A support matrix is only useful when every priority has a shared business meaning, a named owner and a measurable clock. The strongest models connect severity to impact, then keep one team accountable from intake through resolution.
Start with business impact
Define P1 as a critical service outage, P2 as major degradation, P3 as limited impact and P4 as a request or query. This gives operations, engineering and business stakeholders one language during pressure.
Separate response from resolution
A 15-minute P1 response is not a four-hour resolution promise. Track both measures, communicate them clearly and ensure the escalation path reaches senior platform and SRE practitioners without a ticket handoff loop.
Make evidence part of the service
Incident timelines, root-cause analysis, change records and monthly KPI reviews turn an SLA from a contractual number into an operating improvement system.