DPE Lowdown - How Meta does Developer Productivity Engineering with Adam McCormick

In the opening segment, Adam shares his new living situation and discusses the rainy weather in Seattle, while reminiscing about past events and introductions made through Jay Zimmerman at various conferences. Justin highlights the purpose of the webcast, which is centered around productivity engineering and the experiences shared by Adam McCormick from Meta. They emphasize the need for enhancing developer productivity, specifically in large systems such as Facebook with nearly 3 billion users.

Adam describes his role on the Facebook app reliability engineering team, explaining how they work to minimize crashes for a massive user base. He explains the complexity of maintaining reliability at scale and the importance of getting code quality assurance integrated into the development process. The team collaborates with multiple product engineers to establish best practices and investigate issues as they arise, particularly for Android applications.

The discussion shifts to how frequently and the nature of app releases at Facebook. Adam details their build and release processes, including the number of builds they manage daily and the stages of releasing major and minor updates. With about 30 to 50 thousand builds created every day, their approach allows for better deployment and testing of all changes, emphasizing the need for a reliable infrastructure to manage such scales.

Adam clarifies the structure and extent of the productivity engineering group at Facebook, mentioning that they have about 1,000 dedicated engineers within a larger workforce, facilitating developers through tools, testing, and environment optimization. The complexity of the monorepo approach is discussed, highlighting how it contributes to the challenge of handling such large code bases and the subsequent investments made to efficiently manage them.

The focus is on the daily experiences of Meta's software engineers as they work on the Android app. Adam discusses the tools and customized systems they have in place, such as Android Studio and their build systems, to streamline coding, testing, and CI/CD processes, showcasing the heavy reliance on simulations and tests to catch issues before reaching broader audiences.

In this segment, Adam emphasizes the significance of productivity engineering as part of their success in software development and shipping at scale, underscoring that it's not merely about adding more engineers but about investing in structured approaches and solutions that improve overall productivity.

They delve into how Meta measures developer happiness and productivity, sharing that they conduct regular surveys and micro-check-ins to gather feedback and analyze the impact of changes on developer efficiency. They also appreciate the importance of consent in data collection, ensuring developers can opt out of specific productivity tools if desired.

Adam shares insights into notable productivity engineering wins at Meta, explaining how they have standardized processes to reduce the complexity of coding and analysis through consolidated tools and services that streamline common tasks, enhancing overall developer experience.

The conversation touches upon employee feedback regarding potential pitfalls of excessive measurement leading to micromanagement perceptions. Adam notes that they focus on the impact of tooling and systems on productivity without imposing numerical goals on individual engineers, emphasizing that the goal is to foster a culture of effective and happy developers.