Operations

What is escalation matrix?

Published September 10, 2026 · Reviewed by the Fidiora team

Definition

An escalation matrix is a documented map of who becomes involved in an issue, at what severity, and after how long. It removes judgement from the escalation decision by stating in advance that a severity one issue unresolved after thirty minutes goes to a named role.

Why it matters

Escalation matrices exist for the worst moment, not the average one. At three in the morning during an outage, nobody should be deciding who to wake. A matrix converts a stressful judgement call into reading a table, which is why service desk practice has used them for decades.

What to know

  • Define severity by customer impact, not by technical complexity, or engineering and support will disagree constantly.
  • Name roles rather than people, because people leave and rotas change.
  • Include time thresholds, not just severity levels, or nothing escalates until someone loses patience.
  • Publish it where an agent under pressure can find it in seconds, not in a wiki nobody has opened.
  • Test it. An escalation matrix that has never been exercised is a document, not a process.

An example

Example: severity one, meaning multiple enterprise customers cannot use the product, notifies the on-call engineer immediately and the support lead at fifteen minutes. Severity three notifies nobody outside the queue and is reviewed the next working day.

Common mistakes

  • Defining severity by how hard the fix is rather than by customer impact
  • Naming individuals instead of roles
  • Omitting time thresholds so escalation depends on someone getting annoyed
  • Writing the matrix and never running a drill against it
Where Fidiora fits

Fidiora encodes the matrix as rules rather than as a document. Severity, customer tier, and elapsed time can all trigger the right notification automatically, so escalation happens on time regardless of who is on shift.

See Pricing
FAQ

Questions

What should an escalation matrix contain?
Severity definitions based on customer impact, the role notified at each level, the time thresholds that trigger the next level, and the communication channel for each. Anything more and nobody will read it under pressure.
How many severity levels should we have?
Three or four. More levels sound precise and produce inconsistent classification, because the boundaries between five levels are impossible to apply quickly and consistently.
How often should the matrix be tested?
At least twice a year, and after any change to the on-call rota or team structure. An untested escalation path fails exactly when you most need it.
Get Started

See Fidiora resolve a ticket in 60 seconds.

No credit card, no sales call required. Connect your docs and watch it work.

No credit card · Live in under an hour