Every organization claims to be data-driven. Almost none of them are — and the ones that get closest often make worse decisions because of it.

“Data-driven” suggests the data decides. But data doesn’t decide. People do. When teams treat metrics as truth rather than evidence, they optimize proxies, game dashboards, and mistake confidence for competence.

Agile, Lean, DevOps, and Design Thinking all share the same root: disciplined experimentation. Yet many organizations adopt the vocabulary without adopting the epistemology. They measure what’s easy instead of what matters, confuse correlation with causation, and A/B test their way into local maxima.

This session reframes what it actually means to think scientifically at work. We’ll examine how metrics drift from outcomes, why velocity gaming is predictable, and how teams can use data as a tool for learning rather than a substitute for judgment.

If you’ve ever watched a dashboard improve while performance stagnated, you already know the problem. This talk names the system behind it.