Metrics that matter: Measuring developer productivity in the AI era

Mark Quigley is the Head of Platform Engineering at 90. Previously, he worked at Wayfair and other high-growth companies. Mark has a track-record on developer productivity and engineering leadership.

Key Topics Discussed

Rethinking Developer Productivity
Reporting AI Productivity Gains
Presenting Metrics to Leadership

Where to Find Mark Quigley

Mark Quigley: Mark Quigley
90: 90

Transcript

Rethinking Developer Productivity

For Mark, the fact that there is a high correlation between what makes engineers happy and what makes them productive is a “gift of the universe.” He proposes leveraging a Developer NPS (Net Promoter Score) to gage the levels of productivity of your organization.

However, he also explains that Developer NPS is just a starting point. Several other frameworks like DORA are known to move the needle for the organization when tracked adequantely.

Reporting AI Productivity Gains

Microsoft and GitHub have released reports in which they’ve observed productivity gains of 40% for organization using Copilot. However, in Mark’s experience, his teams are experiencing improvements in the 13–22% range.

While this number is not yet what the hype is claiming, the advantage is significant. Developers at 90 are now using more time to do more significant tasks, which can compound to higher productivity gains.

Presenting Metrics to Leadership

The real value of metrics is to support decision making, not the precision of data. Mark warns against over-focusing the discussion on how metrics are calculated or presenting raw results without providing context or a way to act on the data. “Quit debating it, just take the leap,” is Mark’s advice.