Learning From Failure: How We Learned a Better Way
summary
Claims that Five Whys and single-cause RCAs stop learning and that treating "human error" as the end of an investigation hides system faults. Adopting the Etsy debriefing method (a 30-page guide on GitHub) introduces a facilitator, rejects counterfactuals, and focuses on shared learning. After a six-hour outage an original 30-minute RCA expands to 1.5 hours, action items change, and two weeks later a repeat failure lasts three hours and is less severe, showing measurable improvement.
AI-written summary · about
related talks
29mLearning from incidents: how things went right
58mStop Trying to Fix Everything
05mIgnite: Root Cause Analysis Fact and Fiction
22mChoosing Failure
30mHow Did Things Go Right? Learning More from Incidents
40mBuilding Resilience: How to Learn More from Incidents
15mPragmatic Incident Response: Lessons Learned from Failures
16mCreating a Learning Culture