Diagnostics Over Metrics

Four people told me today that my talk Let’s Help Melly was the most inspiring they had seen in a long time. Does that make me feel good? Of course! Is it actionable? Not at all.

One of the most important reasons for measuring things in an organization is so that we learn. When metrics don’t help you to improve, why bother? Eric Ries, of Lean Startup fame, calls them “vanity metrics”. Measuring things that will make you feel good about yourself can be useful, particularly if your ego is in need of a good hug. Which is nice. But it won’t help you test a hypothesis.

Testing hypotheses, or “validated learning”, is what doctors do all the time. Doctors rarely use terms such as metrics or KPI’s. They prefer other terms, such as tests and diagnoses.

Organizational health

I’ve said it before and I’ll say it again: In an ever-changing environment, organizations should be healthy. When we want to improve an organization’s health, we need to run tests. We need to have diagnostics. That’s what metrics and KPI’s should be all about. As soon as we learned what we need to know, when we have validated our hypothesis, we can stop testing. We should kill the metric, and measure something else. There’s no point in a recurring ritual of collecting metrics, unless it’s about monitoring for the absence of a deadly disease.

Instead of measuring how many people I have inspired, I should be measuring how I helped people improve their work. I should measure stories, not compliments. Because the goal of tests, measures, and diagnostics, is to learn how to improve the health of a system. Not to improve the ego of a manager.

Or a presenter.