Measure it or it didnt happen
summary
Moves from meaningless counts (tests run vs failed, bugs opened) to a clear quality goal: release with confidence, measured by a developer confidence poll on a 1–6 scale. Finds >70% of 4,000 test cases never executed, tracks percent flaky unit tests, fixes-in-last-three-betas, regressions and coverage of end-to-end business scenarios, and builds an information radiator dashboard. Takes actions—disable then stabilize flaky tests, add focused e2e automation, test earlier—and shows those metrics and behaviors reliably increase confidence.
AI-written summary · about
chapters (10)
- 0:00Start
- 2:28THE QUALITY GOAL
- 7:47Vanity metrics
- 8:32ACTIONABLE METRICS
- 15:14VISIBLE AND TRUSTED METRICS
- 17:41Unite forces
- 19:00Information Radiator
- 21:15No cheating
- 23:02QUESTIONS
- 29:49CONCLUSION
related talks
43mTo Measure or Not to Measure? That Is the Question
33mDevs should love your tests
24mThe more we test, the worse?
41mThe Tester’s Guide to Quality
46mThe state of the Art in Tackling Flaky Tests
56mThe End Game
29mYour Metrics Are Lying: How to Stop Reporting and Start Learning
48mTesting is Confidence - A Developer's Perspective