Skip to Main Content
To succeed, the team needs to make decisions effectively as a team. This requires discussion, but not as much as most groups think. Whenever a design discussion takes more than about 10 minutes, it's time to run an experiment. It's important to be clear about what you're suggesting: the idea isn't to build two systems, then pick the better-designed one. Instead, we agree to admit when we've stopped arguing on the basis of evidence and instead have descended into pontification and supposition. Sometimes, to evaluate an idea, we have to try it for a while.