Product design question
How will you improve the recommendation engine of a streaming services product?
- ViacomCBS
- Product design
- Medium
Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.
Start a mock interview on this question · Mock interview from a job description
What this question tests
Product improvement thinking for a recommendation system, checking whether you can name specific signal and cold-start problems rather than a vague 'use better AI' answer.
How to approach it
- Identify the current gap: recommendation engines often over-index on recently watched genres, creating a narrow, repetitive feed that misses real user interest breadth.
- Prioritize solving the cold-start problem for new or infrequent viewers, since poor early recommendations are the biggest driver of disengagement for new subscribers.
- Propose improving signal diversity: incorporate implicit signals like time spent browsing a title's details without watching, not just completed views, to better capture intent and hesitation.
- Add explicit signal collection: a lightweight onboarding taste survey for new users, reducing reliance on sparse early behavioral data.
- Address the exploration-exploitation trade-off: deliberately surface a small percentage of diverse, higher-uncertainty recommendations to avoid over-narrowing the feed over time.
- Define success as increased content diversity in what's actually watched, alongside standard metrics like session length and completion rate.
What a strong answer includes
- Names the cold-start problem specifically, the primary weakness in most recommendation systems for new or light users.
- Proposes a concrete signal improvement, time-spent-without-watching as an intent signal, beyond simple watched-or-not binary data.
- Addresses the exploration-exploitation trade-off explicitly, showing awareness that pure exploitation narrows the feed and eventually reduces engagement.
Common mistakes
- Giving a vague 'use more machine learning' answer without naming specific signals or a specific failure mode being addressed.
- Ignoring the cold-start problem, which is the most common and consequential recommendation failure for new or infrequent users.
Likely follow-up questions
- How would you measure recommendation quality beyond just watch completion rate?
- How would you balance showing familiar content against encouraging discovery of new genres?
More product design questions
- How will you design a streaming service product that can cater to customers in the US and Canada assuming they have different needs?ViacomCBS · Product design · Hard
- How would you improve Outlook for the use case when people get overwhelmed by number of emails received after returning from a vacation?Microsoft · Product design · Medium
- You’ve just joined Meta and have been tasked with launching a new product to help people at home during emergencies. What would you build?Meta · Product design · Medium
- How do you design a camera for the elderly?Google · Product design · Medium
- Design a new feature for Twitter that improves the new user’s experience.Twitter · Product design · Medium
- How would you go about designing a new location sharing app for Google?Google · Product design · Medium
More questions from ViacomCBS
Learn the skill behind it
Chapters of the AI PM course that teach what this question tests.
- Chapter 4: Discovery and strategy for AI products
- Chapter 7: AI UX and human oversight: design for a system that is wrong sometimes
- Chapter 14: Get the job: the AI PM interview loop