Product team Performance
Reliability and Speed (Observability SLOs)
Availability
Target:
99.9%
Actual: Nov'25
TBA%
Objective: The API gateway is up 99.9% of the time each month which translates to, at most, 44 minutes of unplanned downtime per month.
Method: To measure this we will measure uptime from synthetic checks every 1 minute from multiple regions.
Speed P95 Latency
Target:
250 ms
Actual: Nov'25
2 mins
Objective: 95% of requests finish in under 250 ms (P95 latency) meaning almost all user clicks feel instant and only the slowest 5% may take longer.
Method: To measure this we will look at latency and errors from request traces and logs at the gateway (and per domain
service).
Error Budget
Target:
39 mins
Actual: Nov'25
300 mins
Objective: We “spend” at most 0.1% of the month on errors/outages. If reliability slips past this allowance then we would pause new launches and fix stability first. The goal being to create customer and team trust, reduce churn, and avoid support blowups during demos or new feature releases.
Method: Manual recording initially
Cost Guardrails (Cost KPIs)
Cloud spend vs. budget
Target:
TBA
Actual: Nov'25
TBA
Objective: Total monthly cloud spend vs. budget (by environment and by product/brand) – simple measure of stay within 100% of budget
Monthly` Spend
Target:
$18,300
Actual: Dec'25
TBA
Objective: Keep team costs within budget.