Ideas first. Evidence next. Investment follows.

Every experiment is written down before it starts: the hypothesis, what we expect, and how we will decide. Then we publish what happened, including the projects we stopped.