Design a feature that helps users discover content they didn't know they wanted.
Model answer
Model answerMost recommendation systems optimize for what users already like. I'd design for productive surprise: a weekly 'outside your bubble' section surfacing content slightly outside usual preferences, curated by editorial signals and social proof. The feature should feel like a trusted friend's recommendation, not an algorithm. I'd A/B test it against standard recommendations and measure click-through, saves, and return visits.
Framework · CIRCLES Method
- CComprehend the Situation
The problem: recommendation systems create filter bubbles, not discovery - IIdentify the Customer
Users who consume regularly but report feeling 'bored' of their feed - RReport Customer Needs
Surprise, trust, and curation — not more of the same - CCut through Prioritization
Weekly cadence — not constant — to make it feel special - LList Solutions
Editorial picks, friend-of-friend signals, trending outside your network - EEvaluate Trade-offs
Editorial is high-quality but doesn't scale; blend with algorithmic signals - SSummarize Recommendation
Outside your bubble' weekly section, measured by saves and return visits
Related questions
- How would you improve the onboarding experience for a B2B SaaS product?
- You're the PM for Instagram Stories. How would you improve it?
- How would you design a feature to reduce churn in a subscription app?
- How would you approach redesigning a core feature after negative user feedback?
- You're the PM for a maps app. How would you improve it for cyclists?