Measuring the impact of a change you can't A/B test
Causal inferenceHardProblem. A change you can't A/B — a default switch, an infra migration, or a marketing campaign — rolls out at the market level. How do you estimate its causal impact, and how does this connect to the forecasting part of the role?
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.