Book: SE@Google Ch 7: Measuring Engineering Productivity

[t-cheepeng]

Book: SE@Google
Chapter: 7. Measuring Engineering Productivity

Summary:

• As scale grows, more personnel is needed. Communication overhead increases and thus, there is a need to make individual engineers more productive so this overhead can be reduced.
• Before even measuring anything, there is a need to find out if the thing is worth measuring
  • What results are expected, and why? If the data supports/rejects the result, what actions will be taken?
  • A not worthwhile measurement could be due to various factors like not being able to change processes immediately, results invalidated by factors, results being used to do a post-hoc explanation for something that is going to happen.
• GSM Framework
  • Goal, desired end result
  • Signal, how you might know that you have achieve the goal
  • Metrix, a proxy for signal. something that is measurable albeit not being ideal
• Identify Productivity Goals with QUANTS
  • Quality of code, Attention from engineers, Intellectual Complexity, Tempo & Velocity, Satisfaction
• Every goal should have >= 1 signal
• Quantitative metrics are useful in aggregating data. Qualitative studies give insights on how to improve the process that is being studied