Build a Podcast Discovery Experience That Helps Listeners Find Episodes Worth Their Time
Company: Meta
Role: Product Manager
Category: Product Design & Strategy
Difficulty: medium
Interview Round: Other
You are asked to design a product that helps people get good podcast recommendations. The prompt is deliberately open: decide who the product serves, where it lives and what a good recommendation means, then design the experience and explain how you would measure whether it works.
```hint Price the cost of a bad pick
A podcast episode takes far longer to judge than a song or a short video, so a poor recommendation costs the listener a lot of time. Let that cost shape both the signals you trust and the way you present a recommendation.
```
### Clarifying Questions
- Is this a standalone podcast app or a recommendation surface inside an existing app that already knows a user's identity and friends, and may it use signals from that app?
- Which surfaces are in scope: a home feed of suggestions, what plays after the current episode, search results, or all of them?
- Are we serving listeners only, or should the product also help new creators get discovered?
- Should success mean turning occasional listeners or non-listeners into regular ones, or deepening engagement for people who already listen every week?
### What a Strong Answer Covers
- A stated goal and platform assumption connected to why podcast discovery is hard: long episodes, a huge catalog, quality that is hard to judge up front, and cold start for new listeners and new shows.
- A chosen listener segment with pain points specific to that segment rather than to podcasts in general.
- Solutions that pair recommendation inputs (stated preferences, listening behavior, content understanding, social signals) with the product surfaces that present them, prioritized by impact and effort.
- A north-star metric that rewards satisfied listening rather than taps, with guardrails for content safety, creator diversity and privacy.
- Trade-offs and risks such as narrowing taste bubbles, misleading or harmful episodes, and the privacy of borrowed signals.
### Follow-up Questions
- How would you recommend a brand-new show that has no listening history yet?
- Listening minutes rise after launch, but listeners rate the recommendations lower in surveys. What would you investigate first?
- How would you keep the system from amplifying misleading episodes without burying smaller creators?
Overview: A product sense prompt asking a PM candidate to design a product that helps people find podcasts worth their time. It tests scoping an open prompt, choosing a listener segment, pairing recommendation signals with product surfaces, prioritizing features and defining satisfaction-based success metrics.