The engineers who write the most code are usually the ones I'd hire last.

Volume looks like productivity, and it's often the opposite. A thousand lines that reimplement something the team already has is future maintenance. It's a new place for the bug to live and one more copy that will quietly drift.

The engineers worth hiring delete more than they add. They notice the third time a pattern shows up and pull it into one shared place. Their PRs are unglamorous and their impact compounds, because they make the codebase smaller and the team faster in the same move.

I built a whole platform around this bias. The rule is simple: if two apps could share it, it goes in one library, and every app consumes it through a thin adapter. A good abstraction is measured by how little code an app needs to use it.

Hire for raw output and you get a codebase that grows faster than your team can hold in their heads. Hire for leverage and it stays legible as it scales.

Disagree? Tell me which you'd rather inherit: 10k lines of clever, or 2k lines of shared and boring. I know my answer.