A service level agreement, or SLA, sets out how quickly an IT provider will respond and what happens when it doesn't. A vague one protects the provider. A clear one tells both sides what to expect.
Priorities, defined in business terms
The SLA should define priority levels by impact: the whole business down, one team affected, one person affected, a request for something new. Check that you agree with the definitions, because the targets hang off them.
Response is different from resolution
A response target is how fast someone starts working on the issue. A resolution target is how fast it is fixed. Many agreements only promise the first. Ask for both, at least for the high priorities.
Hours of cover
Check when the clock runs. A four-hour response during business hours is a different promise to four hours at any time. If you work weekends or across time zones, the hours need to match.
What is excluded
Read the exclusions: third-party outages, unsupported hardware and software, problems caused by changes the provider didn't make. Some are fair. A long list can leave very little covered.
Escalation
The agreement should say what happens when a ticket isn't progressing: who it goes to and when. You should have a named contact above the service desk.
Reporting
You can't hold anyone to targets you can't see. Monthly reporting should show performance against each target, with the misses explained.
What happens on a miss
Service credits are common, but they are small. More useful is a commitment to review repeated misses and a right to leave if performance stays below target.
Your side of it
SLAs put obligations on the customer too: reporting issues through the agreed channel, giving access, approving changes. Know what yours are.
Want your IT looked after for a fixed monthly fee? Managed IT services
All resources