Your question is Risks of Peeking in Ride-Share Test. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You work on a ride-sharing marketplace experiment and your team is monitoring a new product change in an A/B test. After a few days, the treatment looks better on the primary metric, and stakeholders want to check results every morning.
What are the main risks of peeking at experiment results early, and how can early looks distort the final decision? In your answer, explain how you would protect the integrity of the test before deciding whether to ship.