Measuring Developer Productivity: the good, the bad and the data
summary
Argues that productivity must measure outcomes not raw outputs, warns that individual metrics invite gaming and kill collaboration, and recommends team-level DORA/SPACE-style measures plus qualitative surveys. Claims shorter code-review lead times correlate with 50% better delivery performance, documentation included in stories yields a 12-fold increase in performance, and trunk-based development speeds flow. Advises generative culture, flexible infrastructure, Copilot-style tooling as quick wins, blameless postmortems, and a three-step “converse / coach / cut” process for persistent low performers.
AI-written summary · about
chapters (8)
- 0:00Introduction and setup
- 1:27What is developer productivity
- 3:39Good practices for measuring productivity
- 8:27Bad practices and vanity metrics
- 13:08Combining qualitative and quantitative metrics
- 15:22Data-backed improvements for productivity
- 26:19Handling low performance in teams
- 32:46Summary and closing remarks
related talks
43mMeasuring Developer Productivity: The Good, the Bad and the Data
47mCan We Measure Developer Productivity?
34mBuilding a Culture For Developer Productivity
43mThe Past, Present and Future of Developer Productivity Theories and Metrics
31mHow Do You Measure Devel...
49mProductivity Panel
51mDeveloper productivity is waste
21mHow to Increase Developer Productivity and Happiness