What Is an SLA? A Practical Guide for Support Teams
What an SLA is and how to define one: first response time, resolution time, priority levels, business calendars and handling breaches.
An SLA (Service Level Agreement) is a written commitment about how quickly your support team will respond and resolve. An unwritten expectation can never be reliably met, because everyone has a different number in their head.
What an SLA defines
First response time
From the ticket being opened to the first meaningful reply. An automatic “we have received your request” does not count.
Resolution time
Time until the ticket is closed. It needs to vary with complexity.
Priority levels
Critical, high, normal, low — each with its own targets.
Business calendar
Do the clocks run 24/7 or only during working hours? If this is unclear, every breach becomes an argument.
Defining priority levels
Priority should be set by impact and urgency, not by how annoyed the customer sounds. A simple matrix is enough:
- Critical: the service is completely down and work has stopped. Example: the phone system is dead and nobody can take calls.
- High: an important function is broken but there is a workaround. Example: reports will not generate but the data can be read on screen.
- Normal: an irritating problem that does not stop work.
- Low: improvement requests, questions, information requests.
Setting realistic targets
- Measure your current performance first. If your median first response is six hours today, writing a one-hour SLA means breaching the promise in week one.
- Target a percentile, not a median. “90% of tickets are answered within four hours” is measurable and defensible.
- Write down the calendar. Does a ticket opened at 17:55 on Friday start counting at 09:00 on Monday? That sentence belongs in the agreement.
- Define pause rules. Time spent waiting for information from the customer should stop the clock; otherwise you are charged for time you do not control.
- Set up escalation. Who gets told before the clock runs out?
Running an SLA: measurement and visibility
An SLA that lives only in a contract is just text. Three things make it real:
- Every request needs a number and a timestamp. Conversations scattered across an inbox cannot be measured.
- The clock must be visible. Agents should see on screen which ticket is about to breach.
- Breach reports must be read regularly. The valuable question is not how many breaches there were but why.
When a breach happens
Breaches will happen; what matters is what follows. Good practice: tell the customer before they notice, explain the reason and give a new estimate. A delay met with silence feels twice as long as it is.
Common mistakes
- Defining a single target. A password reset and an integration failure cannot share a resolution time.
- Counting the auto-reply as the first response. Customers do not experience it as an answer.
- Measuring closures but not reopens. Closing fast is the easiest and most damaging way to make an SLA look good.
- Ignoring capacity. A two-person team cannot promise a 24/7 critical SLA.
Frequently asked questions
Should the SLA be the same for every customer?
Not necessarily. Enterprise contracts often define tiers — but keep the number of variants manageable, or the team will lose track of which target applies to which ticket.
What is the difference between an SLA and an OLA?
An SLA is the promise to the customer; an OLA (Operational Level Agreement) is an internal commitment between teams — for example how fast support can expect an answer from engineering.
Does an SLA have to be in the contract?
No. It can be run as an internal target that appears in no contract; the measurement and improvement benefits are the same.
Next step: to see how ticket management and SLA clocks work, request a demo.