Technology
SLA. What needs to happen after an outage?
Starting point
Define where availability data comes from and when the clock starts. Separate response time, workaround and restoration. Test the measurement against an incident starting on a Friday evening.
What to agree in practice
Beyond service-level credits, agree communications during the incident and what the post-incident report should contain. For recurring failures, define how to agree a root-cause fix and decide whether to continue the relationship.