Comparing Velocity Across Teams - Why It's Dangerous
Our CTO wants to compare velocity across teams to identify "high performers". Team A averages 60 points/sprint, Team B averages 35. But the teams estimate completely differently. How do you explain why velocity comparison is meaningless?
Use the currency analogy: Story points are like currencies - Team A uses dollars, Team B uses euros, Team C uses yen. Saying "Team A delivers 60 while Team B delivers 35" is like saying "Person A has $60, Person B has €35" - completely meaningless without knowing the exchange rate.
Each team's story points are calibrated to their unique context: team skills, technical debt, domain complexity, tooling. Only the trend within a single team matters. Cross-team comparison creates toxic competition and estimate inflation.
Run a cross-team estimation exercise with the same 5 user stories. Watch teams give wildly different estimates for identical work. When leadership sees it firsthand, the velocity comparison debate dies.
We did this - same "add export to CSV" story was 3pts for Team A (they have a library), 8pts for Team B (legacy codebase), 13pts for Team C (regulatory compliance). All correct for their context. Ended comparison permanently.
Focus CTO on throughput metrics instead: stories completed per sprint, cycle time, deployment frequency, lead time. These ARE comparable across teams.
Team completing 20 stories/sprint vs 12 stories/sprint IS meaningful data. Team deploying 50x/week vs 2x/week IS meaningful. Velocity (points) is internal planning tool only. Shift leadership to metrics that actually measure flow and customer value.
Controversial take: Just normalize the estimates. Have all teams estimate the same 10 reference stories, calculate conversion factors.
If Team A estimates baseline set as 50pts total and Team B estimates same set as 100pts, then Team B velocity = Team A velocity × 2 for comparison purposes. Imperfect but gives the apples-to-apples comparison management demands. Better than nothing.
Real consequence of velocity comparison: Teams start gaming estimates. Once Team B realizes they're being compared to Team A, they'll inflate estimates to boost velocity. Now you have two broken metrics instead of one working metric.
Ask CTO: What decision will you make with this data? If answer is "identify weak performers," you're destroying team culture. If answer is "allocate resources," use customer value metrics instead.