You are a PM at a video streaming service. In tier-3 towns and rural districts, session abandonment within the first 30 seconds is 41% against 12% in metros. Median connection speed there is 1.8 Mbps with frequent drops. Data is a real cost to these users — many are on daily-limit prepaid packs. Content preference skews to long-form regional-language drama. Most watch on one shared household phone, often in the evening.
Design a product change for this audience. Pick a specific user and problem, propose something concrete, and name how you would measure success.
100 points, 60% to pass.
Adaptive bitrate is the obvious answer and probably already exists; a strong answer goes past it. The constraints given are not only bandwidth: a daily data pack makes a failed start an actual financial loss, a shared phone means viewing is negotiated and time-boxed, and long-form drama means the cost of a bad start is repeated nightly. That points at scheduled overnight download of the next episodes on a cheaper window, resume-across-interruption that survives a connection drop, and showing data cost per episode before playback. Measure completed episodes per week per household, not sessions started.