What a Good MSP SLA Actually Looks Like (Bring This Checklist to Any Provider)

“Fast response” is not a commitment. A real SLA has numbers, verification, and consequences. Here’s the checklist we’d bring to any provider — including us.

Response targets by severity — in writing

Critical (total outage, firewall down): minutes, not hours, 24/7. High (one system down): under an hour. Medium and low: same or next business day. If a provider won’t publish the matrix, that’s the answer.

Backup verification, not backup existence

Ask three questions: Who reads the backup logs, and how often? How often are sync jobs checked? When was the last restore test — and can you see the report? Daily log review, weekly sync checks, and quarterly sandbox restores is the standard we hold ourselves to.

Patching with a calendar

Hypervisor and server patching should be monthly and scheduled, firewall firmware updated outside business hours, and licensing compliance checked on a cadence — not “when we get to it.”

Scope honesty

Unlimited helpdesk should say what it covers (OS issues, malware removal, local software) and what triggers a project quote instead. Vague scope becomes a dispute at the worst possible moment.

Compare this checklist against our own published Support & SLA — we put the numbers on a public page on purpose.