202411281857 - Don't Go for 100% Correctness When Reading Engineering Metrics

I cannot stress this enough!
Your engineering metrics (cycle time, throughput, roadmap contribution, etc.) are not financial data - i.e. they are not supposed to be 100% correct, always expect deviation, and never aim for 100% correctness (you don't want to go down that rabbit hole...).

Why?
Because engineering metrics are based on human behaviour, and people make mistakes (we humans aren't perfect) - and your ticketing system (or any other data source for that matter) will never be sane - validate data, set constraints, but don't be a JIRA "natzi".

Your goal should be to have data that are clean enough to be able to be used for decision-making, not perfect.

Post by @jsifalda.bsky.social — Bluesky

This article was originally published on https://craftengineer.com/. It was written by a human and polished using grammar tools for clarity.

Follow me on X (Formally, Twitter) or Bluesky.