Every few months I am shown a service level report where everything is green, and then introduced to a customer who is thoroughly unhappy. Both are telling the truth. The agreement is simply measuring something other than the thing that matters.
The usual culprit is a response-time target that stops the clock when a ticket is acknowledged rather than when the customer can work again. The desk optimises for the measure, because people always do. Acknowledgement gets fast; resolution does not move.
Start from the business outcome
Rewriting an SLA properly means going back to the line of business and asking a question that sounds naive: what does a bad day look like for you? The answers are rarely expressed in minutes. They are expressed in orders that could not be taken, invoices that could not be raised, a branch that could not open.
Translate those into a small number of service definitions, attach targets that reflect the consequence rather than the convenience, and only then work backwards into the operational level agreements that your internal teams have to sign.
The internal half is the hard half
An SLA the business believes in is worthless if no internal team has agreed to the underlying OLAs. This is where most rewrites quietly fail — the customer-facing document gets signed, the internal ones do not, and within two quarters the report is green again for the old reasons.
Do the unglamorous half. It is the half that holds.