.
[GitHub - verschoren/helpcenter\_webhooks: Capture incoming Help Center Webhooks and push them to SlakcCapture incoming Help Center Webhooks and push them to Slakc - GitHub - verschoren/helpcenter\_webhooks: Capture incoming Help Center Webhooks and push them to SlakcGitHubverschoren](https://github.com/verschoren/helpcenter%5Fwebhooks?ref=internalnote.com)
⚠️
If you're not familiar with Cloudflare Workers you could setup the same flow with Zapier. The downside is that Zapier Webhooks (needed to capture the webhook from Zendesk) are a Premium feature.
The worker is pretty straight forward. We first retrieve the `POST` JSON payload Zendesk sends to the Worker.
Next we retrieve the `article ID`, `title`, and `locale` from the payload. We then build a full `url` for the article.
```javascript
export default {
async fetch(request, env) {
const { url } = request;
const article = await request.json();
var article_id = article.detail.id;
var article_title = article.event.title;
var locale = article.event.locale;
var base = 'https://support.internalnote.com/hc/';
var full_url = `${base}${locale}/articles/${article_id}`
var slack = await postToSlack(full_url,article_title);
return new Response('pushed to Slack');
}
}
```
Once we have all the variables we need, we insert them into our message payload and post our message to Slack using the `webhook URL` we got from creating a Slack App earlier.
```javascript
async function postToSlack(full_url,article_title){
var message = JSON.stringify({
"blocks": [
{
"type": "section",
"text": {
"type": "mrkdwn",
"text": "New article published on our Help Center"
}
},
{
"type": "section",
"text": {
"type": "mrkdwn",
"text": article_title
},
"accessory": {
"type": "button",
"text": {
"type": "plain_text",
"text": "View Article",
"emoji": true
},
"value": "article",
"url": full_url,
"action_id": "button-action"
}
}
]
});
const url = "https://hooks.slack.com/services/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";
const init = {
body: message,
method: 'POST',
headers: {'content-type': 'application/json',},
};
const response = await fetch(url, init);
return response;
}
```
## Webhook
Now we can finally link all items together and start using the new Webhook functionality in Zendesk:
1. Go to the Admin Center and choose *Apps and Integrations*. Click on Webhooks
2. Create a new webhook by clicking the button top right
3. Choose *Zendesk Events* and select *Article Published* from the list of options
4. Press Next
5. Enter a Name, Description, and the URL of your Worker. Ignore all other settings.
6. Press Test. You should get a Slack Notification.
7. Once the test runs successful press *Create Webhook* to enable the flow.
## Result
Every published article in Guide will now become a Message in a Slack Channel of your choice!
# So, what's next?
I really like these kind of additions to the Zendesk platform. They make integrating Zendesk into a company or workflow way easier than before. You could imagine flows where a downvoted article generates an alert for a Guide Admin to take a look, or where comments ping a moderator to take a look.
Looking further I'd really love to see the Zendesk Enterprise [approval flows](https://support.zendesk.com/hc/en-us/articles/4408829231770-Reviewing-approving-and-publishing-articles-with-Team-Publishing?ref=internalnote.com) be available as Events too. Instead of emailing an author about approving changes, you could ping them in Slack, or add a task in Asana for them to take a look at the article.
There's always a [wishlist](https://internalnote.com/zendesk-wishlist/) for these new features ;-)
So what are you building?
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** via email or ****share** the article to your colleagues.
### Automatic Ticket Categorisation and Sentiment Analysis via ChatGPT
URL: https://internalnote.com/chatgpt-categorization/
Last updated: 2025-09-08T06:42:49.000Z
These last few months ChatGPT and Generative AI have been all the hype and new use cases for this fun API have been popping up left and right.
Zendesk itself has also picked up these new capabilities and has been actively blogging about its possibilities [here](https://www.zendesk.com/blog/generative-ai/?ref=internalnote.com) and [here](https://www.zendesk.com/blog/zendesk-ai/?ref=internalnote.com).
And last week they also promised some major announcements in this space on their [Relate Event](https://mc.zendesk.com/mc/zendesk-relate/?ref=internalnote.com) in May.
Even though there is a lot of noise and hype around this tool, in the end it is an API, and like any technology, it's more important thinking about how it can fit in your processes and improve them than to think about the technology itself.
> “The question you have to ask yourself is, what is your application designed to do? What problem are you looking for the LLM to solve? Different applications have different requirements and different risks."
> *Jaakko Pasanen, Chief Science Officer, Ultimate*
According to Ultimate, a leading ChatBot and Automation provider, there's four ways Generative AI like ChatGPT can assist in Customer Care, as highlighted in a recent article they posted on their website.
[How to Use ChatGPT for Your Customer Support: Challenges to OvercomeThe challenges and opportunities of LLMs and generative AI in customer service, based on real examples.ultimate.aiGesche Loft](https://www.ultimate.ai/blog/humanized-ai-how-to-use-chatgpt-for-your-customer-support-challenges-to-overcome?ref=internalnote.com)
## The four main impact points
1. **📝 Summarise** support tickets
2. **🔀 Sort and categorise** customer data into intents
3. **💬 Craft example replies** for conversation designers to use as is or as inspiration to brainstorm dialogue
4. **🧙♀️ Transform factual replies** to customer requests into a specific tone of voice
In this article we'll do a deep-dive into two flows that build on top of the Summarise and Sorting powers of ChatGPT and Zendesk.
# 📝 Summarise
The first impact point has been nicely explained by Zendesk itself in a tutorial on the Zendesk Developer website. They published a full how-to on how to create a sidebar app that summarises a conversation each time a ticket is updated.
This makes it easy for anyone looking at a ticket to get the conversation at a glance.
[View Tutorial](https://developer.zendesk.com/documentation/apps/build-an-app/using-ai-to-summarize-conversations-in-a-support-app/?ref=internalnote.com)
# **🔀** Set Ticket Category and Sentiment
The focus of this article lies on the second impact point: Sorting tickets.
The tutorial below will show you how to build a flow that takes any incoming ticket and assigns a category and sentiment to the ticket, which can then be used to set SLA, priority e.a. in your environment.
## Concept
This flow uses a Cloudflare Worker to set the sentiment and category of any created ticket upon ticket creation.
It runs as follows:
1. We have a trigger that notifies a webhook whenever a ticket is created
2. The webhook submits the ID of the created ticket to a Cloudflare Worker
3. The worker uses OpenAI to define a category and sentiment for the ticket and maps it to existing custom field values in your instance.
4. We update the ticket with the sentiment and category value.
Read on for the full tutorial.
# How to build this flow
## Custom Fields
We need two custom fields for this. They can be new Fields, or your existing Category field already in use in your instance.
- **Category**, a dropdown list of categories you want to map your tickets to.
- **Sentiment**, a dropdown with Positive, Neutral and Negative sentiments. Or whichever you prefer for your use case.
You can tweak these drop-downs to your liking or use case, the important thing is that they have a few values.
Note down the Ticket Field ID for both of these Custom Fields
## Worker
The major work of this flow will be done by the Cloudflare Worker. To set it up you can use the repository and copy the `worker.js` code to a new worker via [https://workers.new](https://workers.new/?ref=internalnote.com).
Mine runs at `https://chatgpt-category.internalnote.com`
[GitHub - verschoren/chatgpt-category-worker: A Cloudflare Worker that updates a Zendesk Ticket Category and Sentiment TagA Cloudflare Worker that updates a Zendesk Ticket Category and Sentiment Tag - GitHub - verschoren/chatgpt-category-worker: A Cloudflare Worker that updates a Zendesk Ticket Category and Sentiment TagGitHubverschoren](https://github.com/verschoren/chatgpt-category-worker?ref=internalnote.com)
Once you've initialised your worker, go to the Deployment Settings and add the following five Environment variables:
- `CATEGORY` and `SENTIMENT` are the Custom Fields IDs from the previous step.
- `DOMAIN` is the subdomain of your Zendesk instance
- `OPENAI` is your [OpenAI API Key](https://platform.openai.com/account/api-keys?ref=internalnote.com) (This requires a subscription)
- `TOKEN` is the [Base64 Encoded](https://www.base64encode.org/?ref=internalnote.com) `admin@example.com:token/zendesk_token` of your Zendesk instance with `admin@example.com` any admin in your instance, and `zendesk_token` a [Zendesk API Token](https://support.zendesk.com/hc/en-us/articles/4408889192858-Generating-a-new-API-token?ref=internalnote.com).
### Get Info
The script starts by extracting the `ticket_id` from the URL the worker is triggered from.
We then load the category and sentiment values from your custom fields, as well as the ticket description.
```javascript
export default {
async fetch(request, env) {
const path = new URL(request.url).pathname;
var ticket_id = path.split('/')[1];
var categories = await getCustomField(env,env.CATEGORY);
var sentiments = await getCustomField(env,env.SENTIMENT);
var conversation = await getTicketDescription(env,ticket_id);
// ...
}
}
```
Since the API has a limit of \~4K tokens, we use a bit of code to cleanup the description first:
var description = results.ticket.description;
```javascript
var description = results.ticket.description;
//Remove HTML Code to trim description
const regex = /<[^>]*>/g;
const cleaned = description.replace(regex, " ");
//The limit is 4097 tokens for the API with ~4 characters per token.
//So let's trim the description to 10k characters to be safe.
const trimmed = cleaned.substr(0, 10000);
return trimmed;
```
### Prompt
Once we've got the basic information, we generate two prompts we'll send to OpenAI. These prompts include the ticket field values and ticket description.
> Map the conversation below to this JSON list of categories and return only the value (but not the name or raw\_name) of the best matching category.
You can tweak this prompt to your liking but for me this one worked the most stable.
```javascript
var category_prompt = categoryPrompt(categories,conversation);
var sentiment_prompt = sentimentPrompt(sentiments,conversation);
function categoryPrompt(categories,conversation){
return `
Map the conversation below to this JSON list of categories and return only the value (but not the name or raw_name) of the best matching category.
${conversation}
Categories:
${JSON.stringify(categories)}
`
}
function sentimentPrompt(sentiments,conversation){
return `
Map the conversation below to this JSON list of sentiments and return only the value (but not the name or raw_name) of the best matching sentiment.
${conversation}
Sentiments:
${JSON.stringify(sentiments)}
`
}
```
### Get OpenAI Response
Once we have two good prompts to get the sentiment and category for a ticket, we'll forward those to OpenAI to get the AI to return the best match for the categories and sentiments for the given description.
Note we're using the `text-davinci-003` [model](https://platform.openai.com/docs/models/gpt-3-5?ref=internalnote.com) here since we only need a single text prompt to be returned and don't require a conversation with ChatGPT and require deeper understanding of long text.
```javascript
var category = await openAIRequest(env,category_prompt);
var sentiment = await openAIRequest(env,sentiment_prompt);
async function openAIRequest(env,prompt){
const request = {
model: "text-davinci-003",
prompt: prompt,
temperature: 0.6,
max_tokens: 200,
}
const url = "https://api.openai.com/v1/completions";
const init = {
body: JSON.stringify(request),
method: "POST",
headers: {
"content-type": "application/json;charset=UTF-8",
"authorization": "Bearer " + env.OPENAI
},
};
const response = await fetch(url, init);
const results = await response.json();
return results.choices[0].text.trim();
}
```
### Update Ticket
And finally, once we got a `category` and `sentiment` we can update the Zendesk Ticket with the new values of the Category and Sentiment dropdown fields.
```javascript
if (category && sentiment){
await updateTicket(env,category,sentiment,ticket_id);
return new Response("Ticket updated: " + category+' '+sentiment)
} else {
return new Response('Nothing could be mapped');
}
async function updateTicket(env,category,sentiment,ticket_id){
const url = `https://${env.DOMAIN}.zendesk.com/api/v2/tickets/${ticket_id}.json`;
const ticket = {
"ticket": {
"custom_fields": [
{"id": env.CATEGORY, value: category},
{"id": env.SENTIMENT, value: sentiment},
]
}
}
const init = {
body: JSON.stringify(ticket),
method: "PUT",
headers: {
"content-type": "application/json;charset=UTF-8",
"authorization": "Basic " + env.TOKEN
},
};
const response = await fetch(url,init);
const results = await response.json();
return results;
}
```
🤔
Note, we only check on existance of a value for those two fields, if the value is a non-existing value the API randomly invented the API call will fail.
You can finetune the script to handle those scenario's, or, ignore it like it did. In the case the AI can't figure out a category or sentiment we'll let the agent handle those tickets when they update the ticket.
## Webhook
Create a new Webhook that reacts to a Trigger.
Set the Endpoint URL to the url of your worker, followed by the Ticket ID: `https://chatgpt-category.internalnote.com/{{ticket.id}}`
Leave the other Authentication and Headers options as is. If you've setup the worker correctly you'll get a `200 OK` status and a `Ticket Updated` message after testing the Webhook.
## Trigger
To finalise the setup we need to notify the webhook each time a ticket is created. You can do this by adding a new Trigger that fires when a ticket is created.
Set the Actions to *Notify Active Webhook* and select your Webhook. Remote any URL parameters and leave the rest of the options as is.
## Next Steps
Now that every incoming ticket gets automatically assigned to the right category and sentiment you could use this data to:
- Create triggers that respond to changes in Category and assign the right priority or group to a ticket
- Update ticket priority based on the sentiment
- Have a trigger that calls the webhook each time the customer replies to keep the sentiment updated to the right one
- Create a report that maps sentiment and SLA
# Conclusion
This article showed one (or thanks to Zendesk, two) use cases for ChatGPT to automate and improve your Zendesk Workflows.
This are, on purpose, two flows where we only have an internal impact without any text going to the customer.
When using ChatGPT to generate replies and comments I'm still hesitant. Making sure the facts, tone of voice and quality of the replies aligns with your company values and processes is complex, and I feel it's still early days for fully automated customer interactions.
However, for a future article I'm exploring a flow on summarising Guide Articles in Agent replies.
Also keep an eye on the [YouTube channel](https://www.youtube.com/watch?v=6sSZNSsQQZg&list=PLiM5KikPhEoUiVBLvSER7cf7ToMpn8u7J&ref=internalnote.com) from Premium Plus. My team there has been building some pretty cool apps with ChatGPT too!
### Deep-dive into issues related to Messaging Authentication and JWT User Mapping
URL: https://internalnote.com/deepdive-into-messaging-profiles/
Last updated: 2025-09-08T06:42:58.000Z
⁉️
Apparently old Zendesk accounts can be tagged with an internal flag that prevents External ID matching to work. Zendesk Support removed the flag from my account and External ID tagging now works as expected.
This means Email matching is still an issue, but External IDs do work!
Currently the authentication for Zendesk Messaging as described in my article below has some weird behaviours when it comes to matching to existing Zendesk User Profiles. There's also a rather long thread about these issues on the [Zendesk Community](https://support.zendesk.com/hc/en-us/articles/4411666638746-Authenticating-end-users-in-messaging-for-the-Web-Widget-and-mobile-SDK?page=4&ref=internalnote.com#comments).
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging/)
## What you would expect is:
If a user logs into Messaging and already exists in Zendesk with the same email or external ID, the resulting conversation is mapped against the existing user.
So let's test a few scenario's and log their outcomes:
1. **New user that's never interacted with Zendesk**
A new user uses Messaging for the first time and is logged in via JWT.
Result: --> A new end-user is created without an associated email address in Zendesk.
2. **An authenticated existing user with an email address that exists in Zendesk**
An existing end-user logs into Messaging and starts a conversation. Even though the email address matches an existing user, a new user is created in Zendesk.
3. **An unauthenticated existing user with an email address that exists in Zendesk**
An existing user uses messaging and is **not** logged in. They enter an email address in the *Ask for Details* step. This conversation is added as a ticket linked to the existing user in Zendesk that has a matching email address. (🥳 Finally a good result!)
4. That **same** existing user that has now both messaging and email address linked, now logs into Messaging and starts chatting. This creates a new user without an email, unlinked from the existing user. 🤯
As you can see, none of the scenario's fully match the behaviour one would expect.
Only when you merge a *newly created logged in Messaging user* and the *existing email user profile*, you can achieve a mapping to the same user for any following tickets. But since Zendesk does not show the email address of the logged in Messaging anywhere in the interface, there is no way for an agent to know who to merge too.
## Conversation API
⚠️
****If you have access to Sunshine Conversations,** you can use the Smooch API to do some lookups and retrieve the required values afterwhich you're able to merge the users.
Zendesk also enabled customers to generate a Sunshine Conversation API via de Admin Panel for Suite users.
Below are some example flows that show the actual data and API calls made for these tests.
If you encounter these same issues, please upvote or add a comment to this [Community Post](https://support.zendesk.com/hc/en-us/articles/4411666638746-Authenticating-end-users-in-messaging-for-the-Web-Widget-and-mobile-SDK?page=4&ref=internalnote.com#comments), so we can get this fixed.
# Scenario 1: Test with a new user
## Authenticated Messaging Flow via JWT
This is an example of the JWT token I generate. Note the External ID I use has the email address as part of it to make it show up *somewhere* in the UI for agents. IN real scenarios you'd use an actual ID here from your user database.
```json
{
"alg": "HS256",
"typ": "JWT",
"kid": "app_62965da0ae721700f5743234"
},
{
"scope": "user",
"name": "Peter Parker",
"email": "peterparker@spiderman.example",
"external_id": "user_peterparker@spiderman.example",
"exp": 1682149431
}
```
## Get User Identities in Zendesk
Since the Messaging account info doesn't show up in the interface, I dove into the Identities API to see if something would show up. More info in [this](https://developer.zendesk.com/api-reference/ticketing/users/user%5Fidentities/?ref=internalnote.com#show-identity) Developer article.
```bash
curl --location 'https://d3v-verschoren.zendesk.com/api/v2/users/11109600149778/identities/11109600154130' \
--header 'Accept: application/json' \
--header 'Content-Type: application/json' \
--header 'Authorization: Basic zendesk_token'
```
And yes, there is a `messaging` type in the profiles for these users. No UI in Agent Workspace, but retrievable via API.
```
{
"identity": {
"url": "https://d3v-verschoren.zendesk.com/api/v2/users/11109600149778/identities/11109600154130.json",
"id": 11109600154130,
"user_id": 11109600149778,
"type": "messaging",
"value": "425a9d0b802f5f2ca9cc7463",
"verified": true,
"primary": true,
"created_at": "2023-04-21T07:44:07Z",
"updated_at": "2023-04-21T07:44:07Z"
}
}
```
## Get User in Sunshine Conversations
At first glance this information is not really handy. The `value` on these profiles does not show up on the users, profiles or identities objects in Zendesk. It's only because I have Sunshine Conversations active that checking the `value` against user IDs in SunCo that actual information on the user showed up.
More info on this API can be found in the [Smooch](https://docs.smooch.io/rest/?ref=internalnote.com#operation/getUser) documentation.
🤔
This requires a user to have an active SunCo subscription for its account
```bash
curl --location 'https://api.smooch.io/v2/apps/5f8ecbf276da07000cb2a456/users/425a9d0b802f5f2ca9cc7463' \
--header 'Authorization: Basic sunco_token'
```
```json
{
"user": {
"signedUpAt": "2023-04-21T07:39:09.052Z",
"hasPaymentInfo": false,
"identities": [],
"id": "425a9d0b802f5f2ca9cc7463",
"externalId": "user_peterparker@spiderman.example",
"profile": {
"surname": "Parker",
"givenName": "Peter",
"email": "peterparker@spiderman.example",
"locale": "en-GB"
},
"metadata": {}
}
}
```
# Scenario 2: Start with an existing user that has a external ID
My previous article wrongly noted that creating an Authenticated Messaging user with an External ID that matches an existing Zendesk user would have the system match those users and correctly map the conversation.
Let's test this.
## Create Test User
```bash
curl --location 'https://d3v-verschoren.zendesk.com/api/v2/users.json' \
--header 'Content-Type: application/json' \
--header 'Authorization: Basic zendesk_token' \
--data-raw '{
"user":{
"name":"Green Goblin",
"email": "greengoblin@spiderman.example",
"external_id": "user_greengoblin@spiderman.example"
}
}'
```
```json
{
"user": {
"id": 11109844978962,
"name": "Green Goblin",
"email": "greengoblin@spiderman.example",
"external_id": "user_greengoblin@spiderman.example",
...
}
}
```
## Authenticated Messaging Flow via JWT
I then did a new authenticated API call via https://jwt.internalnote.com with a user that matches the name, email and external ID of this new test user.
This still does not work, I still have two users. One with a Messaging Identity, and a regular Zendesk user.
# A temporary solution
If you have Sunshine Conversations this could be a temporary solution:
- Create a sidebar app that runs next to your tickets.
- Have the app look up the identities of the current requester and get the `type:messaging` identity
- Retrieve the `value` of that identity
- Do a lookup in Sunshine Conversations for that user
- Retrieve the `email` of that user with the information from its `profile:{}` payload
- Search for a user in Zendesk with that same email, and if found merge into that user
- If not found, update the profile of the Zendesk user with the information from its `profile:{}` payload.
You can find a repository that contains a proof of concept of this flow in the following GitHub repository:
[GitHub - verschoren/messaging\_user\_merger: An app that merges Messaging and Email users in ZendeskAn app that merges Messaging and Email users in Zendesk - GitHub - verschoren/messaging\_user\_merger: An app that merges Messaging and Email users in ZendeskGitHubverschoren](https://github.com/verschoren/messaging%5Fuser%5Fmerger?ref=internalnote.com)
# Other remarks
There's also a weird bug: Any regular end-user with an `external_id` set, regardless of them having logged in or not, get a green checkmark with shows "Authenticated" when you hover over it 🤷♂️.
### Learn how to build a full-featured Flow Builder Bot for Zendesk.
URL: https://internalnote.com/flow-builder-dinosaurs/
Last updated: 2025-09-08T06:43:03.000Z
Zendesk Flow Builder has been available for over two years now, and since that time it evolved from a basic flow builder (pun intended) to a complex tool that allows for advanced logic, API calls, authentication and a lot more.
I've [written](https://internalnote.com/tag/flow-builder-and-chatbot/) about how to work with Flow Builder in the past, but since the latest set of releases during the [What's New](https://internalnote.com/whats-new-march-2023/) in March most of the last remaining big ticket items are now available, which means it's time for a deep-dive into Flow Builder.
This article will show you how to build a complete bot and will touch upon every available feature available in Flow Builder. You can find an overview video below, and then you can choose to either watch the full 45min tutorial later in this article, and/or read the article to learn more about the specifics steps used.
# 🦖 Jurassic Park Flow Builder Bot
Tutorials are only as fun as the subject it teaches, so for this how-to will build a bot for Jurassic Park.
We'll start with setting up a few intents and show Help Center articles, before we dive into the really fun stuff with API calls to check on Dinosaur escapes, and use the new authentication metadata to check upon bookings for the park.
Welcome to Jurassic Park. 😎
# How to build a complex Flow Builder Bot
## Customer Intents
As I'll also mention in the tutorial video below, the first step to build a good Chatbot is to define the issues the bot should help the customer with.
In my scenario here, I'm trying to solve four scenarios that will lower agent workload:
1. **How to buy tickets**
Deflect tickets about ticket purchases so that most customers can self-serve and agent only need to handle the complex cases.
2. **Info on Dinosaurs**
Although the Customer Care team should be able to give information about the animals in the park, most of these questions are generic and easily satisfied with a good [article](https://support.internalnote.com/hc/en-us/sections/10793749626514-Jurassic-Park?ref=internalnote.com) or video.
3. **Status of the park**
When things break, you should inform customers in an automated way so that your agents can handle the real emergencies and focus on one to many communication.
4. **Order Information**
Most people that need information on orders either want confirmation or lost their booking emails. Offering that info up front gives agents time to handle real issues with payments or erroneous bookings.
## Mapping it all out
I always start with a flow chart that maps out a rough overview of the flows you expect your customers to follow. This will serve as a guideline for building your Flow Builder bot, and allows for quick iteration before you commit on building the bot.
💡
Since these are four specific flows a customer can take, it's best to build those as four separate intents. This way you can show the intents to the customer upon opening the widget, you can directly launch an intent if a customer searches for it (e.g. I want to buy a ticket), or you can use the intent for [Pro-Active Messaging](https://support.zendesk.com/hc/en-us/articles/5381304334234-About-proactive-messages?ref=internalnote.com).
## Answer Flow Step Types
Flow Builder has around a dozen different Step Types you can use to build your bot. We'll use all of them to build our different flows and intents.
[Understanding answer step typesWhat’s my plan? Tip: Zendesk has renamed our bot capabilities. Answer Bot is now Zendesk bots, Flow Builder is bot builder, and Article Recommendations are autoreplies. Some older features still…Zendesk helpAimee Spanier](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-bot-flow-step-types?ref=internalnote.com#topic%5Fzqr%5Fgwc%5Fk4b)
# Intent #1: Ticket Info
Our first Intent is the *Buy a Ticket* intent. It's a pretty basic flow that offers the customers a set of preselected Guide articles, ask if the issue is resolved and if not, escalates to an Agent.
## Show Help Center Articles
This [step](https://support.zendesk.com/hc/en-us/articles/4408836323738-Understanding-answer-flow-step-types?ref=internalnote.com#topic%5Fgrj%5Fgwc%5Fk4b) allows you to show up to six public articles to your customers. Articles will be shown in the locale the Widget is loaded in and articles are always shown in a carousel in the same order.
Once a customer clicks an article, they'll be redirected to your Help Center. But no worry, if they open the Widget again it'll remember their actions. Any read article will show up in the Context Panel in Agent Workspace [if you've enabled](https://support.zendesk.com/hc/en-us/articles/4408829170458?ref=internalnote.com#topic%5Fnq1%5Fqnm%5Fxsb) those events.
## Ask if question is resolved
Once we've shown the articles we'll ask the customer if the question is now resolved. The answers to this question will be stored in [Explore](https://support.zendesk.com/hc/en-us/articles/4408829761690-Analyzing-your-Flow-Builder-activity?ref=internalnote.com) so you can measure your success-rate. If the customers answers negatively, we'll escalate to an Agent.
## Business Hours
Zendesk Messaging can take your Support schedule into account and can show a different reply inside or outside of Business Hours.
## Authentication Info 🆕
If you allow your users to [login](https://internalnote.com/jwt-messaging/) to your Messaging Widget you can use that information to pre-fill their name and email and make escalation to an Agent easier. You could even have different experiences for VIP users vs regular users, or assign logged in users to a separate support tier.
[Announcing Flow Builder - Messaging Authentication MetadataRollout start Rollout end March 17, 2023 March 20, 2023 We’re excited to announce that you can now build more personalized experiences for authenticated users by using Messaging Authentication…Zendesk helpLisa Tam](https://support.zendesk.com/hc/en-us/articles/5492040520474-Announcing-Flow-Builder-Authentication-Metadata?ref=internalnote.com)
## Transfer to Agent
Once we've offered a few Self Service solutions in attempt to deflect the question, and got confirmation that, yes, the customer does need help, we can transfer the conversation to an Agent in Support.
For unauthenticated users we can use the Ask for Details step (see below) to ask for their name and email, before we attempt to find an available agent.
[Zendesk helpLogo](https://support.zendesk.com/hc/en-us/articles/5378287266586-Announcing-updates-to-Flow-Builder-Transfer-to-Agent-step-?ref=internalnote.com)
# Intent #2: Dinosaur Info
Similar to the basic deflect or escalate flow in our first Intent, this second Intent will follow a similar path.
## Carousel
When the customer chooses the *Dinosaur Info* option we'll use the Carousel step to show them information on three Dinosaurs. Each carousel element included an image (loads an externally hosted photo), a short description, and a button that links to a Help Center (or external) article with information about the Dinosaur.
## Escalate to Agent
Here also, when the customer lets the bot know that they need more information, we escalate to an Agent.
💡
Did you know a right click on any step allows you to copy the step and all steps below? You can then paste that logic anywhere else in your bot and reuse it, instead of recreating a flow from scratch.
# Intent #3: Fence Status
Our third intent introduces some more logic based on API calls and handling the returned data.
## API Call
When a customer chooses the *Status of the Fences* intent we immediately fire an API call. This API call checks an external API resource `https://jurassic-park-bot.internalnote.com/status` and gets a payload back.
```json
{
"status": ["all_good","unlocked_raptor","unlocked_trex","evacuate"]
}
```
We make use of the recently announced [Custom Webhook Headers](https://internalnote.com/custom-authentication-for-webhooks-update/) to reach a protected API resource, and store the returned `payload.status` in to a variable `status`
If the API call fails we'll try to reach an agent for the user by once again copying those steps from the previous Intent.
[GitHub - verschoren/jurassic-park-bot: Cloudflare Worker with API Endpoints for the Jurassic Park Widget IntentCloudflare Worker with API Endpoints for the Jurassic Park Widget Intent - GitHub - verschoren/jurassic-park-bot: Cloudflare Worker with API Endpoints for the Jurassic Park Widget IntentGitHubverschoren](https://github.com/verschoren/jurassic-park-bot?ref=internalnote.com)
## Branch By Condition 🆕
If we do get a valid payload back we use the new Conditional Flows step to show different results based on the `status` value. For each status we show a Bot Message step that includes an externally linked image and a button that links to a Help Center article. And since this Intent often ends with disasters and escaped Dinosaurs, we do not escalate to an Agent for this flow.
[Zendesk helpLogo](https://support.zendesk.com/hc/en-us/articles/5378419926426-Announcing-Flow-Builder-Conditional-Flows?ref=internalnote.com)
💡
You'll notice we use a nested branch condition for this step. We could have used only one, but since this step only supports up to 5 conditions, it's useful to showcase this nested scenario. It's a simple hack that allows you to circumvent the five conditions limitation.
# Intent #4: Order Info
Our final intent is a variation on the previous one. We once again make an API call to an external service, but instead of a hardcoded URL with results that are the same for all visitors, we'll make an API call that contains unique information provided by the customer and returns personalised responses.
## Ask For Details
We use the *Ask For Details* step to ask for an Order Number. This value is stored in a custom field for easier retrieval by an Agent if escalation is needed, and is also automatically stored in a variable `Order Number` we can use anywhere in our Bot Logic.
## API Call with POST/Details
Where the previous Intent did a GET request, we will now make a POST request to `https://jurassic-park-bot.internalnote.com/order` with a JSON Payload:
```json
{
"order_id":"{{Order Number}}"
}
```
The API call either fails, or successfully returns our order data. We store all returned data in variables, and use the *Send a Message* step to return the data in a nicely formatted way.
```json
{
"items": [
{
"name": "Jurassic Park Adventure Park",
"quantity": 2,
"ticket_price": 150,
"ticket_type": "1-Day Park Hopper",
"type": "attraction"
},
{
"name": "Jurassic Park Hotel",
"nights": 3,
"room_price": 350,
"room_type": "Deluxe Room",
"type": "hotel"
}
],
"order_date": "June 2nd",
"order_id": "ORD1002",
"order_status": "confirmed"
}
```
💡
Sadly, Flow Builder does not support looping through arrays to store API call responses into variables. So in a scenario like this one where we get an array of items we need to manually store each `items[i]` as a unique array and the API call will fail if we store e.g. a variable `items[3]` and the next API call only responds with 1 element.
# Conclusion
What started as a fairly simple Flow Chart to handle four clearly defined intents, ended up in a complex tree of over a hundred steps, including links to Help Center articles, external APIs and hosted images.
I really hope the video above, and the documentation in this article helps you to build your own Flow Builder Bot.
## A small ask
Building this bot and tutorial took a lot of preparation. If this **free** **article** was useful to you, share it with a friend or colleague and ask them to subscribe.
If this really offered value, please consider subscribing to a paid tier to make this blog and project viable and allow me to write more of this content in the future.
#### ☕️ Supporter
Most popular way to support this website
[Subscribe Now](https://internalnote.com/#/portal/signup/63ff05d99deb9e003d619e70/yearly)
### Creating a Pokédex with Zendesk Custom Objects 2.0
URL: https://internalnote.com/creating-a-pokedex-with-zendesk-custom-objects/
Last updated: 2025-07-31T08:33:59.000Z
The awesome people at Zendesk just made the [Custom Objects v2 EAP](https://support.zendesk.com/hc/en-us/community/posts/5359269969178-What-is-the-new-Custom-Objects-EAP-?ref=internalnote.com) available.
So why not build a Pokédex inside Zendesk to get to know the APIs 😉
I'll publish a full tutorial soon on [https://internalnote.com](https://internalnote.com/), so subscribe if you want to know more!
I've also started a Wishlist of missing features and improvements:
[Custom Objects EAP FeedbackBelow is a list of feature request and remarks for the Custom Objects EAP.Internal NoteThomas Verschoren](https://internalnote.com/custom-objects-eap/)
## More info:
- [What is the new Custom Objects EAP?](https://support.zendesk.com/hc/en-us/community/posts/5359269969178-What-is-the-new-Custom-Objects-EAP-?ref=internalnote.com)
- [Custom Objects Set-up Guide for Admins](https://support.zendesk.com/hc/en-us/articles/5392409465370?ref=internalnote.com)
- [Developer documentation](https://developer.zendesk.com/api-reference/custom-objects/introduction/?ref=internalnote.com)
### Improve Your Content Discoverability with a Linked List in Zendesk Guide
URL: https://internalnote.com/linked-list-for-guide/
Last updated: 2025-09-08T06:44:41.000Z
As mentioned in last weeks' article on [Federated Search](https://internalnote.com/federated-search/), as businesses grow, their knowledge base expands too. From FAQs to product or marketing pages, there's an abundance of information available to help customers find what they're looking for. By default, the content indexed via Federated Search is only searchable, and not easily discoverable for your users.
In this article, we'll explore how to use Zendesk Federated Search and [Custom Guide Pages](https://support.zendesk.com/hc/en-us/articles/4409012911770-Creating-custom-pages-in-the-help-center?ref=internalnote.com) to create a custom page that provides a curated list of links for end-users. What sets this solution apart is that the added links are browsable just like regular Guide articles in the same UI as Article sections. This approach allows businesses to make even more content available to their customers, not just traditional FAQ articles.
With our custom page, end-users will, for example, have access to App Store install pages, marketing pages, and more. Moreover, we'll also build an admin interface that allows team leads to manage the links indexed on the Help Center and giving them the ability to add or delete links as needed. This new feature provides a more personalised experience for customers, making it easier for them to find the information they need.
In summary, by using Zendesk Federated Search to create a custom landing page, businesses can make more content available to their customers in the same UI as Article sections. This personalised experience enables customers to browse and search through a curated list of links, not limited to traditional FAQ articles. So, let's dive in and explore how to build this custom page and admin interface.
# Tutorial
As always you can find the full code on GitHub:
[GitHub - verschoren/linked-list: Linked List for Zendesk GuideLinked List for Zendesk Guide. Contribute to verschoren/linked-list development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/linked-list/?ref=internalnote.com)
# Prerequisites
Use the Federated Search API or the Search Settings in Guide to:
1. Add a source: "Linked List" and note down its ID
2. Add a few types, eg: "Interesting Articles", "Product Pages".
3. Add one record for each type via the Federated Search API. This makes sure once we retrieve records to show we can at least show one or two articles.
You can refer to my [earlier article](https://internalnote.com/federated-search/), or use the [Zendesk Documentation](https://developer.zendesk.com/api-reference/help%5Fcenter/federated-search/introduction/?ref=internalnote.com)
# Worker
As usual on this website, we'll first build a Cloudflare Worker that handles the API site of our solution. This way we can remove any API keys and complexity from the client side code rendered on your Help Center, and make use of Cloudflare's build-in caching to speed up the page rendering.
Our worker has three end-points:
- `/get` responsible for returned all Federated Search records
- `/add` responsible for adding a record
- `/delete` responsible for deleting a given record
You can copy the example in `worker.js` just make sure to change the top three variables:
```javascript
//your Zendesk subdomain
const subdomain = 'yourdomain'
//Base 64 encoded admin@domain.com/token:{Zendesk API token}
const auth = '123457890qwerty='
//The Source you want to display, "Linked List" in my example
const source_id = '01GT1VEVYWJ10HMZKMR5YRXQHB';
```
You can find the Source ID via [this API call](https://developer.zendesk.com/api-reference/help%5Fcenter/federated-search/external%5Fcontent%5Fsources/?ref=internalnote.com).
The script is pretty straight forward. The only tricky part is filtering the returned records on the `get` call to only show those of the chosen Source and handling pagination correctly:
```javascript
async function getRecords(url,federated_items){
url = url != ''? url : 'https://'+subdomain+'.zendesk.com/api/v2/guide/external_content/records?page[size]=10';
var init = {
method: 'GET',
headers: {
'content-type': 'application/json;charset=UTF-8',
'Authorization': 'Basic ' + auth,
},
};
const response = await fetch(url, init);
const results = await response.json();
merge(federated_items,results.records);
//Check if the results are paginated and if so, call the function again
if (results.meta.has_more == true){
return await getRecords(url + '&page[after]='+results.meta.after_cursor,federated_items);
} else {
//filter federated_items to remove all records with a different source.id
federated_items = federated_items.filter(function(item){
return item.source.id == source_id;
});
return federated_items;
}
}
```
💡
Since these are example pages my scripts do not have any API limiter or security headers. You should make use of e.g. Cloudflare Access or a custom header to protect your scripts!
## Custom Page
Once we have a script that allows us to get, add or delete records, we can start working on the Help Center Page.
To start, create a new Custom Page on your Help Center and copy the code from the GitHub repository in [linked-list.html](https://github.com/verschoren/linked-list/blob/main/linked-list.html?ref=internalnote.com). If you use the default Copenhagen theme, you'll end up with a page that looks exactly like your section pages.
For the page to work you'll need to replace `worker_url` with the URL of your worker.
## Permissions
The script starts with a small section that checks what role current user has and hides the admin interface if they aren't an agent or admin.
```javascript
var admin = false;
if (HelpCenter.user.role != "agent" && HelpCenter.user.role != "manager"){
$('#addlink').remove();
$('.linkform').remove();
} else {
admin = true;
}
```
## Rendering the layout
The page runs on a `getLinks()` function that
- gets all records from the Worker
- creates a section for each `type` it finds
- adds `articles` to that section and adds a delete button if the user is an admin/agent for that article
- adds each unique `type` to the Types dropdown in the form to create new links.
```javascript
function getLinks(){
var settings = {
"url": worker_url + "/get",
"method": "GET",
};
$.ajax(settings).done(function (links) {
$('#main-content').html('');
var sections = [];
links.forEach(function(link) {
if(sections.indexOf(link.type.name) === -1) {
sections.push(link.type.name);
$('#type').append(`${dropdown_option})
$('#main-content').append(`${section}`)
}
$('.article-list[data-articles="'+link.type.name+'"]').append(`${linkeditem});
});
});
};
```
## Adding Records
The HTML page has a form that asks an agent/admin for a title, url, description and type. Based on that info we can ask the worker to create a new record.
Note that records require a unique `external_id`. Since manually added records don't really have a good use for these, we just generate one based on the current timestamp.
We then call the `/add` endpoint of our worker via a POST.
```javascript
$(document).on('click', '#submit', function(event){
var json = {
"record": {
"body": $('#body').val(),
"external_id": 'abc' + Date.now(),
"locale": "en-us",
"title": $('#title').val(),
"type_id": $('#type').val(),
"url": $('#url').val()
}
}
var settings = {...};
$.ajax(settings).done(function (response) {
getLinks();
/*empty form*/
});
});
```
## Deleting Records
To remove a record we have a `data-target` element on each link we display with the record id. Clicking the delete button for a respective element asks the worker to delete that element via `/delete/id`
```javascript
$(document).on('click', '.delete', function(event){
var id = $(this).attr('data-target');
var settings = {
"url": worker_url + "/delete/"+id,
"method": "POST",
"headers": {
"Content-Type": "application/json",
},
};
$.ajax(settings);
getLinks();
});
```
# Home Page
Since Custom Pages don't show up on Guide by default we can update the `home_page.hbs` on your Guide Theme and append an extra category button like so:
Just make sure to replace `/hc/p/linked_list` with the actual link to your Custom Page.
```html
```
💡
Power Tip:
Create a Federated Search Record that indexes your Custom Page. That way your users can find it easily via search. It's a bit like Inception, but does the trick.
# Conclusion
There you have it. We created a custom page that shows a bunch of links for your End-Users, and makes your Help Center even more powerful.
### Expanded API Support via Custom Authentication for Zendesk Webhooks
URL: https://internalnote.com/custom-authentication-for-webhooks-update/
Last updated: 2025-09-08T06:43:08.000Z
Zendesk announced Custom Headers and API Key support for its Webhooks. Where they already supported `Username/Password`, you can now add up to 5 custom headers to your webhook configuration, expanding its support to way more platforms.
[Announcing API key authentication and custom headers for webhooksAnnounced on Rollout starts March 30, 2023 March 30, 2023 Zendesk is excited to announce that you can now use API key authentication and define additional headers when configuring webhooks.…Zendesk helpZach Anthony](https://support.zendesk.com/hc/en-us/articles/5532092885658-Announcing-API-key-authentication-and-custom-headers-for-webhooks?ref=internalnote.com)
I wrote about [Shortcomings of Webhooks](https://internalnote.com/shortcomings-of-zendesk-webhooks/) last fall, and mentioned three scenarios that weren't possible with webhooks at that time:
1. Use webhooks to send data to Zendesk Sell, due to lack of `Accept: application/json` header
2. Access resources behind Cloudflare Zero Trust, due to lack of custom authentication headers
3. Accessing the Asana API, due to lack of `bearer: token` support.
Let's see how this new release changes things.
# Add Sell Lead from Zendesk Support
My first example was a scenario where I wanted to create a lead in Sell each time a ticket was assigned to Sales. You can use it to add leads for tickets you assign to the Sales group, or create tasks in Sell for your Sales team,..
There obviously is a very nice integration for Sell and Support already, but sometimes doing things automatically via triggers is just a bit easier.
## Step 1: Creating a Webhook
Step one in our flow is creating a new webhook.
Use `https://api.getbase.com/v2/leads` as the Endpoint URL, `POST` as the method, and `JSON` as the format.
Next choose `Bearer token` as the authorisation method and enter your [Access Token](https://developer.zendesk.com/documentation/sales-crm/first-call/?ref=internalnote.com#1-generate-an-access-token).
Finally, add a **new** Header with `Accept` as the name, and `application/json` as the value.
## Step 2: Adding a Trigger
Next up, calling our Webhook. For this we'll use a trigger which, each time a ticket is assigned to Sales, will create a new Lead in Sell with the relevant information, and solve the ticket.
The ticket will dissappear from the Inbox of your Support Team, and a new Lead will appear in Sales' Pipeline.
Create a new trigger that looks for:
- group is changed to Sales
And add the following conditions:
- Status is solved
- Notify Webhook (the one you just created) and add the following JSON payload.
```json
{
"data": {
"first_name": "{{ticket.requester.first_name}}",
"last_name": "{{ticket.requester.last_name}}",
"organization_name": "{{ticket.organization.name}}",
"description": "{{ticket.description}}",
"email": "{{ticket.requester.email}}"
}
}
```
## Result
And, if all goes well, each time we now assign a ticket to sales, the ticket will be resolved, and a new lead will show up in Zendesk Sell.
# Protect Cloudflare Worker
As you've probably already noticed on this blog, I'm a big fan of using Cloudflare Workers to handle complex API actions outside of Zendesk.
For example, in the flow below, I use a Worker combined with Zendesk Webhooks to update a newly user with a proper signature, alias and profile picture.
[Webhooks for User and Organisation eventsZendesk recently launched an expansion on their webhooks functionality that allows you to subscribe to changes in Users and Organizations and act upon those actions. In this article we’ll show how you can auto-complete agent profiles with a signature, alias and profile image upon creation.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-user-events/)
Up till now, these workers had to run "unprotected" since Zendesk was not compatible with the way Cloudflare Access works. Now, with this release, I can finally protect these workers by adding these two custom headers to my webhooks.
```html
CF-Access-Client-Id:
CF-Access-Client-Secret:
```
For more information on creating Access Tokens for Workers, take a look at the article below.
[Service tokens · Cloudflare Zero Trust docsYou can provide automated systems with service tokens to authenticate against your Zero Trust policies. Cloudflare Access will generate service tokens …Service tokens · Cloudflare Zero Trust docs](https://developers.cloudflare.com/cloudflare-one/identity/service-tokens/?ref=internalnote.com)
# Adding a task to Asana
So far, that's two out of three issues resolved. Let's check the final one, Asana. In this example we're going to create a new task in Asana whenever a ticket is updated and the ticket type is changed to, you guessed it, Task.
## Step 1: Create a Webhook
First off we create a new webhook that calls the `ttps://app.asana.com/api/1.0/tasks` endpoint.
We use type `JSON` and action `POST` as the option and select `Bearer Token` as our authentication type. See [this article](https://developers.asana.com/docs/personal-access-token?ref=internalnote.com) on how to get a token.
To complete the setup, add a custom header with name: `accept` and value: `application/json` and save your webhook.
## Step 2: Create a trigger Trigger
Create a new trigger that runs when a :
- Ticket is updated
- Type is changed to Task
- Tags does not contain `asana` (this prevents multiple tickets being created)
Set the actions to:
- Alert Webhook (the one you just created)
- Add tag `asana`
In the Alert Webhook action, add the following JSON payload. Don't forget to replace the [project](https://developers.asana.com/reference/goal-relationships?ref=internalnote.com) and [workspace](https://developers.asana.com/reference/workspaces?ref=internalnote.com#workspace) IDs with your own. The result will be an Asana task with the latest comment as the description and the ticket subject as the title.
```json
{
"data": {
"resource_type": "task",
"name": "{{ticket.title}}",
"html_notes": "Zendesk Ticket #{{ticket.id}}: {{ticket.latest_comment}}",
"projects": [
"1204303360323625"
],
"workspace": "1201829988571356"
}
}
```
## Result
And if all goes well, each time we now change a ticket to type 'Task" a new task in Asana is created!
Naturally the native [Asana for Zendesk](https://asana.com/apps/zendesk?ref=internalnote.com) integration is way more useful, but this example shows that the Webhook authentication updates allow for a lot more API integrations now!
# Conclusion
3 out of 3\. I think one of my [wishes](https://internalnote.com/zendesk-wishlist/) came true.
A seemingly small update to the Webhooks feature, makes three otherwise impossible scenarios now possible.
What kind of webhooks will you build next?
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** via email or ****share** the article to your colleagues.
### What's New for Zendesk in March 2023
URL: https://internalnote.com/whats-new-march-2023/
Last updated: 2023-11-28T10:55:50.000Z
Every quarter Zendesk releases their big What's New online Keynote presenting the newest features for customers, support teams, admins and your business overall.
This time the event focusses on four big themes, which closely align with a few trends from their Trends Report for 2023: namely *Keeping up with advances in AI*, *Personalisation* and *Breaking down Silos.*
The four key elements this time were:
1. AI-powered conversational experiences
2. Intelligent operations at scale
3. Open and flexible platform
4. Trust and security
Notably, there's not a lot about Sentiment Analysis in this release. Fingers crossed for some more ML releases for Q2 2023 that might tackle this aspect of CX.
You can get a full recap of the event [here](https://www.zendesk.com/whats-new/?ref=internalnote.com), or read on for my analysis.
As a side note, I have a [running list](https://internalnote.com/zendesk-wishlist/) of wishes and complains, and I managed to remove a whopping 9 items from the list this time 🥳
#### Zendesk releases I could scratch from my Wishlist
1. Security - Groups and organization bug — [documentation](https://docs.google.com/document/d/1NKvTOtQQP5RZUbIJUQVlP2d8mkwDS2S4vdW8OiK9WM4/edit?ref=internalnote.com)
2. Security - Better search in Audit Log
3. Agent Workspace - Enable Followers by default
4. Messaging - Allow settings tags and other metadata.
5. Messaging - Allow access to Authenticated User info for API Call step
6. Guide - Allow for both SSO and Social logins side by side (e.g integrate product accounts and social logins)
7. Allow for both SSO and Social logins side by side (e.g integrate product accounts and social logins)
8. API Support - Deleted Users Add `deleted_users/delete_many?ids=`
9. Messaging Widget - Pro-Active Popup managed serverside
# 🤖 AI Powered Conversation Experiences
Apparently, AI is the new OmniChannel, or the new Self Service if the Marketing of most modern CX tools are to be believed. Leaving the discussion of fad vs feature for another time, one has to admit that since launching Messaging two years ago, the advances in Flow Builder and Conversational Experience across web and social has improved drastically since the old days of Chat and Social DMs as tickets. And the releases of this quarter keep that momentum going.
## Proactive Messages
The biggest release of this What's New was already teased a while back in a [press release](https://www.zendesk.com/blog/proactive-messages/?ref=internalnote.com). Zendesk is adding a way to send messages via the Messaging Web Widget to end-users based on specific actions they take while browsing a website.
#### 📝 Feature Overview
- New Admin section that allows Admins to create triggers for Web Messaging
- Triggers can run based on page title, visitor actions, duration of visit, tags,..
- Target specific customers who've authenticated
- You can choose when it runs based on schedules or agent availability
- Dashboard that shows your running campaigs and where and when a trigger has fired, and what the next actions of the end-user was
- You can choose between an agent or a bot as the first responder
- Available by end of March on all Suite plans
> By adopting a solution of self-service, automation and agent assistance, customers get immediate answers while agents can focus on more complex tasks. Zendesk comes pre-configured so it is easy for admins to schedule when and how often messages will be sent from an agent or bot.
This new feature combines a lot of legacy Chat Triggers but wraps them in the modern and clean UI of Messaging. Sadly it's not live (yet) with no EAP for Partners or customers, but from the short demo we've seen and the screenshots floating around it seems like a powerful tool tool both from a Marketing standpoint (product awareness), Sales (you can interrupt or assist a potential customer in their journey) or Support (pro-active alerts for downtime or issues).
There's a few caveats though. It currently only works for the Web Widget, and it only can only be configured via the Admin Panel. If you want to create alerts via API, or do complex targeting based on data that lives in your CRM, or if you want to target WhatsApp or other social Channels, you'll still need a license for Sunshine Conversations.
When the tool comes available we'll publish a deep-dive on this blog, and 🤞 find some way to interact with this via API. To conclude, CX Today wrote a nice overview from a Business standpoint.
[Zendesk Debuts Proactive Messages, a Conversational Experience ToolCX Today reports on the latest customer experience technology news from around the globe. Read similar CRM news hereCX TodayRory Greener](https://www.cxtoday.com/crm/zendesk-debuts-proactive-messages-a-conversational-experience-tool?ref=internalnote.com)
## Authenticated Metadata in Flow Builder
#### 📝 Feature Overview
- Use metadata of logged in users in Flow Builder
- Build custom paths via the Conditional Flow step for logged in users
- Personalise responses by prefilling the users' name
- Assign logged in users, or VIP users to separate Agent (groups)
This release deserves a big *finally*.** Having the information of the logged in user programatically available makes it easier to give a more personal experience (Hey John!), or preload relevant information for your end-user upon launch of the widget.
I can already imagine flows where a logged in user conditionally triggers an API call, and we serve them with the status of their latest order or their repair as part of the inquiry steps when a bot launches.
This release should also fix the issue where an authenticated Messaging User doesn't always map to an existing Email User with the same email in Zendesk, so overall, a very nice feature addition.
For more information on authenticating Messaging, take a look at this article.
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging/)
## Web Widget Customisation
#### 📝 Feature Overview
- Position anywhere on page
- Accessibility prompts
- Shape Modifier: square, circle, or rounded corners with custom radius.
- Sound on/off
- Disable attachments (security)
A few years ago I wrote the original [Zendesk Widget Configurator](https://widget.guide/?ref=internalnote.com). It turned the extensive Classic Web Widget API into a native UI and made customising that widget (slightly biased here) way easier.
When Messaging arrived I was glad to see that it came with a native interface very similar to my tool, and sad that the new widget seemingly did not have a lot to change or tweak. Over the last two years this has gradually changed, and a lot more features that used to be available to the Classic Widget are now available to the Messaging Widget.
The major items [missing](https://internalnote.com/zendesk-wishlist/) for me now are a Javascript API to tweak behaviour on a per website basis, and a way to pre-fill or skip the *Ask for Details step* in a Flow Builder flow.
If you're interesting in customising the widget even more, take a look at this article:
[Custom Launcher for Zendesk Messaging and Unread CountsZendesk recently added the ability to their Messaging Widget to use Custom Launchers. This article shows you how to easily implement it.Internal NoteThomas Verschoren](https://internalnote.com/custom-launcher-for-zendesk-messaging-and-unread-counts/)
# ⚙️ Intelligent operations at Scale
## Zendesk Explore Improvements
#### 📝 Feature Overview
- Dashboard restrictions: make only a subset of data available to an agent or group.
- Create variants of the same dashboard to dynamically change the data shown based on the selected group.
- Better agent productivity reporting
I'll be the first to admit: Zendesk Explore is not my forte. But these improvements are items I've seen lots of customers request. I've heard from customers who, thanks to the dynamic restrictions, can now turn 25 copies of the same dashboard (one for each of their groups) into 1 dashboard with just a few restrictions configured. Which also means: instead of changing 25 dashboards when a new metric is required, is not turned into 1 dashboard. Sweet.
## Custom Ticket Status
#### 📝 Feature Overview
- Turns the classic Zendesk statusses into Status Categories
- Each category can have multiple custom statusses assigned
- You can have an On Hold - Internal, On Hold - Supplier, On Hold - Development, or Pending - Waiting for Confirmation vs Pending - Waiting for Information
- Comes with full API parity
This is one of those big ticket items that removes a lot of frustration from Zendesk Admins. We're we used to use combinations of Custom Fields and Status to get insights on why a ticket was On Hold, this "new" release turns a few fixed statuses in a flexible and extensible feature in Zendesk. I say "new" since this has been available as a public beta for a few months now. If you haven't tested it yet, here's a few articles to get you started!
[Custom Status API exploringExploring the technical side of Custom StatussesInternal NoteThomas Verschoren](https://internalnote.com/custom-status-api-exploring/)
[Better Pending Ticket flow with Custom StatusesNot every ticket can be solved with a single reply. Agents often need more information from customers in order to get the full context and reply with an answer. Or, even if all info is there, they need confirmation that their suggestions worked before the ticket can be solved.Internal NoteThomas Verschoren](https://internalnote.com/better-pending-ticket-flow-with-custom-statuses/)
[Build a Task Manager in Zendesk Support via Custom StatusesTask Manager to snooze tickets in Zendesk via the new Custom Statuses.Internal NoteThomas Verschoren](https://internalnote.com/custom-status-triggers/)
# 🗺️ Open and Flexible Platform
#### 📝 Feature Overview
- ****Side conversations for Microsoft Teams**
Similar to the existing Slack integration, you can now directly escalate tickets to your colleagues in Microsoft Teams. Useful for escalations to e.g. DevOps, IT or Finance. Still in Early Access
[More Info](https://support.zendesk.com/hc/en-us/articles/5191537451290?ref=internalnote.com)
- ****Zendesk for Google Chat**
Get notified and take quick actions on Zendesk tickets right from Google Chat. Notifications can happen in batch.
[Install the App](https://www.zendesk.com/marketplace/apps/support/913140/zendesk-for-google-chat/?ref=internalnote.com)
- ****Shopify Premium app**
More complex and extensive integration to Shopify then the free default integration.
[Install the App](https://www.zendesk.com/marketplace/apps/support/867416/shopify-premium-for-zendesk/?ref=internalnote.com)
- ****Quick Setup for Sandboxes**
Allows for sandbox creating in minutes instead of hours. You can start working once the setup has been copied, no reason to wait for all (sample) data to be cloned.
- ****New Unity SDK**
Same experience as the Web Widget and Zendesk SDK, but now written for Unity and easier to implement into Games.
- ****Flexible Webhooks**
Use Zendesk Events like creation or modification of users, organizations, groups as a trigger for external API actions.
The Open platform section of the event started with a nice overview of how you can integrate with Zendesk and built on top of it. Most customers start with the Marketplace or play around with Zapier or Workato to automate some processes.
Readers of this newsletter and website know I love to focus on the other end of the platform, by leveraging the what Zendesk calls "Pro-Code" side of integrations. I personally prefer Custom Development as a term but yeah..
This being a customer facing event, most of the releases are focussed on On The Shelf and Low-Code with some a few new Marketplace Apps and a few prebuilt integrations.
Teams for Side Conversations (not to be confused with the [Teams for Support](https://www.zendesk.com/marketplace/apps/support/767198/microsoft-teams-for-support/?ref=internalnote.com) app) complements the existing Slack integration nicely and helps integrating Zendesk into both Google/Slack, and Office365/Teams environments.
This combined with the upcoming UI change where Side Conversations get a nice new location in the Context Panel on the right side of Agent Workspace will makes this an often used solution that "breaks silos".
The only real low-level code feature we got this time was Flexible Webhooks. It's a bit weird this feature got stage time, since its' been around since late last year already. Nevertheless it's a good feature. Instead of needlessly parsing `created_at` dates or listening to an Incremental Export stream, you can now subscribe to a specific Zendesk event (like User Creation) and have your CRM react to it to e.g. augment data and fill in some custom fields. See an exam article below:
[Webhooks for User and Organisation eventsZendesk recently launched an expansion on their webhooks functionality that allows you to subscribe to changes in Users and Organizations and act upon those actions. In this article we’ll show how you can auto-complete agent profiles with a signature, alias and profile image upon creation.Internal NoteThomas Verschoren](https://internalnote.com/zendesk-user-events/)
Would I have loved to see more custom development tools? Sure. Sunshine Events could use some love, Custom Objects desperately need some UI integrations and bulk APIs and ways to programmatically interact with Explore Reports or the Web Widget are sorely missed.
# 🔐 Trust and Security
#### 📝 Feature Overview
- Separate Group and Organization permissions for a more granular permissions and roles.
- Audit Log allows for filtering by type: specific settings, users, and business rules. You can also filter by the name of an object.
- Deleted users are automatically permanently deleted after 30 days
- Multiple SSO sign-in options for end-users and agents - [More Info](https://support.zendesk.com/hc/en-us/articles/5380943678106?ref=internalnote.com)
- Advanced Encryption (BYOK)
This section often feels like the "Let legal also have something to say", but it's impressive to see how Zendesk keeps shoring up their walls while keeping the interface and admin experience easy to use and understandable.
Where other tools have layers of IAM management and permission models, setting up Zendesk roles and permissions still feels fairly simple, even though it's compatible with more and more complexer models over time.
The complex and direct language used in the Advanced Encryption announcement makes me think this feature will differ from what we're used to, but hopefully it comes with that same "beautiful and simple" experience.
As far as User Deletion goes, the new Auto-Delete feature feels more like a way for Zendesk to improve their storage usage than a way to make GDPR easier, and even though the new automation is easy, I still need to resort to the `/search` and `/delete_many` API endpoints or Partner apps like [GDPR Search-And-Destroy](https://www.zendesk.com/marketplace/apps/support/206749/gdpr-search--destroy/?ref=internalnote.com) for easy deletion (not sponsored) for any bulk deletion or compliance automations.
And finally, the multiple SSO feature is one of those, you'll like it when you need it releases. I've seen a few customers with multibrand environments, or with logins on both legacy and migrated platforms who used hacks like [multiple](https://support.zendesk.com/hc/en-us/articles/4408820843802-How-can-I-set-up-multiple-Zendesk-SSO-integrations-in-separate-help-centers-?ref=internalnote.com) Help Centers or multiple [routes](https://support.zendesk.com/hc/en-us/articles/4408886711066-Multibrand-Using-multiple-JWT-single-sign-on-URLs?ref=internalnote.com) to fix this, who can now use this new native release.
# Conclusion
Looking back at Zendesk's Trends Report and this first big What's New of the year it's clear that Zendesk keeps focussing on Conversations and Breaking Silo's.
Messaging, Ticketing and Self Service are still core parts of the Zendesk experience. Although we keep seeing the trench lines moving from Webforms and Help Center articles to Chatbots and Conversations, the first line of defence is still good ticket deflection and self service and Zendesk seems to keep investing in it.
Where Customer experience is one side of the coin, Agent productivity is the other. And by investing in the Agent Workspace, escalation tools like Teams, insights via Explore or the Shopify app, Zendesk keeps giving Agents more context and more tools to give personal support to their customers.
I do feel this is not the entire story yet. The Zendesk Trend Report also focussed on AI Automations and Sentiment, and we haven't seen any major advances of that yet in Zendesk this year.
The [Zendesk-Cleverly](https://www.zendesk.com/newsroom/articles/zendesk-welcomes-cleverly/?ref=internalnote.com) purchase promised a lot with regards to automation, and concepts build on top of ChatGPT promise a lot, but I really hope we see some build-in solutions for content discovery, sentiment analysis, answer assist etc this year.
### Zendesk Roundup for March 2023
URL: https://internalnote.com/link-list-for-march-2023/
Last updated: 2023-05-01T17:45:39.000Z
March started slow with only a few updates, but luckily we had a major What's New in the middle of the month with a LOT of new releases focussing on Conversational ineractions, cross-departemental collaboration, security and.. pro-active messaging for the web widget!
This last one will only go live at the end of the month, but from what Zendesk has shown, it'll be an awesome feature addition.
📥
I'll be publishing my in depth overview of these new releases later this week, so if you don't want to miss it, don't forget to [subscribe](https://internalnote.com/#/portal/signup/free) to the blog. This article will be a members only article!
It wasn't mentioned in Zendesk's Whats New, but the word this month has to be ChatGPT. Zendesk published a full overview on how their ready for AI, ChatGPT's API became finally available, and [every](https://youtu.be/75hvpxLErvU?ref=internalnote.com) [Zendesk](https://www.linkedin.com/feed/update/urn:li:activity:7044631879802650624/?ref=internalnote.com) [Partner](https://www.linkedin.com/posts/robert-cwicinski-15929740%5Fcustomerservice-chatgpt-openai-activity-7041789437780848640-Q0F1?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop) under the sun announced concepts or early looks on ChatGPT.
Zendesk also announced their **Relate 2023** event for May 11th. It'll be an online event, so don't forget to subscribe! and if you're a Zendesk developer you'll probably want to join the **Zendesk Developer User Group** on March 31st!
[Zendesk Relate 2023Zendesk invites you to our global flagship conferenceOpenform](https://event.zendesk.com/zendeskrelate2023broadcast3?ref=internalnote.com)
[Zendesk Developer User Group: Initial Meeting | Zendesk User Groups & Community EventsVirtual Event - Hello! If you are curious about Zendesk development, such as ZIS, ZCLI, Python, Liquid Markup, Zendesk Sunshine, etc, then you are in the right place! This is our first meeting and the intention is to meet each other and find out what the hot topics may be and what interests there ar…Zendesk User Groups & Community Events](https://zdsk.co/42BHrqp?utm%5Fcampaign=611beb0329938d00016c5537&utm%5Fcontent=641dfcf8348e4700015082bb&utm%5Fmedium=smarpshare&utm%5Fsource=linkedin)
# 🎉 New Releases
## Flow Builder improvements
Zendesk's bot keeps getting better month by month. This month we got an The *[Branch by condition](https://support.zendesk.com/hc/en-us/articles/5280598023450?ref=internalnote.com)* step which allows more complex logic based on parameters or choices made by the user. An example flow: If you use *Ask for Details* to ask the customer which country they live, you can then use the new conditions to automatically show a different branch for US, UK or EU users. Or, combined with authentication you can show a more personalised path for known customers, and a more generic one for guests.
🔈
I'm working on a big example bot flow that combines all new Flow Builder features in one both. [Keep your eyes](https://internalnote.com/#/portal/signup) open early april for this one!
## Branded Widget
From time to time new releases create some controversy. And when Zendesk announced their Messaging widget would show a branded footer for all non-Enterprise plans, that did create some negative feedback. I'm a big proponent of: If you use a tool you like you shouldn't hide it, but apparently some Zendesk customers didn't agree. Luckily the choice has already been reversed and Professional users can also hide this footer now.
## ZIS Playground
> [ZIS Playground](https://www.zendesk.com/marketplace/apps/support/651210/zis-playground/?ref=internalnote.com) is a tool for developers, so you can quickly learn to use Zendesk Integration Services capabilities, resources, and syntax. ZIS Playground walks you through the setup process and comes with a simple default Flow, so you can run your first test in just a couple of minutes.
🤔 I **really** should write about [ZIS](https://developer.zendesk.com/documentation/integration-services/?ref=internalnote.com). Although it's pretty early and not completely full featured yet, it's Zendesk's newest integration service that allows easy integration of Zendesk and external platforms via a low-code solution. This new Marketplace place app from Zendesk allows you to explore this new tool and get to know the syntax.
## The rest
- Small tweak in the way [Agent Workspace tabs](https://support.zendesk.com/hc/en-us/articles/5439612833562-Announcing-improved-Agent-Workspace-ticket-tabs?ref=internalnote.com) work.
- You can now merge one or more content tags in [Guide](https://support.zendesk.com/hc/en-us/articles/5000652575898-Creating-and-managing-content-collections-with-content-tags?ref=internalnote.com).
- Merging tickets has a [new alert](https://support.zendesk.com/hc/en-us/articles/5478155610522-Announcing-a-new-ticket-merge-warning-message?ref=internalnote.com) when merging tickets with requester or organisation mismatch. Good for Privacy breached, bad for agents. That flow really should be less complex, not more complex.
# 💡Insights
## ChatGPT
We're seeing more and more useful content about generative AI and ChatGPT as relating to Customer Care and Zendesk. Zendesk itself put a nice overview [on their blog](https://www.zendesk.com/newsroom/articles/generative-ai/?ref=internalnote.com), as well as did the Chatbot [Ada](https://www.ada.cx/posts/new-webinar-preparing-your-cx-organization-to-adopt-generative-ai?ref=internalnote.com).
> The biggest opportunity lies in using AI to eliminate much of the manual workload that can be low value and incredibly time consuming. Imagine an agent going from having to read through pages of text to get a summary of a customer’s previous issues, to getting an accurate, customised summary allowing them to solve customer issues much more quickly. That is the type of scenario LLMs enable and a Zendesk and ChatGPT partnership will provide to businesses. - **Cristina Fonseca**
And if you want to see how the above could work in Zendesk, there's a very good tutorial [on the Zendesk Developer website](https://developer.zendesk.com/documentation/apps/build-an-app/using-ai-to-summarize-conversations-in-a-support-app/?ref=internalnote.com) that shows how to build such an app.
# ⚠ Major Changes
> Important: Starting in May 2023, customers on Zendesk Suite plans will have a new usage-based pricing model for Article Recommendations and custom bots built in Flow Builder that will be based on **Monthly Active Users (MAU)**. The exact date for this change, along with other details, will be communicated to users in the coming weeks.
Since its arrival Answer Bot has always been a paid feature where you're paying Zendesk for each resolved inquiry. Now, Zendesk is also moving Flow Builder to a paid model. For each unique user (MAU) that contacts you over Messaging and interacts with Flow Builder, you're paying Zendesk for that assist. This is not necessary a bad thing. Customer inquiries resolved via Self Service lead to better CSAT and less Agent work.
[About Monthly Active Users for Zendesk botsImportant: Starting in May 2023, customers on Zendesk Suite plans will have a new usage-based pricing model for Article Recommendations and custom bots built in Flow Builder that will be based on M…Zendesk helpAimee Spanier](https://support.zendesk.com/hc/en-us/articles/5352026794010-About-Monthly-Active-Users-for-Zendesk-bots?ref=internalnote.com)
However, this could lead to surprises if you're not prepared. Luckily the change won't happen tomorrow, but the article below has some good suggestions on how to reduce the amount of MAU. Long story short: authentication is the way to go here. The more users authenticate, the less users you're counting double.
[Authenticate Zendesk MessagingZendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.Internal NoteThomas Verschoren](https://internalnote.com/jwt-messaging)
# 🎥 Videos
# 📝 Articles this month
- [Cleaning up Zendesk Organisations with a sidebar app](https://internalnote.com/cleaning-up-organisations-with-a-sidebar-app/)
- [Improve ticket deflection by enabling Federated Search](https://internalnote.com/federated-search/)
- [Make Protected Zendesk Help Center Articles available in Search](https://internalnote.com/zendesk-guide-membersonly/)
# And finally...
Ever needed to do [Account Assumption](https://support.zendesk.com/hc/en-us/articles/4408894200474-Assuming-end-users?ref=internalnote.com) for an Agent?
1. Go to [https://subdomain.zendesk.com/users?role\[\]=4&role\[\]=2](https://d3v-verschoren.zendesk.com/users?role%5B%5D=4&role%5B%5D=2&ref=internalnote.com)
2. Hover right next to the edit button
3. Click Assume
🥳
Thanks for reading this article and the blog. If you liked this content, please consider **subscribing** and **share** this article to your colleagues.
### Cleaning up Zendesk Organisations with a sidebar app
URL: https://internalnote.com/cleaning-up-organisations-with-a-sidebar-app/
Last updated: 2025-09-08T06:44:16.000Z
Are you a Zendesk admin struggling with managing organizations and their associated user data? If so, you're not alone.
Two common issues that Zendesk admins face are organisations being created after their users already exist, and not all users being a part of those organizations. Additionally, organizations can contain bad data and users who don't belong to the organization.
Zendesk has a longtime feature where you can automatically assign users to an organization by mapping its custom email domain. Any user that matches that domain will be added to the organization. And while creating organizations witha [ mapped email domain](https://support.zendesk.com/hc/en-us/articles/4408882246298-Creating-organizations?ref=internalnote.com#topic%5Fnxl%5Fvdt%5Fbc) resolve this issue, it only works for users created after the fact and doesn't always update existing users.
In this blog post, I'll explain how to build/use a new sidebar app that runs in Zendesk Agent Workspace and cleans up your organisations one at a time by leveraging the ZAT and Zendesk APIs. With this app, you'll be able to easily manage your organisation data and improve your customer service workflows.
# The Fix
In this article I'll show you how to use/build **Organisation Cleaner,** a sidebar app designed to help Zendesk admins manage their organisation data more effectively. This app leverages a few APIs to allow admins to quickly clean up any organisation based on the domains setup for that organisation.
The app's primary function is to add missing users or remove bad ones that don't match the domain. This helps to ensure that your organisation data is clean and accurate, which in turn helps to improve your customer service workflows.
To use the app, you can download the latest release from [GitHub](https://github.com/verschoren/organization%5Fmanager/releases/tag/v1.0.0?ref=internalnote.com) and install it as a [private app](https://developer.zendesk.com/documentation/apps/getting-started/uploading-and-installing-a-private-app/?ref=internalnote.com) in your Agent Workspace. From there, you can select the organisation you want to clean up and run the app. The app will then cleanup the organisation in browser so no data is shared with external sources.
[GitHub - verschoren/organization\_manager: This app will help you clean up your organizations in ZendeskThis app will help you clean up your organizations in Zendesk - GitHub - verschoren/organization\_manager: This app will help you clean up your organizations in ZendeskGitHubverschoren](https://github.com/verschoren/organization%5Fmanager?ref=internalnote.com)
🤔
****Why is this a downloadable app and not a Marketplace app?**
The purpose of this blog is explaining how Zendesk works and empower Zendesk Admins and Developers. You can take this app as a starter to make it work specifically for your organization, or use it as is.
If it really provides value for your company, consider subscribing to a [paid tier](https://internalnote.com/#/portal/signup/63ff05d99deb9e003d619e70/monthly) of this blog to make more of these tools possible. And if you need help building this, there's always a [Zendesk Partner](https://www.zendesk.com/marketplace/partners/?query=premium&ref=internalnote.com) nearby!
# The App
The application runs as a sidebar app in the Organisation View of Agent Workspace. Navigate to any organisation and toggle the Apps Panel to get started.
## Test Mode
The app has a build-in test mode . By default we only compare data, but don't update anything in your instance yet.
```javascript
$('#test_mode').change(function() {
if(this.checked) {
test_mode = true;
} else {
test_mode = false;
}
});
```
Similarly, if you want to look at the results in depth, each user logged in test mode has a clickthrough to the user profile, and we show real updates with a notification top right.
```javascript
//show user profile
client.invoke('routeTo', 'user', $(this).data('target'));
//show notification
client.invoke('notify', 'Users added!');
```
## General Logic
The app has two options: **Remove Bad Users** and **Add Existing Users.**
Both use `ZAT` to get info on the current location the user is looking at in our main `handleFlow()` logic.
```javascript
client.get('organization').then(function(organization) {
var domains = organization.organization.domains.split(' ');
var organization_id = organization.organization.id;
//Handle app logic
});
```
## Cleaning up an existing organisation
When we want to clean up an existing organisation we use the `removeUsers()` function.
It uses the `/api/v2/organizations/${organization_id}/users` endpoint to get all users in the organisation you're looking at. It then takes the Domains you added to that organisation and compares it to the Users in the organisation.
```javascript
let options = {
contentType: "application/json",
url: `/api/v2/organizations/${organization_id}/users`,
type: "GET",
};
client.request(options).then((users) => {
var users = users.users;
jQuery.each( users, function( i, val ) {
var domain = val.email.split('@')[1];
if (domains.indexOf(domain) == -1) {
//update user
}
});
});
```
For all users where the domain doesn't match ( `indexOf == -1` ), we add the user to an `unmatchedUsers` array and print the output to a log via `printOutput()`
### Example: removing unmatched users
In the example below I scanned the organisation **Avengers** for any user that didn't have an `@avengers.example` email-address and remove the bad **Hydra** agent `winter.soldier@hydra.example`
## Adding missing users
Adding missing users works similarly. Since an organisation has potentially multiple domains we run the search for each domain setup in our `handleFlow()` function.
```javascript
domains.forEach(function(domain) {
addUsertoDomain(domain,organization_id);
});
```
We first use the search API to search all users that match the domain and aren't part of the organisation via `/api/v2/search.json?query`. Note the `-organization` which excludes users in the organization.
```javascript
function addUsertoDomain(domain,organization_id){
let options = {
contentType: "application/json",
url: `/api/v2/search.json?query=type:user ${domain} -organization:${organization_id}`,
type: "GET",
};
client.request(options).then((users) => {
var users = users.results;
//update users
});
}
```
Once we find our users, we add the user to an `unmatchedUsers` array and print the output to a log via `printOutput()`
### Example: adding missing users
In the example below I had seven dwarfs in Zendesk already, but never bothered to create an organization for them. Once I created the organization **Seven Dwarfs** with email @`sevendwarfs.example` the app found our seven friends, and added them to the organization
## Handling the update
If our `test_mode` is false and we have an `unmatchedUsers` array, we can handle updating the users via `createOrUpdateMany()`. This does a basic POST to the Zendesk endpoint that allows us to update up to 100 users at once.
The reason we want to use the `create_or_update_many` endpoint is since this a single API call that ends up on a Job in Zendesk. We don't need to wait for multiple calls to succeed, and we don't overload the API limits.
```javascript
function createOrUpdateMany(unmatchedUsers,message){
let options_nomatch = {
contentType: "application/json",
url: `/api/v2/users/create_or_update_many`,
type: "POST",
data: JSON.stringify(unmatchedUsers)
};
client.request(options_nomatch).then((update_many) => {
client.invoke('notify', message);
});
}
```
However, since this all happens in the background, we do want to show some output. This happens with our `PrintOutput()` function. It adds a list of log lines to our app nicely formatted with an icon (✅/❌) and the users name/email.
[GitHub - verschoren/organization\_manager: This app will help you clean up your organizations in ZendeskThis app will help you clean up your organizations in Zendesk - GitHub - verschoren/organization\_manager: This app will help you clean up your organizations in ZendeskGitHubverschoren](https://github.com/verschoren/organization%5Fmanager?ref=internalnote.com)
# A few caveats
Zendesk returns paginated results. This app only takes into account the first page of results.
For cleaning up an organization I figured chances of organizations having more than 100 users is slim. If you want, you can always update the app to add pagination and add a pull requests [here](https://github.com/verschoren/organization%5Fmanager/issues?ref=internalnote.com).
To add missing users you're in luck. Each time you press the button we'll pull in the next 100 missing users (if there are any) since the API search excludes already added users.
The `readme.md` on GitHub also comes with a sample payload to import. You can use these two user sets to easily test and validate the app.
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** and ****share** this article to your colleagues.
### Improve ticket deflection by enabling Federated Search
URL: https://internalnote.com/federated-search/
Last updated: 2025-09-08T06:44:20.000Z
Federated Search lets your customers seamlessly search across multiple platforms and websites simultaneously.
In a Zendesk context, federated search enables customers to search for the information they need across the website, blog, and FAQ. This means that customers can easily find the answers they need without having to submit a ticket. One key advantage of federated search is that it allows customers to search for blog articles and website sales content within the same search experience as regular FAQ articles.
This can greatly benefit customers as they can access a wider range of information in one go, without having to switch between different search interfaces or platforms. By including blog articles and website content in the search results, customers can also benefit from a more holistic understanding of the company's products and services. Furthermore, by providing customers with more relevant information, they are more likely to find the answers they need and reduce the likelihood of submitting a ticket and improve CSAT.
Federated search can also help to reduce the workload of customer support teams. By enabling customers to find the information they need on their own, support teams can focus on more complex tickets that require their expertise. This can improve the efficiency of the support team, as they can allocate their time and resources more effectively and reduce their workload.
[About Zendesk Federated SearchWhat’s my plan? Help center federated search lets your end users see content in your help center search results that is external to your help center. This means that when an end user searches in…Zendesk helpNova Dawn](https://support.zendesk.com/hc/en-us/articles/4408830243482-About-Zendesk-Federated-Search?ref=internalnote.com)
🔒
Note: Federated Search is only available for Zendesk (Suite) Enterprise users.
# Federated Search in Zendesk
There's a two ways you can enable Federated Search in Zendesk:
## Search crawler
This is the easiest way to setup Federated Search. Zendesk looks at your website's sitemap and indexes all publicly available content on a nightly basis.
It has the benefit of being an almost no-code integration with the exception of one verification tag to be added to your websites ``
## Federated Search API
This API endpoint lets you manually add records to the search database of your Zendesk instance. You can do this one article at a time, or you can use a tool like Zapier or a custom script to update the database whenever a new page/article/item is added to your website.
# Demo
The Demo Help Center on support.internalnote.com has Federated Search enabled. It indexes this website and also has a bunch of loose websites added to its index.
If you search for e.g. LEGO it'll show support articles from the FAQ, an article from this Blog, and a few external links in its results.
[DEMO: Search for LEGO](https://support.internalnote.com/hc/en-us/search?utf8=%E2%9C%93&query=lego&ref=internalnote.com)
# Automatic Search Crawler
Most customers I've seen have a lot of duplicate content on both their website and Help Center. The website explains the what/how of their offerings from a commercial standpoint, and the help center has almost identical article so that customers contacting them via web/email/social can find that same content via search or Answer Bot. For this reason I recommend to index at least your website inside your Help Center so you can remove a lot of this semi-duplicate content.
## Setup
Setting up an automatic Search Crawler is five easy steps:
1. Add Website Sitemap to Crawler (e.g. `https://internalnote.com/sitemap.xml`)
2. Validate ownership by adding `{{article.body}}
`
4. Replace it with the code below.
```html
{{#if signed_in}}
{{article.body}}
{{else}}
{{excerpt article.body characters=300}}
This article is for members only.
{{/if}}
{{article.body}}
`
4. Replace it with the code below.
```html
{{#each article.labels}}
{{#is identifier 'preview'}}
{{#if ../signed_in}}
{{../article.body}}
{{else}}
{{excerpt ../article.body characters=300}}
This article is for members only.
{{/if}}
{{else}}
{{../article.body}}
{{/is}}
{{/each}}
CF-Access-Client-Secret:
```
### Which, currently, is not compatible with Zendesk Webhooks.
### **Asana**
And finally, one last one we ran into lately: [Asana](https://developers.asana.com/docs/create-a-task?ref=internalnote.com), where once again the requirement for two additional header items defining accepted content-types are impossible to submit, and thus impossible to use.
```curl
curl -X POST https://app.asana.com/api/1.0/tasks \
-H 'Content-Type: application/json' \
-H 'Accept: application/json' \
-H 'Authorization: Bearer {access-token}' \
-d '{"data": {"field":"value","field":"value"} }'
```
### **Workaround**
Mentioning issues is one thing. Fixing them is another.
Currently my workaround is the following:
1. I create a Cloudflare Worker as a proxy for each webhook I need in Zendesk
2. My Webhooks in Zendesk POST to the Worker
3. The Worker processes the request, appends a `header` and forwards the request to the real API
4. The API responds to the request call
5. The Worker relays the response to Zendesk
See [this Worker](https://developers.cloudflare.com/workers/examples/alter-headers/?ref=internalnote.com) as a generic example.
### **Zendesk Feature Request**
A more structural solution would be Zendesk adding support for custom headers in Webhooks.
If you like to see this too, please upvote [this](https://support.zendesk.com/hc/en-us/community/posts/4971033907994-Need-for-Custom-Headers-in-Webhooks?ref=internalnote.com) community post and add your feedback.
[See Community Post](https://support.zendesk.com/hc/en-us/community/posts/4971033907994-Need-for-Custom-Headers-in-Webhooks?ref=internalnote.com)
### Flow Builder - Ask for details
URL: https://internalnote.com/flow-builder-ask-for-details/
Last updated: 2025-09-08T06:45:05.000Z
Last summer Zendesk added the ability to [make API calls from within Flow Builder](https://www.tripelhop.dev/blog/zendesk-flow-builder-api-call?ref=internalnote.com). Instead of having a Chat bot that *just* replies with Help Center articles, you can now reply with there data types too by calling your own or external APIs.
One of the improvements we listed was the ability to ask the customer for data, and use that as part of our query. Example: ask for order number, and retrieve status of that order.
Luckily, Zendesk has just enabled this feature, making the range of options for Flow Builder way bigger by adding an *Ask for Details* step.
### **Concept**
[OMDb API](https://www.omdbapi.com/?ref=internalnote.com) is a free API that enabled movie lookups from within IMDB. We’ll use this API to have Flow Builder ask our customer for a movie title, and reply with its full title and year of release. You can test this bot via the Widget below.
In reality you’ll probably use this API to show the user order information, system status updates, product pricing or other types of more useful data.
`https://www.omdbapi.com/?apikey=[yourkey]&t=[query]`
## Setup
### Custom Field
The *Ask for Detail* step stores the customer reply in a custom field. You can use an existing one, or create a new one for this specific step. Make sure you choose Text Field if you want the customer to enter any kind of text.
For this demo I’ve added a *DEMO Movie* text field.
#### **Add a new step to gather details**
Go to Bots in the Admin Center and open an existing answer, or create a new answer for your Bot.
Once in edit mode you can add the first new step to your bot flow:
- In our example we already have an existing Bot that *Presented Options*, so we added a new option called *Movie Database.*
- Underneath that option we add an *Ask for Details Step*
- Give it a title (Enter a movie title) and link it to the new Custom Field you created above.
#### **Add API call**
Next we’re going to add the actual API call
- Under the *Ask for Detail* step we add an *Make API call* step
- Give it a name (Finding Movie Details), description, and add the API details. Important here is to add the API and then choose *Add a variable* to select the *Response from the customer.*
`[GET] https://www.omdbapi.com/?apikey=[yourkey]&t=[your custom field]`
No header info needed
Once you added the API call we can validate it by running a test.
- In our test we search for Blade Runner
- The API gets called and returns a response
- You can browse through the response and store specific keys as variables to use in your responses.
- In our example we store the *Name* and *Release Date* keys for the first matched movie as variables *year* and *title.*
The variables are stored and worked for all successful API calls. So for each query your customers execute we store the first matched movie as the variables for a reply.
#### **Reply to the customer**
And we can now add two final steps to our Flow:
1. Send Message under successful API call
2. Send Message under failed API call
The success message used the variables we created above. So we’ll always reply with:
> “The movie \[title\] was released on \[year\]”
And the failed message can reply with:
> “No match found for \[DEMO Movie\]”
### **Customer Experience**
Once you enable the new flow, this is the experience your customers get:
### **Conclusion**
This new *Ask for Details* option allows for more personalised Flow Builder experiences that can pull in external data based on the customer request instead of generic information or only Help Center articles.
Currently the *Call an API* option is still fairly limited in its reply option. You can basically reply with hardcoded variables in text form.
I’d love to see richer formats. Loop across results and reply with a list of results. Or parse the returned data and render it as images, links or buttons.
Hopefully we’ll see this sooner than later.
### Custom Launcher for Zendesk Messaging and Unread Counts
URL: https://internalnote.com/custom-launcher-for-zendesk-messaging-and-unread-counts/
Last updated: 2025-09-08T06:45:15.000Z
Zendesk recently added the ability to their Messaging Widget to use Custom Launchers. This article shows you how to easily implement it.
A longstanding request for both the Classic and Messaging Widget is the ability to customize the Widget completely with custom icon, color, shape or positioning.
Although Zendesk offers the option the change color and text, fully replacing the design used to require hiding the widget and managing its visibility for every scenario manually.
Recently Zendesk [added the ability](https://support.zendesk.com/hc/en-us/articles/4500747797914?ref=internalnote.com#id%5Fjnl%5Fxtm%5Fgnb) to their Messaging Widget to implement Custom Launchers with a lot more ease. By enabling the custom behaviour, Zendesk will never load its own launcher button, and will give feedback on the current Chat session via a callback function so you can create a custom launcher, and easily reflect read/unread status.
⁉️
When using the Custom Launcher for Messaging, [proactive](https://internalnote.com/proactive-ticketing-for-messaging/) messages are ****not supported!** See [this](https://support.zendesk.com/hc/en-us/articles/5381304334234/comments/5575083752986?ref=internalnote.com) comment
Parallel to this new design option, Zendesk also added a function that allows you to get the unread count of your conversation. You can use this to show the amount of unread messages to your user, or react when the unread count reaches a certain treshhold.
The Default Widget layout and examples of custom designs.
## Feature Overview
By enabling the Custom Launcher in Zendesk, you can embed the widget without the traditional launcher button floating bottom right.
You can enable the new behaviour by going to your Zendesk Admin Panel and [toggling the appearance](https://support.zendesk.com/hc/en-us/articles/4500747797914?ref=internalnote.com#id%5Fjnl%5Fxtm%5Fgnb) from square/circle to custom. This will hide the launcher button from any website the widget has been added to.
Once enabled you can:
- Add a *Contact us* link on your webpages which will open the widget if clicked.
- Design and load your own launcher button which can open the widgetto align your Support experience with your brand.
- ...
## Demo Page
I created a working demo [here](https://jwt.internalnote.com/customlauncher.html?ref=internalnote.com) to test out this new behaviour.
It allows you to launch the Messaging Widget via a Contact Us button on the middle of the page, or (better) use a custom designed launcher button to open the Messaging Widget.
Since Zendesk also offers a function to get updates on read/unread messages, the pages also contains a demo of this feature.
When you start a conversation the Answer Bot and close the widget the notification will appear and reflect the amount of unread messages.
Alternatively, by pressing the *Demo →* button , the Launcher Button will display a hardcoded demo of 9 unread messages.
Naturally, this page is a demo page, so you can adapt the design and flow to reflect your own use case.
### **Code**
Let's dive into the code to see how this works:
[GitHub Repository](https://github.com/verschoren/zendesk%5Fwidget/blob/main/docs/customlauncher.html?ref=internalnote.com)
#### **Contact Us button**
The contact us button is pretty straightforward. Clicking the button triggers an *OpenWidget()* function that well.. opens the widget.
```html
Contact Us
```
#### **Custom Launcher**
The Custom Launcher works similarly, but consists of a staticly placed *div* containing the launcher bottom right, and a second *div* placed relative to that launcher to show the counter.
We hide the counter by default, and use the Unread Counter to show/update its count.
```html
```
#### **Unread Counter**
The Unread Counter is a live updating function that provided continuous feedback on the amount of unread messages in the widget.
By implementing a callback that updates our Launchers' Notification Badge we can choose when our Launcher should show a notification, and update the count live.
Or, if you want, you can even decide to open the widget once the amount of unread messages reaches a certain treshold:
```javascript
zE('messenger:on', 'unreadMessages', function (count) {
// if there are unread messages, show badge
if (count > 0){
$('#counter').html(count);
$('#notification').show();
}
// if there are more than 4 unread messages, show the widget
else if (count >= 4){
zE('messenger', 'open');
}
//else hide the counter
else {
$('#notification').hide();
}
})
```
### Customize and brand your Classic Zendesk Widget
URL: https://internalnote.com/customize-and-brand-your-zendesk-widget/
Last updated: 2024-08-19T20:40:18.000Z
Zendesk has a nice Web Widget that allows you to embed your contact channels and FAQ on any website. This allows your customers to have the same experience when they contact you via your website, web shop or FAQ and makes your Help Center content available anywhere on your website.
Questions about compatibility while looking at a product on your website? The FAQ search in the Widget has got you covered. Urgent questions about an expired promo codes while checking out? Live Chat's just a click away.
The widget can be embedded on any website, and Zendesk [offers a native customisation](https://support.zendesk.com/hc/en-us/articles/4408836216218?ref=internalnote.com) to tweak the Widget's primary color so it aligns to your branding.
But what if we take it one step further? Completely customise the Widget's branding with custom colours, logos and titles?
This not only differentiates your Support offering from any other Zendesk customer, it also allows you to add some personality to your Customer Care with custom colours, logos and naming conventions.
Note, this guide applies to the Classic Zendesk Widget.Zendesk Messaging currently only offers a [limited set](https://support.zendesk.com/hc/en-us/articles/4409103246874-Creating-a-messaging-Web-Widget?ref=internalnote.com) of options via the Admin Panel.
# How does it work?
Customising the widget requires adding some extra Javascript code on your website just below the embed code Zendesk provides.
See example here:
[Internal Note - Zendesk Classic Widget CustomizationInternal Note - Zendesk Classic Widget CustomizationZendesk Classic Widget CustomizationInternal Note](https://widget.internalnote.com/classiccustom?ref=internalnote.com)
## Colors
You can overwrite the default color and assign custom colours to a couple of elements:
```javascript
window.zESettings = {
"webWidget": {
"color": {
"theme": "#007aff",
"launcher": "#FFCC00",
"launcherText": "#142b39",
"button": "#007AFF",
"resultLists": "#007AFF",
"header": "#FF443A",
"articleLinks": "#007AFF"
}
}
}
```
- Theme: the primary color set by the Widget Settings
- Launcher and Launcher Text: the color of the button and its text specifically.
- Button: Any button in the Widget UI
- Result List: color of search results from the Guide search
- Header: Color of the Widget's header
- Article Links: Color of links in text.
## Launcher
By default the widget has Support or Live Chat as labels for the Widget for when chat is respectively offline and online.
You can override this text to have a specific call to action:
```javascript
window.zESettings = {
"webWidget" : {
"launcher": {
"label": {
"*": "Customer Care"
},
"talkLabel": {
"*": "Call Us"
},
"chatLabel": {
"*": "Live Chat"
}
}
}
}
```
Putting it all together creates a redesigned widget like this:
## Answer Bot and Chat
Going one step further we can also tweak the design of your Answer Bot and Chat Agents:
- Answer Bot allows for a custom Avatar, name and title.
- Chat allows a custom Avatar, title and subtitle.
```javascript
window.zESettings = {
"webWidget": {
"answerBot": {
"avatar": {
"url": "https://website.com/avatar.png",
"name": {
"*": "Internal Note Bot"
}
},
"title": {
"*": "Customer Care"
}
},
"chat": {
"concierge": {
"avatarPath": "https://website.com/avatar.png",
"name": "Customer Care Team",
"title": {
"*": "We take care of you"
}
},
"title": {
"*": "Chat to Us"
}
}
}
}
```
Note that the title and name can be Locale aware by specifying more locales:
```javascript
...
"name": {
"*": "Our Support Bot",
"nl": "Onze Hulp Bot",
"fr: "Notre Bot D'Aide"
}
...
```
# The result
[Internal Note - Zendesk Classic Widget CustomizationInternal Note - Zendesk Classic Widget CustomizationZendesk Classic Widget CustomizationInternal Note](https://widget.internalnote.com/classiccustom?ref=internalnote.com)
Putting all the code together:
```javascript
window.zESettings = {
"webWidget": {
"contactOptions": {
"enabled": true
},
"color": {
"theme": "#007aff",
"launcher": "#FFCC00",
"launcherText": "#142b39",
"button": "#007AFF",
"resultLists": "#007AFF",
"header": "#FF443A",
"articleLinks": "#007AFF"
},
"answerBot": {
"avatar": {
"url": "internalnote_social@2x.png",
"name": {"*": "Internal Note Bot"}
},
"title": {
"*": "Customer Care"
}
},
"chat": {
"concierge": {
"avatarPath": "internalnote_social@2x.png",
"name": "Customer Care Team",
"title": {"*": "We take care of you"}
},
"title": {
"*": "Chat to Us"
}
},
"launcher": {
"label": {
"*": "Customer Care"
},
"talkLabel": {
"*": "Call Us"
},
"chatLabel": {
"*": "Live Chat"
}
}
}
}
```
## **Further reading**
The Zendesk Widget offers a whole lot more customisation options, going from showing/hiding parts of the UI, changing label text, filtering search results e.a.
Check out the [API](https://developer.zendesk.com/api-reference/widget/introduction/?ref=internalnote.com) or the awesome [Widget Guide](https://widget.premiumplus.app/?ref=internalnote.com) I build for Premium Plus.
### ticket.sendMessage() for Support Apps
URL: https://internalnote.com/ticket-sendmessage-for-support-apps/
Last updated: 2025-09-08T06:45:45.000Z
Zendesk's release notes [last week](https://support.zendesk.com/hc/en-us/articles/4686897341466-Release-Notes-Through-2022-05-28?ref=internalnote.com) had a short entry on a new action for sidebar apps:
> We’ve launched functionality for **Zendesk Support Apps to send messages to end-users on behalf of agents** through [Zendesk Messaging](https://support.zendesk.com/hc/en-us/articles/4408882490778-Introducing-messaging?ref=internalnote.com) and [Social Messaging channels (through Sunshine Conversations)](https://support.zendesk.com/hc/en-us/articles/4408836484378-Adding-Sunshine-Conversations-channels-to-the-Zendesk-Agent-Workspace?ref=internalnote.com) in Agent Workspace. Previously, this API only worked for chat. Developer documentation lives [here](https://developer.zendesk.com/api-reference/apps/apps-support-api/ticket%5Fsidebar/?ref=internalnote.com#ticketsendmessage).
This new action is basically an extension of the existing `appendComment()` action that already [existed](https://developer.zendesk.com/api-reference/apps/apps-support-api/ticket%5Fsidebar/?ref=internalnote.com#commentappendtext) for ticket based channels (email, webforms, talk,..), but this time for Conversation based channels like (Web) Messaging and Chat:
## ticket.sendMessage
Sends a message to the end-user, on behalf of the agent. Only works during live chat (channel = chat) and messaging (channel = messaging) conversations for the time being.
```javascript
client.invoke(
'ticket.sendMessage',
{
channel: value1,
message: value2
}
)
```
## So, what can you do we it?
Basically this allows app developers to create richer experiences for agents with more capabilities than traditional macros.
Where a **macro** allows agents to have pre-formatted content to send to customers, including placeholders like name, custom fields, ticket id,..., they do not allow for any external data or complex variables.
This is where this **action** is useful. You can use it as a final step in a sidebar app after you've pulled in data from external resources like an order system, CRM, Marketing Tool, Invoice system....
#### **Example use cases**
- When a customer asks for their Gift Card saldo, we can pull in the data from our invoice tool and reply with "You currently have 42$ available on your gift card"
- When a customer has ordered an item, we can pull in the order status and reply with "Your order is currently shipped with tracking number TEEKAY-421"
- When a customer has send in a repair we can reply with the status of their repair
- When a customer has setup an appointment we can reply with "Your next appointment is confirmed for May 4t at 11:38AM"
- ...
### **Demo App: send order status to customer**
Imagine the following scenario:
- A customer inquires about their order status
- The agent quickly asks for their order number via a button in a custom sidebar app
- The customer replies with an order number
- The agent fills in that order number in the sidebar app, which then retrieves the order status via API from the webshop back-end.
- The agent can then chose to send the customer the order status (It's shipped!) or an overview of the entire order.
Sounds cool right? That's exactly what the app we're building in this article does!
#### **Code overview**
You can find the full app in the repository above, but below are a few highlighted items specific to this feature.
[GitHub - verschoren/internalnote\_sendmessage\_example: Demo App to showcase the sendMessage App featureDemo App to showcase the sendMessage App feature. Contribute to verschoren/internalnote\_sendmessage\_example development by creating an account on GitHub.GitHubverschoren](https://github.com/verschoren/internalnote%5Fsendmessage%5Fexample?ref=internalnote.com)
#### **Channel filtering**
Since the `sendMessage()` action only works for chat and messaging, we first retrieve the current ticket the agent is looking at and store its channel in a variable.Since Zendesk calls *messaging* '*native\_messaging*' internally, we rename the channel for convenience later.
```javascript
client.get('ticket').then(function(ticket) {
var channel = ticket['ticket'].via.channel
channel = channel == 'native_messaging'? 'messaging' : channel
if (channel != 'messaging' && channel != 'chat'){
//Do nothing
} else {
//Do something
}
});
```
### Sending a message
This function is the core of the app and maks use of the new `sendMessage()` action.Note that *channel* has to be either chat or messaging, it won't work for any other channels and throw an error.
Message can be any kind of plain text. You can add line breaks via `\n` but no markup in HTML or Markdown is supported, and neither are attachments.
> You can call this action multiple time in a row if you want to send out a burst of short replies instead of one big text block, making it more native feeling.
```javascript
function sendMessage(channel,message){
client.invoke('ticket.sendMessage', {
channel: channel,
message: message
})
}
```
#### **The rest of our demo app is pretty basic:**
- We have a fake JSON dataset which in production should pull data from a live web API based on the entered Order ID.
- We have 3 click `onClick()`\-events that send out specific messages based on the agents' choice.
- Since we already capture the ticket data to check the channel, we also pull in the requester name to make the messages more personal.
```javascript
$( "#info" ).click(function() {
sendMessage(
channel,
`Hello ${name}, before we can assist you we need your order number.`
)
});
```
🥳
Thanks for reading this article and the blog. If you liked this content, please consider ****subscribing** via email or ****share** the article to your colleagues.
### Classic Web Widget Authentication
URL: https://internalnote.com/classic-web-widget-authentication/
Last updated: 2024-08-19T20:32:25.000Z
You can authenticate your Zendesk Web Widget to identify your users, and to make restricted Help Center content available to users.
By default Zendesk' Web Widget has no idea who's looking at the website. Similarly, when searching the Help Center via the widget, it will only show publicly available articles, and no articles assigned to a specific segment of users or your agents.
To fix this you can pre-fill or authenticate the Web Widget.
- [Authenticated users in Chat SDK](https://develop.zendesk.com/hc/en-us/articles/360052354433-Enabling-authenticated-users-with-the-Chat-SDK-?ref=internalnote.com)
- [Authenticated Visitors in Widget](https://support.zendesk.com/hc/en-us/articles/360022185314-Enabling-authenticated-visitors-in-the-Chat-widget?ref=internalnote.com)
Setting it up is easy, as long as you have a working JWT endpoint. If you haven't got one, this guide will show you how to set one up using Cloudflare Workers.
[GitHub - verschoren/zendesk\_widget: Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand.Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand. - GitHub - verschoren/zendesk\_widget:…GitHubverschoren](https://github.com/verschoren/zendesk%5Fwidget?ref=internalnote.com)
💡
Note that this example code does not validate the user against any directory. It trust the input and returns a valid JWT for Zendesk to use.
## How it works
Let's start at the end of the flow and show a working scenario.
1. A user visits [https://jwt.internalnote.com](https://jwt.internalnote.com/messaging.html?ref=internalnote.com)
2. They enter their name/email and press login
3. The website logs in the user and the widget authenticates
4. Zendesk recognises the user and fills in their name and email in the prechat form and form fields.
5. When the user searches the Help Center, restricted content becomes available.
The agent doesn't really see that the user was logged in during the Chat, but he will see the correct name, email and page view history.
## Setting up Authentication
Authenticating a user requires the following steps:
1. Get the necessary secrets from Zendesk for Chat and Guide
2. Create a web service (e.g. via Cloudflare Workers) to generate a valid Valid JWT
3. Have a function on your website that calls the web service when a user logs in and generates a JWT based on their email, name and ID
4. Push that JWT Tokens to the Widget
### Credentials - Chat
Setting up authentication for Chat first requires generating a Secret.
That's done via [https://subdomain.zendesk.com](https://subdomain.zendesk.com/admin/account/security/end%5Fusers?ref=internalnote.com#messaging)[/chat/agent#widget/widget\_security](https://d3v-verschoren.zendesk.com/chat/agent?ref=internalnote.com#widget/widget%5Fsecurity)
### Credentials - Guide
Setting up authentication for Guide requires generating a Secret and filling in a list of allowed d0mains:
This can be done via [https://subdomain.zendesk.com/admin/channels/classic/web-widget](https://subdomain.zendesk.com/admin/channels/classic/web-widget?ref=internalnote.com)
Note that you need to enter all websites where you want authentication to work. You can do this via \*.domain.com to capture any subdomain for your website(s).
To generate a login for the web widget you'll also need a name, email and external ID for your user.
## Generating the JWT
Based on the above items you can generate a working JWT. You can find an example below.
There's a few important caveats:
- External ID has to be unique for each user, so use a GUID, UUID, or database ID
- Our code does **not** validate if the user exists, it just accept whatever input is there.
#### **Chat**
```javascript
var input = {
"name":"Thomas Verschoren",
"email":"thomas@verschoren.com",
"external_id":"123456"
}
const secret = "abc123";
var token_raw = {
iat: Math.floor(Date.now() / 1000),
name: json.name,
external_id: json.external_id,
email: json.email,
};
var token = jwt.encode(token_raw, secret);
```
#### **Guide**
```javascript
var input = {
"external_id":1906365876753,
"user_email":"thomas@verschoren.com",
"user_name":"Thomas Verschoren"
}
var secret = "abc123";
var token_raw = {
jti: Math.floor(Math.random() * 10000000),
iat: Math.floor(Date.now() / 1000),
name: json.user_name,
email: json.user_email,
};
var token = jwt.encode(token_raw, secret);
```
For the example website above I created a Cloudflare Worker that handles the JWT scenario for both Messaging and the Classic Widget.
You can find a working code example via the link below.
[GitHub - verschoren/zendesk\_widget: Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand.Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand. - GitHub - verschoren/zendesk\_widget:…GitHubverschoren](https://github.com/verschoren/zendesk%5Fwidget?ref=internalnote.com)
#### **Logging In**
The end result is a JWT token that you pass to the widget via:
```javascript
//Prefill Forms and Chat
zE('webWidget', 'identify', {
name: user.name,
email: user.email,
});
zE('webWidget', 'prefill', {
name: {
value: user.name,
readOnly: true
},
email: {
value: user.email,
readOnly: true
}
});
//Authenticate
window.zESettings = {
webWidget: {
authenticate: {
jwtFn: function (callback) {
callback(jwttoken);
}
}
}
}
```
### **What about your Help Center?**
You can also easily use the above example to authenticate the widget for users browsing your Help Center.
1. Uncheck "Add Widget to Help Center" on [https://subdomain.zendesk.com/admin/channels/classic/web-widget](https://subdomain.zendesk.com/admin/channels/classic/web-widget?ref=internalnote.com)
2. Copy the Widget Embed Code on that same page.
3. Open your Zendesk Guide Theme and [edit its code](https://support.zendesk.com/hc/en-us/articles/4408832558874-Editing-the-code-for-your-live-help-center-theme?ref=internalnote.com)
4. Open `document_head.hdbs` and paste the Widget Embed Code at the bottom.
5. Also paste the following code just below the embed code.
[Internal Note - Zendesk Classic Widget CustomizationInternal Note - Zendesk Classic Widget CustomizationZendesk Classic Widget CustomizationInternal Note](https://widget.internalnote.com/guide%5Fclassic/?ref=internalnote.com)
From now on, whenever a user logs into your Help Center, the widget will authenticate itself.
### Authenticate Zendesk Messaging
URL: https://internalnote.com/jwt-messaging/
Last updated: 2025-12-02T17:57:51.000Z
⚠️
Apparently old Zendesk accounts can be tagged with an internal flag that prevents External ID matching to work. Zendesk Support removed the flag from my account and External ID tagging now works as expected. [Look at this article for more info](https://internalnote.com/deepdive-into-messaging-profiles/).
This means Email matching is still an issue, but External IDs do work!
Zendesk recently added the ability to authenticate users in the Zendesk Messaging Web and Mobile SDK. This article shows how to set it up with sample code.
Zendesk recently added the ability to [authenticate users in the Zendesk Messaging](https://developer.zendesk.com/documentation/zendesk-web-widget-sdks/sdks/web/sdk%5Fapi%5Freference/?ref=internalnote.com#authentication) Web and Mobile SDK. This allows any website or app that has logged in users to pass that information to Zendesk so that you're sure you're talking to the right person, and removes the need for your customer to enter any credentials.
Setting it up is easy, as long as you have a working JWT endpoint. If you haven't got one, this guide will show you how to set one up using Cloudflare Workers.
Note that this example code does not validate the user against any directory. It trust the input and returns a valid JWT for Zendesk to use.
[See Demo](https://demo.internalnote.com/messaging.html?ref=internalnote.com)
## How it works
Let's start at the end of the flow and show a working scenario.
1. A user visits [https://demo.internalnote.com/messaging.html](https://demo.internalnote.com/messaging.html?ref=internalnote.com)
2. They enter their name/email and press login
3. The website logs in the user and the widget authenticates
4. Zendesk recognises the user, and starts a conversation.
If there's still an ongoing conversation from earlier, the widget will show the conversation regardless of the user having used that browser/device before.
When the agent opens the conversation in the Agent Workspace he notices a green checkbox next to the users' name showing them the user has logged in correctly.
💡
Note: Zendesk has different approaches on how to map Messaging users to existing user profiles. Take a look [here](https://internalnote.com/messaging-authentication-identify-and-merge-existing-users?utm%5Fsource=demo) for more information
## Setting up Authentication
Authenticating a user requires the following steps:
1. Get a Secret and App ID from Zendesk
2. Create a web service (e.g. via Cloudflare Workers) to generate a valid Valid JWT
3. Have a function on your website that calls the web service when a user logs in and generates a JWT based on their email, name and ID
4. Push that JWT Token to the Widget
### Credentials
Setting up authentication for Messaging first requires generating a Secret and App ID. That's done via [https://subdomain.zendesk.com/admin/account/security/end\_users#messaging](https://subdomain.zendesk.com/admin/account/security/end%5Fusers?ref=internalnote.com#messaging)
```html
App ID: app_12345abcde1234567890
Secret: some-very-long-string-with-digits-and-numbers
```
To generate a login for the web widget you'll also need a name, email and external ID for your user.
## Generating the JWT
Based on the above items you can generate a working JWT. You can find an example below.
There's a few important caveats:
- External ID has to be unique for each user, so use a GUID, UUID, or database ID
- The credentials do not expire unless you set an expiration date
- We pass `email_verified: true` so that Zendesk will handle the email address as a verified address to work with the new [email identity](https://internalnote.com/messaging-authentication-identify-and-merge-existing-users?utm%5Fsource=demo) rules.
💡
Note that our sample code does ****not** integrate with your actual directory and as such we can not validate if the user you're logging in actually exists. That's something you need to implement yourself!
```javascript
var input = {
"external_id":1906365876753,
"user_email":"john@example.com",
"user_name":"John Smith"
}
const app_id = "app_123";
const secret ="abc123";
const key = await crypto.subtle.importKey(
"raw",
utf8ToUint8Array(secret),
{ name: "HMAC", hash: "SHA-256" },
false,
["sign"]
);
const header = JSON.stringify({ alg:"HS256", typ:"JWT", kid:app_id });
const payload = JSON.stringify({
exp: Math.floor(new Date().getTime() / 1000.0) + 86400,
scope: "user",
name: json.user_name,
email: json.user_email,
external_id: external_id,
email_verified: true
});
const partialToken = `${base64URLStringify(utf8ToUint8Array(header))}.${base64URLStringify(utf8ToUint8Array(payload))}`;
const signature = await crypto.subtle.sign(
"HMAC",
key,
utf8ToUint8Array(partialToken)
);
const jwt = `${partialToken}.${base64URLStringify(new Uint8Array(signature))}`;
```
For the example website above I created a Cloudflare Worker that handles the JWT scenario for both Messaging and the Classic Widget.
You can find a working code example via the link below.
[GitHub - verschoren/zendesk\_widget: Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand.Zendesk has a nice Web Widget to embed your contact channels and FAQ on any page of your website creating a consistent experience that aligns with your brand. - GitHub - verschoren/zendesk\_widget:…GitHubverschoren](https://github.com/verschoren/zendesk%5Fwidget?ref=internalnote.com)
### Logging In
The end result is a JWT token that you pass to the widget via:
```javascript
zE('messenger', 'loginUser', function (callback) {
callback(jwttoken);
});
```
### Logging Out
If for any reason you wish to logout your user, you can use the code below.
```javascript
zE('messenger', 'logoutUser');
```
## What about Zendesk Guide?
You can also easily use the above example to authenticate Zendesk Guide.
1. Open your Zendesk Guide Theme and [edit its code](https://support.zendesk.com/hc/en-us/articles/4408832558874-Editing-the-code-for-your-live-help-center-theme?ref=internalnote.com)
2. Open `document_head.hdbs` and paste the[ following code](https://demo.internalnote.com/guide%5Fmessaging?ref=internalnote.com) at the bottom
3. From now on, whenever a user logs into your Help Center, the widget will authenticate itself.
### Zendesk Flow Builder API Call
URL: https://internalnote.com/zendesk-flow-builder-api-call/
Last updated: 2025-08-01T12:06:41.000Z
Zendesk announced the ability to add an API call step to their Messaging Flow Builder. This allows you to get external data and show it your users. So what's possible? Let's build a quick mockup and find out!
Zendesk just announced the ability to add an API call step to their Flow Builder. This allows you to *get, post, update or delete* data on a remote server or API as a step in the Bot flow.
As usual, their [documentation](https://support.zendesk.com/hc/en-us/articles/4572971586586?ref=internalnote.com) is very clear, but lacks some real world examples or detailed steps.
# Concept
We have a Support widget on our website. We notice that customers often contact us with questions regarding the status of our systems.
So instead of pointing them to an external status page, we'll let the bot get the current system status, and hopefully return an "All good, we're online". Or if not, we let them know we're offline with a link to more information.
You can test this flow via the Widget below.
# Setup
For this example we'll use [GitHub Status API](https://www.githubstatus.com/api?ref=internalnote.com). It has a Summary endpoint we can use to retrieve the current status of GitHub:
## Endpoint
`GET curl https://www.githubstatus.com/api/v2/status.json`
## Response
```json
{
"page":{
"id":"kctbh9vrtdwd",
"name":"GitHub",
"url":"https://www.githubstatus.com",
"updated_at": "2022-05-30T10:27:36Z"
},
"status": {
"description": "Partial System Outage",
"indicator": "major"
}
}
```
# So how do we set this up?
First we go to our Flow Builder setup and setup some boilerplate.We've created a Bot flow that basically asks the customer if they need to know the current status, or need help.
## Make an API call
Next we add a step underneath the "*Are all your systems online?*" question, and choose an API step.
Enter the following and press *Make API Call*
- Subject: Get System Status
- Description: This retrieves the status of the GitHub systems.
- Request Method: GET
- Endpoint URL: [https://www.githubstatus.com/api/v2/status.json](https://www.githubstatus.com/api/v2/status.json?ref=internalnote.com)
If all goes well the system returns a valid response and a list of variables returned.
For our example, we're only interested in the *status > description* value. Go to this entry via the dropdown and click *Save*. Choose *system\_status* as the variable name.
We've now:
1. Retrieved the current status from GitHub
2. Saved the status description as a variable for later reuse.
Each time your customers open the widget and select "Get System Status", the Bot will pull in the current status and save it for reuse.
## Showing the status
What's left is showing that status to the customer.
- Under the *Success option* select the "*\[This will be the API step\]*" *m*essage and edit the text to say "*Our current status is:* "
- Click the "*Select Variable*" button and choose the recently created *system\_status* option.
- The text should now say "*Our current status is: {{system\_status}}*"
- Similarly, under the *Fail* option, add a text block that handles the scenario where we can't retrieve the status.
# The result
You can test out the flow by going to [support.internalnote.com](https://support.internalnote.com/?ref=internalnote.com).
# Improvements we'd like to see
- The ability to retrieve an array, and use those as values in an options list.Example: list items in a customers' order, and have them select the item before going forward
- Ask for the name of the currently JWT authenticated user and use that to personalise replies
## Ask For Details
This article was written before Zendesk enabled *Ask for Details* in Flow Builder. See this other article if you want to find out about more complex API calls
[Flow Builder - Ask for detailsThe new Ask For Details option in Flow Builder allows you to pull in contextual information via API into your Zendesk Chat Bot.Internal NoteThomas Verschoren](https://internalnote.com/flow-builder-ask-for-details/)
### Designing Relate 2020
URL: https://internalnote.com/designing-relate-2020/
Last updated: 2023-11-28T10:09:38.000Z
What WWDC is for Apple users, Relate events are to Zendesk users. I’ve done a couple in London and they were always a delightful combination of tons information, both technical and commercial, good speakers and a nice calming design.
Pity/understandable that this one was cancelled due to COVID, from the images in the article [linked](https://medium.com/zendesk-creative-blog/i-really-definitely-and-seriously-did-attend-relate-2020-heres-what-happened-b653be9b72c0?ref=internalnote.com) the look and feel alone would have made this event a joy to walk around in.
> But I had to document what this event would’ve been. This article is an ode to the amazing team of Zendesk creatives and marketeers, from 3D to sound and video and slides, as well as our colleagues from [Sparks](https://wearesparks.com/?ref=internalnote.com), who dedicated sweat, time, tears, talent (and, let’s be real, probably a little blood) to create beautiful things that never saw the light of day.
### Zendesk App Tools Github Action
URL: https://internalnote.com/zendesk-app-tools-github-action/
Last updated: 2023-11-28T10:09:09.000Z
This [Github Action](https://github.com/marketplace/actions/zendesk-app-tools-deploy-apps?ref=internalnote.com) creates or updates an existing app in your Zendesk instance after a successful commit in your GitHub repository.
[Zendesk App Tools - Deploy Apps - GitHub MarketplaceCreate or Update an existing app in a Zendesk instance. Requires a .zat fileGitHub](https://github.com/marketplace/actions/zendesk-app-tools-deploy-apps?ref=internalnote.com)
Checkout the YouTube video below to see how it works.
### Zendesk Talk Status
URL: https://internalnote.com/zendesk-talk-status/
Last updated: 2023-11-28T10:08:16.000Z
Zendesk Talk is Zendesk’s integration VOIP solution. We use it at the office for both incoming support calls from our clients as well as, naturally, outgoing calls.
Now weirdly, Zendesk does not offer a mobile app for their phone solution. They have great apps for Support and Sell, but you can’t access or manage your Talk account go.
This leads to a few issues. Once I leave the office I can’t put my account offline, or redirect calls from browser to mobile.
So, I solved this little hassle with a custom web app. It allows anyone with a Zendesk account to change their status or redirect calls to mobile and web.
Needless to say, I track nothing, can’t hear your calls and don’t store any data.
