Sign in Book a demo
Glossary
Glossary Escalation & handoff Updated Oct 9, 2026

What Is Ticket Escalation? (And How to Build a Support Escalation Process)

Ticket escalation is moving a support ticket to someone with more expertise, authority, or system access when whoever is handling it, a person or an AI agent, can't resolve it. Good escalations carry the full context along.

Also called ticket escalationsupport escalation process

In short
  • Ticket escalation moves an issue to someone with more expertise, authority, or access.
  • Common types are functional, hierarchical, automatic, and AI-to-human.
  • A good escalation passes the full context, so customers never repeat themselves.
  • McKinsey: almost 70% of care leaders say empathy and trust will always need humans.

In a McKinsey survey of 440 customer care leaders, almost 70% agreed that empathy and trust will always require human involvement. Even among the most advanced teams, 64% named customer preference for a live person as a top barrier to digital adoption.

Forrester's 2024 CX Index names underwhelming chatbot experiences as one reason US customer experience quality hit an all-time low.

At Vrio, Brainfish now resolves 80 to 85% of developer questions automatically. Engineers who used to field 30 to 50 messages a day now see only the complex edge cases.

What are the types of ticket escalation?

Escalations usually fall into one of four types, depending on what triggers them.

Type What triggers it Where the ticket goes Example
Functional The issue needs skills or system access the current tier doesn't have A specialist team, like billing or engineering A bug that needs server logs pulled
Hierarchical The issue needs a decision above the handler's authority A team lead or manager A refund outside standard policy
Automatic A rule fires, like an SLA timer about to run out The next tier or an on-call owner An urgent ticket untouched for two hours
AI-to-human The AI can't answer confidently, hits a policy limit, or the customer asks for a person A human agent, with the conversation attached A customer asks to discuss a contract change

How do support escalation tiers work?

Many teams organize escalation into tiers, each handling harder problems than the last.

  • Tier 0: self-service. Help center, in-app help, and AI agents.
  • Tier 1: frontline support. Generalists who handle how-to questions and known issues.
  • Tier 2: technical specialists. Deeper product knowledge, plus access to admin tools and logs.
  • Tier 3: engineering or product. The people who can fix the product itself.

What does a good support escalation process look like?

A support escalation process decides when a ticket moves up, where it goes, and what travels with it. Strong ones do six things well.

  1. Name an owner for every tier, not just a shared inbox.
  2. Write clear triggers so agents don't have to guess.
  3. Pass the full context: the conversation, what's been tried, account details, and any logs.
  4. Set response times by tier and priority.
  5. Keep the customer informed on who has the ticket and what happens next.
  6. Learn from every escalation by grouping the reasons and fixing the gaps behind them.

McKinsey's research on agentic customer experience recommends involving frontline teams from day one to define escalation thresholds and handoffs between AI agents and people.

When should a ticket be escalated?

Escalate when a ticket:

  • Needs skills or access the current tier doesn't have
  • Needs a policy exception, like a refund or contract change
  • Is about to miss its SLA
  • Involves a high-value or at-risk account
  • Touches security, legal, or compliance
  • Comes from a customer who asks for a person, which should work on the first ask

How do you measure ticket escalation?

The core metric is escalation rate: escalated tickets divided by total tickets, times 100. If 1,200 of 10,000 monthly tickets move up a tier, the rate is 12%.

A low rate isn't automatically good, since it can also mean customers are getting stuck. Track it alongside:

  • Escalation reasons, to spot recurring gaps
  • Time to escalate, or how long customers wait to reach the right person
  • Resolution time after escalation
  • Bounce-backs, or tickets sent back down because they didn't need escalating

How is AI changing ticket escalation?

As AI agents take on more Tier 0 and Tier 1 work, the question becomes when the AI should hand off. There are two ways to get it wrong. An agent working from thin or outdated knowledge escalates far too much, while one built to deflect at all costs traps customers in a loop.

Cybersecurity company Huntress set out to avoid the second mistake. Its support leader had been on the receiving end of bots that keep stalling people who ask for a human, and the team refused to ship that experience. As he puts it:

"Deflection is when you don't want people to come to support. Prevention is when you don't need them to but if they do, by all means, come to us." Adam Wasmus, VP of Global Support, Huntress

The goal is to answer well enough that most customers never need a ticket, and let the ones who do reach a person easily.

Brainfish builds AI support agents for B2B teams that need accurate answers. They ground every answer in your product knowledge with a cited source and follow the escalation rules you set, including routing high-value accounts straight to a person. They run alongside the helpdesk you already use, like Zendesk or Intercom.

For complex tickets, Ballet by Brainfish turns a plain-English description of a process into a workflow that checks your systems, applies your policy, takes the action, and escalates when it should. Teams can also start from prebuilt templates, like one for log investigation that pulls the logs and attaches a first-pass diagnosis before an engineer opens the case.

Want to see which of your escalated tickets an AI agent could resolve? Bring your trickiest ones to a Brainfish demo, or check pricing.

Frequently asked questions.

What's the difference between ticket escalation and ticket routing?

Routing sends a new ticket to the right queue when it first comes in, based on topic, language, or plan. Escalation happens later, after someone has already worked the ticket and found they can't resolve it. Better routing usually means fewer escalations.

What is a good escalation rate?

There's no single benchmark, since it depends on product complexity and how much your frontline tier can do on its own. A falling rate paired with steady CSAT usually means real progress. A falling rate with rising complaints often means customers are getting stuck.

Should AI agents escalate tickets to humans?

Yes, whenever the AI can't give a confident, sourced answer, the request falls outside policy, or the customer asks for a person. The aim is to escalate the right tickets, with full context attached, so the human can pick up exactly where the AI left off without starting over.

How does Brainfish handle ticket escalation?

Brainfish AI support agents answer from your real product knowledge with a source behind every answer, and follow the escalation rules your team configures. When a person is needed, workflows built with Ballet by Brainfish can check systems, pull logs, and attach a diagnosis first, so the human starts with the investigation already done.

See it working on your own tickets.

Bring a real ticket to the demo and we'll show you Brainfish resolving it, grounded in your docs and systems.

Book a demo

No engineering sprint required.