Discovering Music Through Friends — A Spotify Case Study

Spotify houses over 40 million songs to discover and fall in love with. Still, myself and many other users fall into the routine of listening to the same music on repeat.

There are so many ways Spotify attempts to suggest music to me. But, browsing through hit-or-miss playlists and auto-generated suggestions makes the discovering music disorganized and time consuming. I’m sure lots of work goes into curating Spotify playlists or predicting users’ likes, but algorithms cannot always align with music tastes.

To find out if others felt the same, I dove into interviews with other music lovers.

Understanding How People Discover Music

1. People usually listen to their own library

“If you make your own playlists, it’s so easy to just keep listening to your own music.”

2. People discover new music the most through friends and word of mouth

“Friends will just tell me what they’re listening to, or play a song out loud in person.”

“I have one friend who shares my music taste, and I like to stalk their playlists or send them links”

3. One bad experience can make people distrust Spotify discovery

“There was a period of time where I used Discover Weekly a lot because it consistently gave good music but later I didn’t really like it as much so haven’t been back in a while.”

Users tend to avoid browsing music suggestions in Spotify because they distrust Spotify, or are lazy — and they would rather discover through friends and word of mouth. But, there is still a disconnect between receiving a recommendation from another source and listening to it.

Current ways to share with friends disrupt the flow of browsing of other social media/conversation

I discovered that people are most likely to listen to a new song if recommended by a friend. They have the confidence that their friend knows their music taste well. But, when sharing music to social media or in person, they still either

  • forget about them, or
  • don’t feel like switching between messaging and music apps.

The Problem

People want to discover music that fits their taste, but it is hard because:

Suggestions generated by algorithms cannot match our wide music tastes.

It’s easy to forget about music recommendations from friends.

Final Interaction

Prototype for music discovery (left) and sharing (right) interactions

How I got there


  • Social interactions: message inbox, groups (“taste buds”), sharing feed
  • Customization: A/B testing to better determine taste, user-created mood board
  • Organization: “Listen to Later” feature

After taking a vote I decided to focus on the social solution space. How might we make finding new music more human and intimate?

How Social is Too Social?

Sketches of inboxes and conversation screens

I continued with different variations ranging from message-heavy to no messaging at all.

Explorations for a messaging inbox (left) and chat (right) feature

Though I attempted to find a way to emphasize the music shared rather than the chatting aspect, I realized that this solution was straying from the People Problem about music discovery. It also may be too dramatic of a change and doesn’t align with Spotify’s mission of sharing music (not conversation). So, I tested three different prototypes with users:

  1. Typical messaging inbox
  2. Inbox but no messaging
  3. A smaller inbox built into existing “Made for You” page
Pros and cons of three different prototypes

User feedback reiterated my new hypothesis. People don’t a separate messaging space for music, but still want more convenient sharing between friends.

Option C places emphasis on the songs themselves by displaying them front and center. This sacrifices some of the personal connection to a friend since the friend’s photo and name are less prominent. But, it takes up less space and prevents the app from becoming too much like social media. Still, a few people liked how B incorporated groups and focused on the friends.

I iterated on feedback from users and created two more explorations of the horizontal scroll.

I chose A to focus on easy access to music

For A, I enlarged the images for visual consistency with the rest of the home screen. Then, in B I explored categorizing the inbox by friend rather than showing all of the suggested songs.

After these iterations, a majority of people still value the music over the friends who send it.

Merits of Option A:

  • Distanced from the typical inbox/chat format. You can send recommendations without needing to provide a message for context.
  • Suggestions are only one click away from the home screen.
  • Enlarging the image emphasizes the music and is consistent with Spotify’s visual design system.


  • Less personal, since suggestions aren’t grouped by the friend who sent them.
  • Fewer songs viewed at a time

Content and Visuals

An audit of the different visual treatment for different information hierarchy

The home screen content all follows a similar hierarchy of information (above). To display recommended songs, I modeled the visual design similarly to podcast episodes, because of the similar amount and type of information needed. I laid out the potential and essential content requirements and visual design that I decided on here:

I decided to keep a rectangular album cover, song name, and artist name to make the music-related information clear and highest priority.

The sender’s photo layered over the album cover provides an easily recognizable way to see the user’s friend who suggested it. Though this breaks from the current design system of Spotify, I felt a different visual treatment from other design patterns is needed to emphasize that these suggestions come from a different source.

With the basics down, I decided to explore other ways to improve the listening experience.

Bring back the playlist!

This allows for:

  • playing the recommendations in order
  • condensing the recommendations in a list format
  • searching and filtering through recommendations.

The two different views allow for (1) a visual, at-a-glance inbox and (2) a more organized and condensed playlist.

Now that I settled on a way to discover and receive recommendations, I explored the sharing flow.

Fitting in a new way to share

The top-level icon in A makes the feature easy to discover, but also makes the action impossible to access unless the song is currently being played. On the other hand, D is hardest to discover but fits with how people are used to sharing songs.

I decided that B or C made more sense since sharing natively in the app should be distinct from sharing externally to social media and more prominent. Ultimately, C accomplishes this well while placing the entry point higher in the hierarchy to encourage discovery and use.

Final flow for sharing song recommendations to friends

Visual Design

The typography, spacing, and colors I used


Looking Forward

This was my first UX case study, and it was a great learning experience. I hope to continue to grow in product design, especially in product thinking and prototyping. Special thanks to everyone who helped me with user research, brainstorming, and user tests, and to Cornell AppDev for the learning experience!

I am in no way affiliated with Spotify. This was a case study for Intro to Digital Product Design. I would love to hear any feedback, comments or critique!

learning @cornell 🌱

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store