Hypothesis tracking

An experiment without a written hypothesis is hard to learn from. If nobody wrote down what was expected and why, a flat result is just a shrug, and a win gets remembered as whatever story is told afterwards.

Pertento keeps the hypothesis on the experiment, alongside its PXL score. That means the reasoning is visible while the test runs, not just while it is being planned, and it is still there when the result arrives.

The payoff is in the losses, which are most of them. A recorded expectation that was wrong is a fact about your customers. An unrecorded one is nothing at all.

Because scoring and hypothesis live on the same record, the backlog is also the reasoning: the list of what to run next is a list of arguments, ranked.