Amazon Music Challenge
Design a feature for Amazon Music that increases user engagement and retention.
Music has always been a shared experience — passed around in cars, negotiated at parties, carried between people. But current streaming built the opposite: a platform optimised for one.
A collaborative listening feature where a host starts a mix and sets a vibe, and others join instantly to add songs — creating a shared, evolving queue seamlessly blended by AI-powered mixing.
Saved mixes, top-liked tracks, and nostalgia notifications turn each session into a memory worth returning to.
View Figma prototype ↗The two most critical interactions in BeatDrop — starting a mix as a host and joining one as a guest.
Guests join a session by scanning a QR code from their own phone. The session simultaneously broadcasts to any screen in the room.
BeatDrop allows users to save mixes and track top-liked songs, turning each session into lasting memories they can relive for years to come.
One tension surfaced across all 52 participants: music brings people together, but the interface puts one person in control. Others struggle to contribute without interrupting. Passing the device creates friction. The experience feels transactional rather than shared.
We ran task-based usability sessions with 15 participants to pressure-test the concept.
Testing revealed that some users were unfamiliar with DJ mixing concepts — so the AI Automix toggle ("Let AI choose") was introduced, lowering the barrier to entry without removing customisation for more advanced users.
- Mixed-methods research report — 52 participants, surveys + interviews, Double Diamond framework
- Full interaction design: collaborative queue, AI DJ layer, QR code guest access, memory features
- Usability testing with 15 participants — 93% strong interest, 40% willing to pay premium
- High-fidelity Figma prototype with annotated handoff documentation
- Presented at Amazon Music NYC offices — selected as winning concept