Spotify: Re-imagining the Blend Playlist Experience
PRODUCT DESIGN • UI DESIGN
Overview.
Spotify Blend playlist merges two users' musical preferences into a personalized playlist. To Blend, users invite a friend and Spotify's algorithms analyze their listening habits to make a unique playlist that is updated daily. The playlist also offers insights on musical compatibility, making it a fun way to discover new music with friends.
Specs.
Project Type.
Research, Experience Design, UI Design
Tools.
Figma, Notion, Zoom
Team.
Kaila Lee
Duration.
Dec. 2022 - Jan. 2023
Problem.
The Blend Playlist curates music that is not emblematic of user’s musical tastes. While the algorithm analyzes your listening habits, it can pull songs that you may have listened to just once or twice. This can cause the Blend playlist to sit in a user’s library and go un-listened.
PROJECT FOCUS
How might we create a compelling playlist experience that is automated yet emblematic of user’s shared musical tastes?
Solution.
To improve the accuracy of the Blend Playlist and create a personalized space to share listening habits, I introduced:
Blend Onboarding
Top Artist and Song Highlights
Song Replacement
Adjacent Song Suggestions
HEURISTIC ANALYSIS
What’s the current flow?
As both a product designer and avid Spotify user that frequently creates playlists, my initial step was evaluating the flow for creating a Blend playlist to highlight strengths and weaknesses based on my design intuition.
Flow: Creating a Blend Playlist
Flow: The Blend Listening Experience
DISCOVERY RESEARCH
Playlist Patterns
To gather feedback from an outside perspective, I spoke with 6 Spotify users to better understand their own playlist use.
🖍 Custom Playlists
The users I spoke with fell into two categories. The first set of users (4) preferred highly curated and organized playlists. Some even have a playlist for each month of the year.
🙉 1,100 Songs
The second set of users (2) I connected with had long playlists between 300-1,100 songs. This demographic was not concerned with having organized playlists and relied heavily on search.
COMPETITIVE ANALYSIS
Evaluating alternative approaches.
To better understand the problem space, I conducted research on a variety of product experiences from tax software, to design suites, to mental health tools.
Guiding Question: How do current software experiences tailor content to unique user needs and wants to provide accuracy and relevance?
SYNTHESIS
Opportunity.
🙋🏻♀️
01. Curation through Onboarding
Users would benefit from a quick onboarding experience where they can select criteria that will curate the songs selected for their Blend.
💿
02. Song Replacement
Once a user has created their Blend playlist, it would be optimal to be able to replace songs that may not reflect their listening habits.
✨
03. Adjacent Song Suggestions
To keep the Blend listening experience spontaneous, Spotify can suggest songs in adjacent categories for users to add to their playlist.
Hypothesis.
By offering users the opportunity to indicate their music preferences before the Blend playlist is created, the playlist is more likely to reflect accurate listening habits. Providing song replacement will help to overcome the inherent unpredictability of AI-generated playlists but maintain the personalization of this feature.
Quick Iteration.
After creating a user flow diagram, I began to visualize the intended flow for a better curated Blend Playlist. I experimented with different playlist page layouts and onboarding screens. I sketched them out during a Crazy 8’s exercise.
Lo-fidelity wireframes.
Mid-fidelity wireframes.
FINAL ITERATION
Design
When refining the Blend playlist experience, my goal was to create an area on Spotify that feels accurate, personal, and compelling.
Curated Onboarding
The Blend Listening Experience
Smart Song Replace
To improve the accuracy of the Blend listening experience, I create a two-step onboarding experience that asks users for just two genres they would like to share and one song that reminds them of the user they are Blending with.
After two-step onboarding, users will have the ability to continuously revise their playlist. Although Blend is regularly updated, providing users with the ability to replace songs they don’t prefer and add in adjacent song suggestions will create a listening experience personal to users and their friends. The Shared Top Song and Artist story is a compelling part of the Blend experience, providing users with a glance at their shared listening habits.
In an instance where the user wants to remove a song, they can replace it from the Song Actions List. The song’s replacement is generated by Spotify’s algorithm, which is continuously improving how it understands the user’s preferences.
Shout out to Jacque for chatting with me about this project for over 2 hours! 💕
User Interviews
To gauge attitudes towards this an optimized Blend Playlist experience , I interviewed 3 participants to determine the impact of these changes to the power playlist user.
Structure of the interviews:
About the interviewee and their playlist use
Prototype feedback
Feedback.
Getting feedback from power playlist users was overwhelmingly positive. Some users were curious about the replacement feature, suggesting it may be an improvement to allow simple delete.
☀️
“I wish Blend actually worked like this. My girlfriend and I just make each other playlists, but if this worked properly I’d use it right away.”
4.5/ 5
average rating on an improved, more personalized listening experience.
NEXT STEPS
Reflection
I really enjoyed working on this project! I’m an avid Spotify user and create a new playlist every month to capture my experiences. I even create my own artwork for each playlist I make (yes - I know this is impossibly nerdy!). Music is a big part of my day, and being able to share that with others is exciting. My hope was to create an area on Spotify that is personal and fosters closer connections with others through music sharing.
This was one of the first times I recreated an existing app’s UI. This project offered an opportunity to further improve my UI design skills.
If I continued working on this project, I would like to:
1. Consult with SWE to gauge the feasibility of my proposed solutions
2. Address edge cases
3. Collaborate on a second iteration of this with my best friend, Ramyani, who is a super talented designer 💖