This week, I read A Human-Centered Approach to Developer Productivity. This paper takes on how we can measure and improve developer productivity without too many metrics.
The productivity of developers is usually thought to be measurable using single or straightforward metrics. However, in reality, it is much more complicated than that. Measuring the productivity of software developers is difficult because it involves evaluating humans engaging in a complex and creative task. Their productivity is influenced by both technological and social factors, which need to be added to the equation together to gain a deep understanding of developer productivity.
Understanding Developer Productivity
The question “How do we make developers more productive?” isn’t just about shipping code faster or building and deploying in crazy short times. It’s about supporting humans—humans who happen to be developers, yes, but humans at all times.
Google’s research team took a deep dive into the nuances of engineering tools, infrastructure, and processes, and has been asking the hard question, “How can we measure productivity without associating people with numbers?”
Well the Spoiler: It’s not easy!
The study highlights that humans are messy, and so is measuring productivity in software engineering. While a framework or tool might speed up some parts of the process, it’s the deeper, more complex human factors that often make or break developer productivity.
My Summary of the Paper
Google’s team broke down the challenges of improving developer productivity into several key points. It’s a story of trying to answer: “How can we support developers in their work without turning them into cogs in a machine?”
- Memory and Decision-Making: The research highlights how developers face cognitive limits. They can only hold so much information at once, and interruptions or shifts in focus can derail their progress. If a build takes too long, for instance, developers might switch tasks and lose track of their flow, making it harder to pick up where they left off.
- Task Complexity: Not all complexity is necessary. Some tasks come with what Google calls “accidental complexity”—obstacles that don’t need to be there and only serve to slow down progress. If we can reduce these obstacles, we can make things easier and more productive for developers.
- Team Dynamics: Working with others adds layers of complexity to any project. In these times of hybrid or WFH working models, geographic distance, time zones, and differences in experience all impact how teams collaborate. It’s the human side of teamwork that can make the biggest difference here. How well do people support each other? How effectively do they communicate? These are just as important as the tools they use.
- Organizational Context: The productivity of a company is significantly influenced by its culture and organizational structure. These are the factors that are often overlooked. For example: if an organization rewards speed over quality, developers will feel pressured to deliver quickly, even if the long-term impact on code quality is negative.
The best thing about this study is its focus on the cognitive work behind coding. Often, we don’t see the thinking, problem-solving, and research that developers do before they even write a line of code. These cognitive efforts don’t result in immediate outputs, but they are crucial for long-term productivity gains.
One surprising but relatable example is how build speeds affect developers. When builds are fast, developers can stay in their flow, maintaining focus on the task at hand. But when builds slow down, they switch tasks, lose focus, and end up taking longer to regain their previous momentum. This is a classic example of how small technical hurdles can have bigger human impacts.
There is more than just coding
The research stressed the need to move away from traditional productivity metrics like counting lines of code, as it doesn’t truly capture the complexity of software engineering. The key is to understand that productivity also involves factors like problem-solving, strategizing, and learning, which are hard to measure but truly fundamental to outcomes.
Taking a Holistic View of Productivity
Google’s team recommends taking a holistic approach to productivity, considering human factors, collaboration, and long-term impacts. I recall listening to Dr. Nicole Forsgren on Lenny’s podcast How to Measure and Improve Developer Productivity where she emphasized the need for a comprehensive perspective.
In line with this thinking, frameworks like DORA (DevOps Research and Assessment) and SPACE (Satisfaction, Performance, Activity, Collaboration, and Efficiency) offer a multi-dimensional way to measure productivity. The main idea is to balance short-term wins with sustainable, effective team maintenance. Recognizing the importance of inclusivity, fairness, and collaboration directly affects developer output and retention.
Final thoughts
Developer productivity is so much more than tracking lines of code or measuring speed. The research by Google makes it clear that if we really want to boost productivity, we need to focus on the humans behind the work—their cognitive load, how they collaborate, and the environment they work in. By taking a more human-centered, holistic approach, we can create spaces where developers aren’t just faster, but more effective, creative, and happier in what they do. It’s not just about getting app-builds out the door—it’s about setting developers up, to truly make an impact and succeed.
Referenced:
- DORA Framework: https://dora.dev/
- A brief intro to DORA: https://devops.com/what-is-dora-and-why-you-should-care/
- SPACE Framework: https://queue.acm.org/detail.cfm?id=3454124
That’s all folks!
Every two weeks, I share research and perspectives on tech, product growth, developer productivity, and engineering excellence.
Subscribe here to stay in the loop.
Thanks for reading!
— Ibtihaaj