Agile Metrics: Velocity is NOT the Goal
summary
Argues velocity is a lagging indicator and setting targets triggers Goodhart's Law and the Hawthorne Effect, so boosting velocity often hides costs (longer hours, lower quality). Demonstrates fixes: apply standard deviation to velocity (example SDs 7 and 3.1) to get confidence ranges, use scatter plots to link low velocity to high cyclomatic complexity and low test coverage, use cumulative flow diagrams to spot approval bottlenecks, and track team mood with a git checkin plugin (joy proves a leading indicator). Uses Pivot Tracker/Jira APIs, burn-with-SD and distribution charts to make predictions actionable.
AI-written summary · about
chapters (11)
- 0:00Introduction and setup
- 1:02What is velocity in Agile
- 8:00Velocity as a lagging indicator
- 14:22Pressure to increase velocity
- 22:10Causes of variable velocity
- 26:02Using scatter diagrams to find root causes
- 36:44Cumulative flow diagrams explained
- 41:20Measuring team health beyond velocity
- 51:22Team Joy and leading indicators
- 53:28Metrics for teams not managers
- 58:23Q&A on defects and velocity
related talks
41mEscape Velocity - Better Metrics for Software Teams
48mVelocity Deep Dive
01h 31mAgility ≠ Speed - Agile Development
37mDebugging Engineering Velocity
43mPredictable Delivery
15mChoosing Quality to Improve Your Team's Velocity
26mMetrics and Modeling – Helping Teams See How to Improve
29mYour Metrics Are Lying: How to Stop Reporting and Start Learning