You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
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
The text was updated successfully, but these errors were encountered:
Book: SE@Google
Chapter: 7. Measuring Engineering Productivity
Summary:
The text was updated successfully, but these errors were encountered: