Workhuman increases ROI from AI assistants 21% with DX
Workhuman, a leader in employee engagement software, relies on an engineering team of 250+ to build and deliver its category-leading platform. As part of a broader effort to improve developer efficiency, the company invested in GitHub Copilot. Early usage showed promise but varied across teams. To maximize its benefits, Workhuman needed a way to measure Copilot’s impact and identify adoption gaps.
Measuring Copilot’s effectiveness at scale proved challenging. “We knew Copilot was helping but didn’t have a way to understand which teams were using it, to what extent, and how it was helping,” says Andrew Keogh, Director of Engineering.
Workhuman was already using DX to measure developer experience and saw an opportunity to extend it to Copilot. Keogh’s team implemented DX reporting to track adoption, time savings, and its correlation with key metrics like velocity and codebase experience. This gave Workhuman clarity into how Copilot was being used, where adoption could improve, and its broader impact on developer productivity and experience.
DX data revealed a trend: while Copilot significantly boosted productivity for fully onboarded developers, adoption remained inconsistent across teams and tenure. Keogh’s research, using DX, showed that some developers needed more support and education to integrate Copilot effectively into their workflows.
“DX helped us see where Copilot was working, where it wasn’t, and what to do next. That’s what turned early adoption into measurable impact,” Keogh explains. Using these insights, his team launched training and documentation initiatives based on the practices of the most successful teams to drive broader adoption.
Since then, Workhuman has seen a 21% increase in developers achieving productivity gains with Copilot, unlocking over 12,500 developer hours annually. Copilot users also report an 11% higher developer experience score and 15% greater velocity than non-users.
With clear ROI data and measurable developer benefits, leadership confidently expanded its investment in Copilot. Today, Copilot adoption exceeds 80%, with AI generating 18% of all code.
“DX gave us the ability to prove Copilot’s time savings and tie them to improvements in cost, throughput, and developer experience, which changed the conversation with leadership,” Keogh explains. “Instead of just saying ‘we hear engineers like it,’ we can quantify its impact and make informed decisions.”
As Workhuman broadens its use of AI in engineering, DX will continue to measure its impact and guide decisions with clear data.
