I’ve found that GTM experiments tend to start structured and then drift into docs, sheets, and tribal knowledge.

For people running a lot of experiments: do you use any system to track hypotheses, outcomes, and decisions, or accept that this stays messy?