Spotify
Enhancing music discovery and education in the Spotify ecosystem.
ROLE
Product Designer
TIMELINE
2025 - 4 Weeks
TEAM
2 designers
TOOLS
Figma
Overview
Designing AI-powered tools to transform music discovery and learning on Spotify.
WHAT IS “SPOTIFY”?
Spotify is one of the biggest music streaming platforms globally, giving users access to millions of songs and audio content from artists across the world.
PROBLEM SPACE
Spotify’s algorithm often faces criticism for restricting music discovery and only exposing users to narrow range of artists and genres. Additionally, many of its discovery features are passive, making it difficult for users to shape and personalize their recommendations.
SOLUTION
My design partner and I created AI-driven features that empower users to play a more active role in personalizing their music recommendations and discovery experience.
JUMP TO FINAL DESIGNS
→
Ideation
THE ECHO CHAMBER
Spotify users aren’t satisfied with their music recommendations
Spotify has over 670 million users worldwide, but many share similar concerns when it comes to the algorithm and music discovery experience. It is often difficult to find new music because many discovery features are passive, making the experience feel limited and uninspiring.
Motivations
WHY DOES IT MATTER?
To maintain and attract users, Spotify needs to improve their music discovery experience. But how?
Even Spotify recognizes that their recommendations can improve. We took a look at Spotify’s research blog and what work is being done to help users expand their listening behavior.
The answer? Allowing users to refine their recommendations.
SO THEN,
How might we enhance music discovery in Spotify to empower users to explore diverse music in a more personalized and engaging way?
Design Principles
Outlining core themes to guide our solutions
Based on insights from our user interviews and secondary research, we formulated design principles that guided our explorations and decision-making moving forward.
Ideation
We organized our ideas by effort and customization.
Based on our principle of low cost, we compared the potential effort level of our ideas to existing Spotify features that users commonly use.
Because many existing features already covered low effort and low customization spaces, we focused on ideas that offered higher customization.
OUR FINAL DESIGNS
A blend of active and AI-driven features reimagine how Spotify listeners can actively shape their discovery journey.
AI Deep Dive
AI Deep Dive educates users about the makeup of their playlists and albums, making personalized recommendations and introducing new music in a way that deepens users’ understanding of their own musical taste.
Radio Refresh
Users found that radios often prioritized songs they liked over ones that truly matched the sound of the original song. Radio Refresh lets users fine-tune radios based on specific aspects of a song they love, making music discovery more personal and accurate.
Jam Recommendations
Don’t know what to add to your jam? Jam recommendations recommend a mix of popular tracks and personalized picks to keep the vibe going.
Daylist Refresh
With Daylist Refresh, users can tweak their Daylist by amplifying the moods and styles they enjoy most.
Daylist - Exclude Songs
Users were frustrated that Daylist often included many songs they already liked, rather than introducing fresh recommendations. Our feature allows users to toggle between including liked songs or focusing on seeing new discoveries.
Explorations
Early ideas that informed our final direction.
Through testing and iteration, we learned which of our ideas truly resonated with users and which ones helped shape our thinking, even if they didn’t make it into the final experience.
RADIO REFRESH: INITIAL ITERATIONS
While these versions didn’t move forward, they surfaced important considerations about how much flexibility users would want. They also highlighted areas where interaction patterns and visual feedback could be refined for greater clarity and ease of use.
DAYLIST VIBE REFINE
Users found the Vibe Refine feature to be helpful, but its use case was quite niche. It would only be used when users have multiple vibes for a single Daylist time, which may be infrequent.
AI SONG COMPANION
AI Song Companion was originally the song counterpart of our AI Deep Dive feature, but we felt that it was too open-ended for a genAI experience. When testing with users, the use case was confusing when compared to other song-specific features like Song Radio.
BEHIND THE SCENES WIDGET
We originally explored a Behind the Scenes widget to better connect listeners to broader aspects of a song. However, we thought its impact on music discovery would ultimately be limited compared to other features.
Takeaways
EXPLORE FROM EVERY ANGLE
I typically approach ideation with a focused mindset, but this project helped me realize that loosening the reins can significantly broaden the scope of my ideas. Collaborating with my design partner, who encouraged us to diverge and explore a wider array of possibilities, taught me the value of thinking more expansively and freely.