The Debrief Is the Product
Running the exercise is only half the work. The debrief — a structured post-exercise review — is where the value is extracted. Without a debrief, a GameDay is just a stressful afternoon. With a good debrief, it becomes a prioritized list of improvements.
A structured debrief follows the same format as a blameless post-mortem:
- Timeline reconstruction: Walk through what happened, in order, without blame.
- Findings: What gaps were exposed? Categorize them: access issues, knowledge gaps, process failures, tooling failures.
- Prioritization: Not all findings are equal. A missing SSH key that blocked incident resolution is P1. A slightly confusing runbook step is P3.
- Action items: Each finding gets an owner and a due date. "We need a shared secrets vault" → assigned to the Platform team, due in two sprints.
Tracking Maturity Over Time
Organizations that run resilience testing regularly track their exercise maturity — how much better they get over successive exercises. Early exercises surface basic access and knowledge gaps. Mature organizations find subtler issues: communication latency between teams, unclear escalation thresholds, or runbooks that are technically correct but too slow to execute under pressure.
A useful metric is Mean Time to Correct Action (MTTCA) during exercises — how long does it take the team to take the first right step after an injection? Tracking this across exercises shows whether training is working.
Key Point: The goal of resilience testing is not to pass the exercise — it is to fail safely. An exercise where everything goes perfectly is either a sign that your team is genuinely excellent, or a sign that the injection wasn't hard enough. Experienced facilitators deliberately design exercises that the team will partially fail, because partial failure produces the most actionable findings.
A common SRE interview question is: "How do you measure the effectiveness of your on-call training?" Most candidates describe documentation or shadowing programs. A stronger answer introduces measurable outcomes from exercises: "We run quarterly GameDays and track Mean Time to Correct Action across exercises. We also track the ratio of findings that are access/permission issues versus knowledge gaps — that tells us whether our tooling or our documentation is the bottleneck."