You randomized viewers but need a creator metric — why can't you read it?
Experiment designHardProblem. You're testing a new Reels ranking model to see if it improves creator outcomes (creators getting more reach, posting more, staying active). You randomize viewers 50/50. Why is this the wrong design for a creator-side metric, and what would you do instead?
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.