Before you measure AI coding tools, listen to your engineers.
Companies roll out AI coding tools and start counting pull requests and lead time. The better first move is to ask the people doing the work what helps and what gets in the way.
The short answer
Put people before process. Before you track pull request velocity or lead time, talk with your engineers about what is and isn't working with the new AI tools, then act on what you hear, again and again. Metrics can come later and evolve with the new workflows. In the end, the metric that matters most is customer and client satisfaction, and a team that's happy with what it does and how it works is the best start.
The most common mistake is putting process before people. Companies invest in AI coding tools and start measuring pull request velocity and lead time, instead of first asking their developers what is working and what isn’t. Track metrics on top of a flawed process and you get low-quality outcomes, and you make your team members’ lives miserable.
Why does measuring first backfire?
Because any initiative, however well intended, will fall short if the team isn’t part of the decisions. And any metric attached to it will be useless.
AI tools and the paradigms around them are changing fast, so this can’t be a one-time conversation. It has to happen often. Learning, adapting and overcoming the difficulties along the way is what creates real value for your organization. Metrics can come later, and they can change to fit the new workflows you put in place.
Trust your team, give them a voice, and they will find a way.
What separates good companies from amazing ones is a team that feels safe and supported to keep reflecting on how it works, and to keep improving it.
What does the research say?
The evidence points the same way, and much of it comes from software teams.
- Safety comes first. Google studied 180 teams, 115 of them engineering project teams, and ranked psychological safety first among the five dynamics of effective teams: “a shared belief held by members of a team that the team is safe for interpersonal risk taking.”
- Learning is part of the job. DORA’s 2024 research calls transformational leadership and an environment of continuous learning critical.
- AI amplifies the team you have. The 2025 DORA report puts it plainly: “AI doesn’t fix a team; it amplifies what’s already there.” In its words, “the value of AI is unlocked not by the tools themselves, but by the surrounding technical practices and cultural environment.”
What should you measure first?
Start by measuring how satisfied your team is with its daily work, tools and processes. Learn about their pain points, and work with them to address those.
At the end of the day, the most important metric is your customers’ and clients’ satisfaction. Making sure your team is happy with what it does, and how it does it, is the best start you can have.
You’re supposed to hire people smarter than you, so listen to them. Work together to raise the quality of the work, and leave the organization in a better place than you found it, every day.
What I’d do this quarter
- Start real, genuine conversations with your engineers. Not surveys, not interviews.
- Use them to find out what's blocking people from delivering at their best, then act on it.
- Make these conversations a habit and a critical part of your culture.
Sources: Google re:Work · DORA · Google Cloud
Written by Luciano Ayres. Edited with AI assistance.