Metrics that Matter
summary
Argues that metrics reduce subjectivity and create predictability, claiming velocity stabilizes after three sprints for predictability (example: ~23 points) and that a DevOps brainstorming session produced 150 unique metrics that resist simple taxonomy. Recommends choosing metrics from business strategy, prioritizing get data for free and automating collection, exposing metrics publicly to drive improvement, tracking lead/process/wait time, and insists metrics must be comparative over time rather than as isolated numbers.
AI-written summary · about
chapters (8)
- 0:00Start
- 1:38Symptoms Leading to Measurement
- 11:49A Categorization of choices
- 18:26Lead Time = Process Time + Wait Time
- 21:00A Lifecycle of Metrics that Mater
- 21:20Perspective
- 22:00Principles of Measurement
- 25:31Takeaways
related talks
01h 01mFinding Metrics that matter and using them safely
47mWhat's Measured Improves: Metrics that Matter
06mDelivery Metrics: the good, the bad, and the totally ridiculous
39mTest Metrics – a thread or a threat
34mThe Road More Traveled: Putting Transformation Back on the Rails
15mDelivering DevOps Business Metrics that Matter
52mMetrics Serving the Agile Performance of Your Transformation
30mMetrics for good, not evil