Zendesk automation to close stale tickets after multi step reminders: The definitive guide

Stevia Putri
Written by

Stevia Putri

Amogh Sarda
Reviewed by

Amogh Sarda

Last edited October 29, 2025

Expert Verified

Let’s be honest, every support team knows the feeling. You open your queue, and there they are: stale tickets. Those unresponsive conversations that just sit there, gathering dust. They’re more than just a messy backlog; they mess with your resolution metrics and have your agents wasting time on manual follow-ups that go nowhere.

You could have your team chase down every single one, but who has time for that? It’s just not a long-term solution.

The good news? You can build a workflow to handle this automatically. The not-so-good news? Trying to piece it together with Zendesk’s native tools can feel a bit like building with duct tape and hope.

In this guide, I'll walk you through how to set up a Zendesk automation to close stale tickets after multi step reminders. We'll also look at why this native approach has its limits and show you a simpler, AI-powered way to solve the problem for good.

What is Zendesk automation?

Before we jump into the setup, it’s helpful to get a handle on Zendesk's two main automation tools: Triggers and Automations. They sound almost the same, but they do very different jobs.

  • Triggers are based on events. They fire off the second something specific happens, like a customer creating a new ticket. Think of the instant "We got your request!" email you receive, that's a trigger at work.

  • Automations are based on time. They run on a schedule, usually about once an hour, and scan your entire ticket queue. They’re looking for tickets that meet time-based rules, like "this ticket has been pending for more than 48 hours."

To build our multi-step reminder workflow, we're going to lean heavily on time-based Automations to nudge customers and, eventually, close out those inactive tickets.

Closing stale tickets with native Zendesk automation

Getting this done in Zendesk isn’t a one-and-done thing. It’s more like a chain reaction of separate automations that need to work together perfectly. The goal is simple: remind the customer once, give them a final warning after a bit more time, and then close the ticket if you still hear nothing back.

Here’s how to build it, step by step.

Step 1: Create a tag and set the first reminder

First off, we need a way to flag tickets that are in this follow-up process. A simple tag is perfect for this. Let's create one called "stale_ticket_reminder_1".

Now, let's build the first automation:

  1. Head over to Admin Center > Objects and rules > Business rules > Automations.

  2. Click Add automation.

  3. Give it a clear title, something like "Stale Ticket - First Reminder".

  4. Under Meet ALL of the following conditions, you’ll want to add these rules:

    • Ticket: Status | Less than | Solved
    • Ticket: Hours since pending | Greater than | 48 (or whatever time frame works for you, maybe 72 hours)
    • Ticket: Tags | Contains none of the following | "stale_ticket_reminder_1"
  5. Under Perform these actions, set up the following:

    • Notifications: Email user | (requester) and write your first friendly follow-up message.
    • Ticket: Add tags | "stale_ticket_reminder_1" This automation keeps an eye out for any pending ticket that’s over 48 hours old and hasn't already gotten a reminder. Once it finds one, it sends the email and adds our tag so the customer doesn't get pinged with the same message over and over.

Step 2: Set the final reminder

Okay, on to the next one. We need a second automation for that final warning. This one will look for tickets that have already received the first reminder but have been quiet for a few more days.

Let’s create a new tag: "stale_ticket_reminder_2".

  1. Go back to the Automations page and click Add automation.

  2. Title it "Stale Ticket - Final Reminder".

  3. Under Meet ALL of the following conditions, add:

    • Ticket: Status | Less than | Solved
    • Ticket: Hours since pending | Greater than | 120 (so, 5 days total, for example)
    • Ticket: Tags | Contains at least one of the following | "stale_ticket_reminder_1"
    • Ticket: Tags | Contains none of the following | "stale_ticket_reminder_2"
  4. Under Perform these actions, add:

    • Notifications: Email user | (requester) with a message that gives a clear heads-up, like, "Just a note that we'll be closing this ticket in 48 hours if we don't hear back."
    • Ticket: Add tags | "stale_ticket_reminder_2" This makes sure only tickets that have gone through step one and waited a bit longer will get that final nudge.

Step 3: Automatically solve the ticket

Last but not least, we need an automation to close the loop. This one finds tickets that have received both reminders and have hit the final deadline.

  1. Create one more automation and call it "Stale Ticket - Auto-Solve".

  2. Under Meet ALL of the following conditions, add:

    • Ticket: Status | Less than | Solved
    • Ticket: Hours since pending | Greater than | 168 (so, 7 days total)
    • Ticket: Tags | Contains at least one of the following | "stale_ticket_reminder_2"
  3. Under Perform these actions, just add one thing:

    • Ticket: Status | Solved And there you have it. It works, but it feels a bit clunky, right? You're now managing three separate automations, a couple of tags, and some very rigid time-based rules that don't know anything about the actual conversation.

The hidden costs and limitations of native Zendesk automation

While the process we just walked through gets the job done, it highlights some of the headaches that come with relying only on native Zendesk automation.

  • It’s complicated and easy to break. You’re juggling three automations and a tagging system for just one workflow. If someone on your team accidentally removes a tag or tweaks a condition, the whole thing can fall apart. That’s a lot of admin work that only grows as you add more workflows.

  • It’s not smart. The system is basically a glorified stopwatch. It has no clue if the customer's last message was "Thanks, that fixed it!" or "I'm still stuck, just trying to find the error log." It treats both exactly the same, which can lead to closing tickets too early and leaving customers frustrated.

  • It doesn't give you any insights. This workflow won't tell you why your tickets are going stale. Is there a gap in your knowledge base? Is a certain product question consistently tripping people up? Native automations can’t give you that kind of feedback.

  • It’s inflexible. What if you want to change the follow-up time from 48 hours to 55? Or maybe you only want this to run for non-VIP customers? Every little change means digging back into the automation rules, which adds more work and more chances for something to go wrong.

This whole setup solves the immediate problem of closing a ticket, but it does nothing to fix the root cause or help you improve your support operations.

A better way: Using AI to manage stale tickets intelligently

Instead of building a complicated, rigid system inside Zendesk, you could use an AI layer like eesel AI that plugs right into your helpdesk. This approach makes the whole process way simpler and adds a layer of intelligence that native tools just don't have.

With an AI agent, the workflow becomes one smart rule instead of three disconnected ones. The AI can actually understand the context of a conversation, take custom actions, and even learn from your past tickets to give better answers.

Here’s a quick look at how the two approaches stack up:

FeatureNative Zendesk Automationeesel AI Agent
Setup ComplexityHigh: Needs 3+ automations and multiple tags.Low: A one-click integration and a single, flexible rule.
IntelligenceNone: It's all based on timers.Context-aware: Understands what a conversation is about.
FlexibilityLow: Stuck with rigid time-based conditions.High: You can customize prompts, actions, and when it runs.
TestingLive only: You have to turn it on to test it, which can lead to mistakes.Safe Simulation: Test on thousands of your old tickets first.
InsightsNone: Doesn't tell you where your knowledge gaps are.Actionable: Reporting shows you trends and what docs are missing.

Go live in minutes, not months

Unlike tools that want you to tear out your whole helpdesk and start over, eesel AI integrates with just one click. Because it's radically self-serve, you can connect your Zendesk account, train the AI on your knowledge sources (including your past tickets), and get your stale ticket workflow running in minutes. No need to sit through a sales call or a mandatory demo.

Simulate and test your workflow with confidence

Nervous about an AI closing tickets when it shouldn't? eesel AI's simulation mode lets you test your setup on thousands of your own historical tickets. You can see exactly how the AI would have responded, what it would have done, and what your automation rate would look like. This lets you tweak everything with total confidence before it ever interacts with a real customer.

Comparing pricing: Native Zendesk automation vs. an integrated AI solution

When you’re looking at costs, it’s easy to get focused on the price tag. But it’s just as important to think about what you’re actually getting and whether your bill will be predictable.

Zendesk pricing breakdown

The automation features we talked about are available on most Zendesk Suite plans, which start at $55 per agent/month (billed annually) for their Team plan. But if you want more advanced features, like custom workflows for different brands or better analytics, you'll have to jump up to higher tiers like Suite Growth ($89/agent/month) or Suite Professional ($115/agent/month). Those costs go up with every agent you add to your team.

eesel AI's transparent pricing model

On the other hand, platforms like eesel AI offer pricing that's a lot more predictable. Plans are based on a flat monthly fee tied to how many AI interactions you have, not how many agents are on your team. This means your costs don't automatically jump just because you hired someone new.

Best of all, eesel AI has no per-resolution fees. It doesn’t matter if the AI handles 100 tickets or 1,000 in a month, your bill stays the same. This helps you avoid the bill shock that’s common with other AI tools, where a busy month can lead to a surprisingly big invoice.

Stop chasing stale tickets, start automating intelligently

Dealing with a backlog of stale tickets is a headache for pretty much every support team. And while Zendesk’s built-in tools give you a way to automate the cleanup, it's a clunky and high-maintenance solution. Building a multi-step reminder system means juggling tags and automations that can't understand the context of a conversation, creating more work than it saves.

A modern, AI-powered approach is a much better way to go. By integrating a tool like eesel AI with your helpdesk, you can swap that tangled web of rules for a single, smart workflow.

With the ability to learn from your past tickets, understand context, and test everything in a risk-free simulation, you can finally move beyond simple timers. You can build a system that not only clears your queue but helps you understand why tickets go stale in the first place, giving you the power to actually scale your support.

Ready to see how an AI agent can transform your Zendesk workflows?

Start your free eesel AI trial today and go live in minutes.

Frequently asked questions

Setting this up natively in Zendesk is quite complicated, requiring you to manage three separate automations and multiple tags. This multi-step process can be fragile and prone to breaking if conditions or tags are accidentally altered.

The main advantage is that it helps clear your queue of unresponsive conversations, reducing backlog and improving resolution metrics. It also frees up agents from time-consuming manual follow-ups on inactive tickets.

Yes, you can customize the "Hours since pending" conditions for each step to define the timing of reminders and the auto-closure. You can also tailor the email messages sent to the requester in each reminder automation.

If a customer replies, the ticket's "Hours since pending" timer would reset, and its status would likely change from pending. This would typically prevent subsequent automations in the Zendesk automation to close stale tickets after multi step reminders chain from firing, effectively stopping the auto-closure process.

Native automations lack intelligence, treating all inactive tickets the same regardless of context, and offer no insights into why tickets go stale. They are also inflexible, requiring manual adjustments for every minor change, and can't be easily tested without going live.

With native Zendesk tools, testing is primarily live, meaning you have to activate the automations to see how they perform, which can lead to unintended consequences. In contrast, AI solutions often offer a safe simulation mode to test on historical data.

The automation features required to build a Zendesk automation to close stale tickets after multi step reminders are generally available on most Zendesk Suite plans. This includes plans starting from Suite Team, which is their entry-level Suite tier.

Share this post

Stevia undefined

Article by

Stevia Putri

Stevia Putri is a marketing generalist at eesel AI, where she helps turn powerful AI tools into stories that resonate. She’s driven by curiosity, clarity, and the human side of technology.