Product design question

How would you design a recommendation system for Disney+ for new customers (with less data)?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

AI and technical product thinking under a real cold start constraint: can you design a recommendation approach for users with little viewing history.

State the core challenge: collaborative filtering needs history, which new users do not have, so a different approach is needed early on.
Propose an onboarding step: ask new users to pick a few favorite franchises or genres at signup to seed a preference profile.
Use content based signals, matching genre, cast, and franchise metadata to the stated preferences, before enough behavioral data exists.
Blend in popularity and editorial curation, like trending titles and staff picks, as a safe fallback for the first sessions.
Transition users to behavioral collaborative filtering once they accumulate enough watch history, roughly after a handful of completed titles.
Define success as engagement and completion rate in the first two weeks, since that is when cold start risk is highest.

Explicitly separates the cold start phase (content based plus onboarding signals) from the mature phase (collaborative filtering).
Uses an explicit onboarding preference step, a concrete and testable design choice, not just says use less data.
Picks a metric focused on the vulnerable early window, first two weeks, rather than steady state metrics.

Proposing a generic recommendation algorithm without addressing why cold start is different.
No plan for transitioning users out of the cold start approach as data accumulates.

Chapters of the AI PM course that teach what this question tests.