How Does AI Ticket Routing Work in Zendesk? Triage, Assign and Classify (2026)

AI ticket routing in Zendesk works by having an AI agent read each new ticket on creation, set its type, priority, tags and custom fields, then assign it to the right group or agent. You can do it with Zendesk's own intelligent triage (included on Suite and Support Professional and above, but using its predictions in workflows needs the Copilot add-on at $50 per agent per month) or with an agent layer like Macha that writes the fields and makes the assignment itself.

The cost of bad routing is easy to miss because it doesn't show up in first-response time until it's too late. A ticket sits in the general queue before someone notices it belongs in billing. It's labeled "question" when it's really an "incident." It sits unassigned because an assignment script broke. The customer doesn't see that as your team being busy. They see nobody answering, and the problem grows with the team.

What does ticket routing actually involve?

Routing is three decisions:

1. Classification: what kind of ticket is this? Question, incident, problem or task? Which product area, language and customer tier?
2. Prioritization: how urgent? VIP customers and outages get high priority. "How do I cancel" usually doesn't.
3. Assignment: who or which queue should own it? Billing, technical, account managers, or a specific on-call rep.

A good routing system makes all three. Most teams' rules cover one or two and leave the rest to whoever reads the queue first, which is where the delay comes from.

What can an AI agent do that Zendesk triggers can't?

Zendesk triggers are rule-based: "if subject contains 'refund', tag as refund-request." That handles tickets where the customer uses your keyword. It misses the ones where they don't, where the request is ambiguous, or where it spans two categories.

Zendesk's own answer is intelligent triage, which classifies every incoming ticket by topic, sentiment, language and entities. Zendesk rolled the classifications out to Suite and Support Professional plans and above in July 2026 at no extra cost, but acting on them in triggers and other workflows requires the Copilot add-on, $50 per agent per month. That pricing is per seat, so the bill tracks the size of your team rather than the number of tickets it routes. The predictions still feed triggers you write, so the routing logic stays yours to maintain.

An AI agent reads the ticket the way a person would and acts on it directly. It understands that "I want my money back for the order I never received" is both a refund request and a shipping problem. It notices a frustrated customer and bumps priority. It sees a Spanish ticket and assigns it to the Spanish-speaking team. It can check the customer's history (a VIP, a fourth ticket this month) before deciding.

The Zendesk tools that make this work in Macha:

• Update Type: set the ticket as problem, incident, question or task
• Update Priority: low, normal, high or urgent
• Update Tags: add or remove category and topic tags without overwriting existing ones
• Update Fields: set custom field values, with dropdown options checked and dates formatted
• List Groups: list every active Zendesk group so the agent can resolve a group name to an ID
• Assign Ticket: assign to a group, a specific agent, or both

The agent uses these together. A typical autonomous routing flow: read the ticket, classify and set type, set priority, set custom fields and tags, look up the right group, assign.

Which routing workflows are worth automating first?

1. Tier-based routing for B2B support

Goal: route enterprise tickets to the dedicated account team and everyone else to general support.

How the agent does it:

1. Reads the ticket and identifies the requester
2. Looks up the customer in your CRM via a custom API tool
3. If tier is "Enterprise": tags tier-enterprise, sets priority to High, and assigns to the Enterprise Support group
4. Otherwise: assigns to General Support

The flow runs as soon as the ticket is created, so enterprise tickets land with the right team before anyone opens the general queue.

2. Language-based routing

Goal: German tickets go to the DACH team, French to the FR team, English to the global team.

How the agent does it:

1. Reads the subject and body and detects the language
2. Sets the language custom field
3. Calls List Groups to find the right team (for example, "Support DACH")
4. Assigns the ticket to that group

This is where custom fields with labels matter. The agent reads and writes field values by their human-readable labels, so the routing logic reads "set language to German" rather than "set field_1900003089273 to 1900012345."

3. Incident vs question disambiguation

Goal: outage reports go to the on-call engineer immediately; how-to questions go to the general queue at normal priority.

How the agent does it:

1. Reads the ticket and decides: is this a report of something broken (incident) or a question about using the product?
2. If incident: sets type to "Incident", priority to "Urgent", assigns to Engineering On-call, and posts a Slack alert through the Slack connector
3. If question: sets type to "Question", priority to "Normal", and lets the AI try to answer from the Help Center before a human picks it up

Keyword rules struggle here. A model can tell "the page won't load" (incident) from "how do I change my password" (question) even though both might mention pages, loading or passwords.

How do you set up a routing agent in Zendesk?

If you've already done basic agent setup (see the complete setup walkthrough), it takes five steps:

1. Create a dedicated routing agent, separate from your customer-reply agent. Its instructions cover classification and routing, not customer-facing replies.
2. Enable the routing tools: Update Type, Update Priority, Update Tags, Update Fields, List Groups, Assign Ticket.
3. Write the routing instructions in plain language: "Classify the ticket as incident, problem, question or task. Set priority based on customer tier and ticket sentiment. For language-specific tickets, assign to the matching team."
4. Set the trigger to Ticket Created, so routing happens before any customer-facing reply.
5. Test before going live. Use Test Run on a sample of about 20 historical tickets and check the routing decisions against what your team would have done. Tune the instructions for every miss.

You can also chain agents. The routing agent assigns the ticket to a group, and if that group has its own reply agent, that agent picks up where routing left off. Two narrow agents are easier to test than one that does everything.

How do you know the routing agent is working after 30 days?

Check these at the end of the first month:

• Tickets land in the right queue on creation. Compare time-to-correct-queue before and after.
• Priority levels mean something. Urgent tickets are urgent because the agent checked the signal, not because customers picked the option.
• The misroute rate is low and falling. Hand-check a sample of routed tickets each week. Misses should be edge cases worth an instruction change, not a repeating pattern.
• Custom fields stay populated. Reports get accurate because the data is consistent, not because someone remembered to fill in the form.
• Humans handle exceptions, not classification.

Is AI routing worth setting up if you already have triggers?

Routing sounds dull until you've run a month with it. When every ticket starts in the right place at the right priority with its fields filled in, the queue stops being a mix of urgent and trivial and arrives sorted and assigned.

Macha fits teams already running Zendesk, Freshdesk, Gorgias, Front, HubSpot or Intercom who want the agent writing fields and assignments inside the ticket, and it's billed per ticket rather than per seat, from $299 a month for up to 750 tickets on the pricing page. Combined with autonomous reply agents handling tier 1 traffic, the team's work becomes the hard tickets, escalated to the right person, in the right order.

To set this up, start with the complete setup walkthrough, then add a routing agent as a second specialist. Or see Macha for Zendesk for the full picture.