Explain and derive CUPED (it was invented here)

Variance & ratio metricsHard

Problem. Your success-rate experiment is underpowered for the small effects you care about. Explain CUPED, derive why it reduces variance without adding bias, and say how you'd apply it at Bing.

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.

← Success up, ad revenue down, with a novelty curve — ship?1000s of experiments and daily peeking — control the error →