The AI Strategy Playbook: What engineering leaders need to know about driving AI impact
Why some teams see 20-point jumps in productivity while others see performance decline
Justin Reock
Deputy CTO
This post was originally published in Engineering Enablement, DX’s newsletter dedicated to sharing research and perspectives on developer productivity. Subscribe to be notified when we publish new issues.
Earlier this year, we published The Guide to AI-assisted Engineering, a practical resource that engineering leaders could share with their teams to enable AI best practices. It shared effective prompting techniques to use AI tools in ways that are proven to save developers time. The response was great: I heard from a lot of leaders who shared that they’d distributed it internally and received positive feedback from their teams.
That guide was written predominantly for engineers. It only had a brief section at the end for leaders, with recommendations on how to support and drive AI impact. Recognizing interest in that topic, I analyzed data from DX and collaborated with leaders, and eventually expanded our recommendations into a talk that I’ve been giving at conferences and internal company workshops for the past few months. We just published the written version of that talk into a new guide.
You can download the guide here, or read on for a summary of the main takeaways.
Why this guide matters now
AI is driving the biggest change in software development practices since DevOps. Many organizations are seeing major gains in developer velocity and satisfaction, but others are seeing chaos, confusion, or stalled adoption. The difference isn’t really in the tools, which are leapfrogging each other and rapidly reaching feature parity. The difference is in leadership strategy.
This new guide distills data and observations from DX’s research team. It’s aimed at senior engineering leaders who want to enable and accelerate their organization’s integration of AI.
Key takeaways for leaders
- Psychological safety comes first. Teams that view AI as an augmentation for existing engineers see faster adoption and higher overall satisfaction
- Developer productivity measurement remains the same, but new metrics are needed to get a lens into the impact of AI. This guide recommends the AI Measurement framework, a mixed-methods approach to AI measurement that balances telemetry, surveys, and experience sampling across utilization, impact, and cost.
- Strong adoption follows strong enablement. The most successful teams we spoke to invest in structured education, controlled pilots, experimentation, and clear compliance before scaling out to the whole organization.
- Leaders should be thoughtful about how they introduce AI tools and ROI metrics. They should position AI as a performance multiplier: a way to help teams move faster, improve quality, and unlock new capabilities across the organization. It’s also important to emphasize that these tools are meant to enhance engineers’ abilities and expand what’s possible, not replace their roles.
- The biggest opportunity isn’t code generation. By leveraging AI automation across the SDLC, focusing on reviews, documentation, incidents, and refactoring, we fix the real bottlenecks that constrain throughput.
Whether you are just starting with your first AI adoption plan or beginning to scale existing pilots, this guide should help you bolster your organization’s AI rollout.