We killed lines-of-code metrics, so why are we measuring AI productivity by merged PRs?

brandon_galang · x · 2026-09-29

The author questions the trend of citing "hundreds of merged PRs per week" as evidence of AI coding tool impact: we stopped treating lines of code as productivity, so why revert to merged-PR counts? A single PR can have more impact than hundreds of inconsequential repo updates. The real question is what outcome—what changed for the people using the product—are we actually measuring.

Original post →

More from coding & agent

coding & agent channel →