Frameworks for statistically significant tests?

Hey all,

Does anyone have a good framework or principles for thinking about whether to launch/GA something when it’s not technically stat sig from the A/B experiment, but the results are directionally positive? I want to be as data-driven as possible but I don’t want to suppress my intuition about the product feature. Fun times being a PM!

1 Like

This topic was automatically closed after 180 days. New replies are no longer allowed.