Set budgets per journey
Define p95 and p99 latency objectives for signup, checkout, and API calls that map to revenue.
Track error budgets for latency the same way you do availability.
Slow is down—latency burn should page the same way availability burn does.
Alert on burn rates
Configure Watch.Dog to alert on fast and slow burn for latency SLOs so you see both spikes and trends.
Route latency alerts to owners with runbook links for quick tuning.
Prioritize fixes visibly
Add latency budget usage to team dashboards and sprint planning to keep performance work funded.
Use feature flags to ship gradual performance improvements without risking regressions.