Inside the DevEx Metrics Compass with André Meyer

Engineering Enablement by DX

0:00

-46:02

Inside the DevEx Metrics Compass with André Meyer

Researcher André Meyer shares how teams can measure AI’s impact beyond adoption, choose meaningful productivity metrics, and understand what makes developers’ work rewarding.

Brian Houck

Nov 12, 2026

Listen and watch now on YouTube, Apple, and Spotify.

Are we measuring the impact of AI on engineering, or just how much it’s being used?

In this episode of Engineering Enablement, I talk with researcher André Meyer about his DevEx Metrics Compass, why traditional productivity metrics often miss what matters, and how to measure the value AI actually delivers.

We also revisit André’s Today Was a Good Day study to explore what makes developers’ work rewarding, how AI is changing that experience, and practical ways to balance focused work, collaboration, and growing code review demands.

Some takeaways:

Why traditional metrics fall short

AI adoption and output metrics don’t demonstrate business value. Metrics like AI usage, acceptance rates, and pull request counts reveal activity, not outcomes.
Collaboration is largely overlooked in developer productivity measurement. Developers spend a significant portion of their time communicating and collaborating, yet few metrics capture its impact.
Measuring customer and business value remains a major challenge. Connecting engineering data to customer satisfaction, feature adoption, and business outcomes requires looking beyond traditional development metrics.

Choosing the right DevEx metrics

The DevEx Metrics Compass maps more than 120 metrics from research and industry. It helps teams navigate existing measurement frameworks and identify relevant metrics.
There’s no universal set of DevEx metrics. Organizations should select measures based on their specific goals, challenges, and level of measurement maturity.
More metrics don’t necessarily produce better insights. Focus on a manageable set of priorities, with more detailed metrics available to investigate specific problems.

Building a more effective measurement system

Telemetry tells you what changed, but not necessarily why. Pair quantitative data with developer surveys and qualitative feedback to understand the full picture.
Metrics should be evaluated and retired as needs evolve. Regularly assess whether they remain actionable, inform decisions, or show signs of being gamed.
Measure AI’s impact across multiple dimensions. Look beyond activity to developer satisfaction, quality, efficiency, and business outcomes, tracking changes over time.

What makes a good day for developers

Developers find satisfaction in making meaningful progress. They report better workdays when they can create something valuable, encounter fewer blockers, and have time to focus.
AI doesn’t necessarily diminish the satisfaction developers get from their work. Developers can still enjoy creating software when they retain control over what gets built and how decisions are made.
Not all context switching is harmful. Observational research revealed that the cost depends on the nature of the task switch, something automated telemetry alone can’t capture.

Protecting focus time and improving collaboration

The timing of meetings matters as much as their frequency. Scattered meetings fragment the workday and make sustained focus difficult.
Teams should coordinate focused work and collaboration time. Shared schedules can reduce interruptions while preserving opportunities to communicate and solve problems together.
AI-generated code is increasing the burden of peer review. Clear PR expectations, author accountability, and risk-based review practices can help reduce cognitive load and protect code quality.