I've been leading Engineering teams for fifteen years.In that time, I've watched the industry get very good at measuring the wrong things.

Output metrics tell you how fast the machine is running: pull requests merged, velocity points, deployment frequency. They don't tell you whether the team is healthy, whether people feel safe raising problems, or whether what looks like productivity is actually a team quietly burning out. I've always believed that Engineering management is fundamentally about taking care of people. The metrics should serve that, not replace it.

So I built my own. Predictability over velocity. Planned versus unplanned work composition. Defect trends. Culture surveys alongside delivery metrics, because the DORA research established that organizational culture is one of the strongest predictors of software delivery performance. Nicole Forsgren's public work gave me pieces to work with over time, but nobody handed me a survey methodology. I assembled that myself, because my teams needed it and nothing else existed.

For years, I ran all of this by hand. Spreadsheets, Google Forms, manual data pulls, charts I built myself, because the tools that existed never surfaced what I thought actually mattered. I brought feature requests to Jira. I reached out to other tooling companies and offered to partner on better dashboards. Nobody bit. So I kept doing it manually. Eventually I got tired of waiting and built it myself.

The methodology didn't start in Engineering. When things go wrong, the failure is usually in the conditions, not the people. Engineering generates value, real value, when the people doing it can deliver what they say, when they say, given a clear and unobstructed path.