MTTR vs. MTTD: Which Metric Matters More?
MTTR and MTTD measure different parts of incident response. Strong reliability teams track both detection speed and recovery speed.
Detection and recovery are different problems
MTTD, mean time to detect, measures how long it takes to know something is wrong. MTTR, mean time to resolve, measures how long it takes to restore service after the problem begins or is detected.
Both matter because customers experience the whole incident, not just the repair window.
Why MTTD is often overlooked
Many teams focus on recovery while ignoring detection. But if customers report outages before monitoring does, the response is already late. Uptime monitoring, synthetic checks, server monitoring, and alert escalation should reduce the time between failure and awareness.
MTTD is especially important for silent failures: broken checkout, failed cron jobs, expired certificates, DNS mistakes, or degraded APIs that still return partial success.
MTTR shows operational strength
MTTR improves when teams have clear ownership, useful runbooks, rollback paths, good observability, and practiced incident response.
The better question is not which metric matters more. MTTD protects awareness. MTTR protects recovery. Together they show whether your reliability program can detect, understand, and fix customer-impacting problems quickly.