Developer productivity is the question of how much useful software work a developer or team can deliver, and at what cost to quality, review burden, coordination, learning, and long-term maintenance.
In AI-assisted software development, productivity claims need careful measurement. More merged pull requests or faster local implementation may be real signals, but they are incomplete unless paired with defects, reverts, review load, flow, and delivered value.
Links
- DORA
- The SPACE of Developer Productivity
- Microsoft rollout study: Adoption and Impact of Command-Line AI Coding Agents