Stale alert thresholds turn your regression pipeline into noise. When baselines drift — new features, added dependencies, heavier pages — thresholds set at launch either fire constantly and train the team to ignore them, or sit so loose that real regressions slip past unseen. Both outcomes erode the signal that observability tooling is supposed to provide, and users pay the cost in LCP, TBT, and abandoned sessions the team never hears about.
Info because stale thresholds degrade monitoring quality rather than directly breaking user-facing performance or availability.
Put a recurring quarterly review on the calendar and record each pass in a tracked doc so the cadence survives team turnover. Capture the current baseline for each Core Web Vital, the threshold, and the date reviewed. Use a file like docs/performance-baselines.md:
## Q1 2026 Review (2026-03-15)
- LCP: baseline 1.8s, threshold 2.5s
- INP: baseline 180ms, threshold 200ms
- CLS: baseline 0.05, threshold 0.1
ID: performance-deep-dive.regression-prevention.alert-thresholds-reviewed
Severity: info
What to look for: Count all documented performance baseline measurements. Enumerate evidence of quarterly reviews in git history or docs. Look for documentation on alert threshold review schedule. Check git history or team docs for evidence of quarterly reviews. Look for baseline measurements and trend data.
Pass criteria: Alert thresholds reviewed and updated quarterly. Thresholds are based on business requirements and user expectations. Baselines adjusted as product grows. At least 1 documented review within the last 90 days.
Fail criteria: Alert thresholds never reviewed, or outdated relative to current performance baseline.
Skip (N/A) when: Never — thresholds should evolve as product grows.
Cross-reference: For alert configuration, see regression-alerts-configured.
Detail on fail: "Alert thresholds set in 2023 — app now 5x slower, thresholds outdated" or "No evidence of threshold review since initial setup"
Remediation: Schedule quarterly reviews of performance targets and alert thresholds.
<!-- docs/performance-baselines.md — quarterly review -->
## Q1 2026 Review
- LCP baseline: 1.8s (threshold: 2.5s)
- Reviewed: 2026-03-15