SE-Radio Episode 317: Travis Kimmel on Measuring Software Engineering Productivity

Metrics Importance

In the realm of software engineering, the significance of well-defined metrics cannot be overstated. Travis Kimmel highlights how engineering has lagged behind other industries in adopting metrics-driven management, despite its role in enabling such practices elsewhere 1. He argues that without benchmarks, engineering narratives lack the quantitative backing needed for effective communication with stakeholders. Kimmel notes, "There's really no benchmarks to compare this quarter to last quarter," emphasizing the need for metrics to bridge this gap 1.

   

Challenges

Implementing engineering metrics comes with its own set of challenges. Travis Kimmel points out that many agile metrics rely heavily on self-reporting, which can lead to inaccuracies like overestimating tasks to meet arbitrary targets 2. He stresses the importance of developing metrics that provide accurate signals without adding unnecessary overhead to teams. "Do it iteratively," he advises, suggesting a focus on one aspect at a time to avoid overwhelming the team 3.

   

Visualizing Data

Effective visualization of engineering data is crucial for improving communication and decision-making. Travis Kimmel emphasizes the power of visual tools like D3 for creating high-visibility updates that can depoliticize discussions by presenting clear, objective data 4. He states, "Data depoliticizes everything," highlighting its role in transforming subjective opinions into fact-based conversations 5. This approach not only aids in internal communication but also aligns engineering efforts with broader business objectives.