Measuring Engineering
9 months ago
- #productivity
- #software-development
- #cycle-time
- Software development is unpredictable and requirements constantly change.
- Cycle time is the duration between a ticket opening and closing, often seen as a key productivity metric.
- A study analyzed 12,000 contributors across 200 organizations to understand cycle time factors.
- Factors affecting cycle time include merged PRs, comments per PR, coding days, defects, and collaboration.
- More merged PRs and coding days reduce cycle time, while more defects and comments increase it.
- Variability in cycle time is high, making individual comparisons meaningless.
- Developer productivity should be viewed as variable, like weather, not a fixed attribute.
- Focus on team-level trends and system improvements rather than individual metrics.
- Cycle time simplicity can lead to incorrect assumptions without context.
- Aggregate trends and qualitative signals are more meaningful than individual snapshots.