Does the app being slower actually cause churn?

Causal inferenceHard

Problem. A new version made app startup ~200ms slower. Someone plots startup time vs retention across users and says "slower users retain fine, so it doesn't matter." Is that right? How would you measure whether performance causally affects retention?

Before you reveal: say your answer out loud, as if you were in the real interview — get your reasoning across clearly first. There is no single correct answer: reading what the interviewer is really after and defending your own thinking is what makes an answer strong.

← Only 20% of users are affected — why does the metric look flat?Sends are up week one — but will it retain? →