Your question is Experiment Design for Feature Impact. 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).
How would you design an experiment to measure the impact of a new feature on user engagement?
Assume this is a hands-on design exercise. Explain how you would define success, choose the randomization unit, size the experiment, analyze the results, and decide whether to ship. State any assumptions explicitly, and address statistical validity risks such as peeking, novelty effects, sample ratio mismatch, and interference between users.