YouTube Music Search Redesign
Personal Project
2024. 8
Context
YouTube Music is a streaming music platform developed by YouTube, which was acquired by Google. It was launched in 2015 and has over 100 million paid subscribers as of January 2024. It provides various resources, including songs, videos, podcasts, and episodes. However, the user experience is not as intuitive as that of other music streaming services, such as Spotify and Apple Music.
ROLE
Product Designer
PROCESS
User research, Journey flows, Wireframing, Prototyping, Iterations
TEAM
Solo
Challenges
How can we redesign YouTube Music app's search experience so that users to find the songs that they are looking for in a faster way?
User Research
In order to get as many data points as possible, I visited over 20 online YouTube music reviews and a number of public posts on Reddit. I collected 35 pieces of feedback and organized them into groups.
Key Insights:
Recommendation: Recommendation doesn't effectively help with music discovery.
Music search: Difficult to navigate and search for the music users want intuitively.
Local file support: Unable to upload music from local storage.
Meanwhile, I also looked into other music streaming software that is a direct competitor to YouTube Music, including Spotify, Apple Music, SoundCloud, and Amazon Music. It is crucial to understand YouTube Music's strengths and areas for improvement compared to its competitors before making any adjustments.
Comparison Analysis
Based on the takeaways, YouTube music can do a better job at the music searching and selecting user experience. Below are two example of the visual comparisons.
Genres Comparison
Search Comparison
Problem Statement
Based on the user research and competitor analysis output, I decided on the focus of this redesign project:
It’s hard for YouTube Music users to intuitively navigate and search for songs that they want.
Ideation
In order to break the problem into pieces, with the help of some target users, I identified 2 user cases: public song search and local library search. Then I listed out some potential improvements.
Ideation Process
Design
Low-Fidelity Design
I started by sketching the low-level flows of the search features and added some improvements. Since low-fidelity sketches are easy to iterate on or change direction, they are great for the initial design.
1. Local Song Search
For current YouTube Music users, if they download a lot of songs, it is very hard for them to identify the songs they want to listen to. To help users find songs easily, I added a local music search bar feature and a filtering feature.
Local Music Search Bar: Find downloaded songs within the library.
Filter: Filter local songs based on genres, languages, and moods, and then play a single song or several songs in the filtered result.
Local Song Search + Filter
2. Public Song Search
Many users have reported that the current YouTube Music search engine is not as intuitive as those of other apps. Since any users can upload content to YouTube, it can be challenging to find official songs. Additionally, the search results are often cluttered with a mix of songs, videos, podcasts, episodes, and profiles, which diminishes the app's focus.
Therefore, I proposed the following changes:
Redesign the public search flow to focus on content rather than autocomplete.
Differentiate official songs from other results.
Reorganize the search results to emphasize songs instead of videos, episodes and profiles.
Provide suggestions to assist users in finding and searching for songs.
Public Song Search
Mid-fidelity Flows Exploration
My next step was to create mid-fidelity flows. Mid-fidelity prototypes are great for user testings and iterations.
Local song search - Search by song name
Local song search - Search by artist name
Local Song Search - Filter
Public song search - Official and other results
User Testing and Iterations
I scheduled user testing sessions with 5 target users throughout the journey. Below are some questions I asked.
"Search for a song locally. The name is of the song is Love Story, and the artist is Taylor Swift. Play the song after you find it."
"Search for a serious of hip hop & rap songs and add them all to the queue."
"Search for love songs written by Taylor Swift in YouTube Music publicly. Play the song after you find it."
"Do you think the flow is intuitive? Will you use this feature?"
Here are some of the takeaways:
For local search, users prefer to search by song name rather than artist name. Overall, the flow is very intuitive.
Some suggestions on the search page is very helpful (different from the current Youtube Music version)
Users care most about if they can find the right song quickly. They don't care too much about the approaches.
Differentiating official songs can help with the song finding process.
None of users prefer the local music filtering feature. They are most likely not going to use the feature and search for songs in the same genres online instead.
Due to the unpopularity of the filtering feature, I decided to let go of the feature. Meanwhile, I made some iterations for other flows.
Local Search - Search Page
Local Search - Search Result
Public Search - Search Page
Public Search - Live Search Result
Final Product - High-Fidelity Prototypes
Below is the final high fidelity search prototypes made with Figma.
Redesign Comparison
Success Metrics
In the future, future user testing sessions should be scheduled and collected both qualitative and quantitive data. Meanwhile, below are some of the metrics to measure the success of the redesign.
User engagement - How much time users spend on the YouTube Music app compared to pre-launch period.
User retention rate - How many old users decide to continue using YouTube Music app.
Task completion time - How much time users spend on average to search for a song.
Customer satisfaction rate - How pleasing users are when they use the YouTube Music app.