CAROL LEE
Developer Productivity
Problem: Lack of alignment on how teams and orgs define and measure productivity, leading to misconceptions about developers' work and effort.
Outcomes & Solutions: During a field study, we found that although IC developers, managers, and leaders would all like to prioritize performance over productivity or production, they disagree on how to measure performance and often inadvertently measure production or productivity instead. To mitigate this, we developed and disseminated the Three-Layer Productivity Framework to enable teams to create alignment on how performance, productivity, and production are measured.
Purpose & Method: Field study examining how IC developers, managers, and leaders define and measure production, productivity, and performance.
Outcome: Overall, participants reported a strong preference for an increased use of code quality metrics than is currently used, and a decreased use of raw counts of code commits and pull requests than is currently used. Of note, zero participants selected number of code commits and number of code reviews as their preferred metrics, despite some reporting organizational use of these metrics.
Outcome: Observationally, both engineering leaders and ICs ranked usability and value of work to the org as their current top measures of productivity, but indicated a preference for code quality. However, this pattern was flipped for managers: managers reported that code longevity, followed by code quality, were the primary metrics currently used, but indicated a preference for usability and value of work to the org.
