YouTube Music — UX Enhancements
Through this UX study, I demonstrated a logical enhancement on handling playlists and using the app's algorithm to improve that user experience.
Tools used
- Figma — High-fidelity design tool
- Pen-paper & FigJam — Early concepts and Low-fidelity design tool
- MailChimp — Surveying tool
Observations
Current scenario
People have defined different music for different activities, such as driving, workout, chilling and relaxing, etc.
To achieve this goal, people have the following methods:
- They create separate playlists for that different activities.
- They rely on a defined category / mood-based list created by the application.
- Or they use a combination of the above methods.
But there are specific problems associated with each process. Let's discuss them.
Problem with playlists
- Creating and managing these playlists is a cumbersome task.
- It also requires constant interference of users to keep the playlist updated with newly discovered music tracks.
How categories solved this problem
To solve this problem, Apps introduced category-based music, which resonates with the most prominent activities for a user.
Each category creates multiple auto-generated playlists that include tracks that will likely belong in that category based on criteria defined by the app.
For example — Rock music for the Workout category and Blues, Jazz music for the Relaxing variety, etc.
So….Problem solved?
Yes, mostly. It allowed people to listen to music on their terms and current activity.
But there's still one small problem with this category feature,
- The app still defines these categories,
- And the rules created for categories are also fixed,
- And, people can't change that even if they want to.
But before we discuss the problems and solutions, It's essential to identify if the general audience agrees with this thinking or not.
Quantitative Analysis
I started a small survey to understand people's music-listening behaviour and how they correspond to certain situations.
This survey will also help me find common ground between my hypotheses (shown below) and the actual scenarios.
Survey Design
- Tool Used: MailChimp
- Total Questions in Survey: 18 questions
- Entire duration for the survey: ~ 3 weeks
- Total Respondents: 418 respondents(Statistically significant)
An overview of results
- The primary audience consists of people between 18–30 years old.
- More than 60% of the respondents use Spotify as their primary listening platform, followed by Apple Music and YouTube Music.
- 55.7% of respondents use a combination of app recommendation and their search methods.
- 3/5 (Average) — On the ability of discoverability and finding new tracks, the respondent rated the app's recommendation.
- 66.7% of respondents create specific playlists for different activities.
- 66.4% of respondents rarely encounter unwanted music during playback.
Comparison with actual Hypothesis
Before starting the project, I created hypotheses of — What I believe is the problem.
And the data I gathered from the surveys helped me verify my thoughts against hypotheses.
Here is the list of hypotheses I devised and the result of the survey:
- Hypothesis #1
People use music streaming services for their music needs.
Result: True
- Hypothesis #2
Most people use the app's recommendation engine to explore new songs.
Result: True with conditions
Conditions: I thought everyone was using the ONLY recommendation engine to explore, but on the contrary, people use combinations of both AI and self to achieve the task.
- Hypothesis #3
People segregate their music based on their activities — Walking, Exercise, Traveling, Focus, etc.
Result: True
- Hypothesis #4
For specific use cases where Categories don't work as per people's needs, People can make different playlists that serve the segregation purpose.
Result: True
Out of those who use playlists — 50.5% create new playlists every six months
5.8% of respondents rely solely on the app’s recommendation
- Hypothesis #5
They are rigid, and people need to add the specific music track as per their needs.
Result: True
- Hypothesis #6
People don't update their playlists regularly but let recommendation engines add similar music to the following music tracks.
Result: False
- Hypothesis #7
The process of updating a playlist is complicated and tedious, to say the least.
Result: True
- Hypothesis #8
The difference between Playlists and Categories is the content adding flexibility.
Result: True
The problems identified
I converted the opportunities into goals for this project to enhance the application's functionality.
- Rigid Categories
Since categories defined in the application are fixed and rigid, users can't update the rules or create a new one. - Tedious Playlists Updation
People unanimously agreed that they need to update their playlists as soon as they discover a piece of new music that satisfies their criteria. The process to update an already defined playlist is itself a tedious task. - Juggling multiple playlists
People agreed that they have to create multiple playlists independently depending on their activities, and handling all of these playlists is a difficult task.
Opportunity discovered
What enhancements/extensions can I make to available features in the application to create playlists more flexible, captivating and easy to manage.
Goals for this project
I converted the opportunities into goals for this project to enhance the application's functionality.
- Rigid to Flexible
The categories need to be flexible enough to accommodate a person's changing taste and update the playlists based on that change. - Tedious to Captivating
Categories will continue to create auto-generated playlists that can update themselves based on new rules and people's listening habits. People won't have to add new tracks every time into rigid playlists. - Juggling to Easy to manage
The new feature will create and manage multiple auto-generated playlists that resonate with the flexible criteria following the application's algorithm, thus eliminating the various playlists handling phenomenon.
The Solution
What If we devise a newer feature where each person can create high-level custom categories based on their own criterias.
And let the app create multiple playlists based on these new criteria defined by the user.
This way, users don't have to deal with multiple custom playlists that had to be updated by the user.
Instead, they can just define what TYPE of music they want to hear, and the app's recommendation engine does all the heavy lifting.
Designing Phase
With the goals in place, I needed to visualize the working of this new enhancement through data-flow diagrams.
Data Flow Diagram
Through this flow diagram, I visualized the flow of the data and defined the high-level steps of the process.
Proposed steps
- User clicks on — Add New Category.
- Enter the category's name, Image associated with the category — which will be displayed as a banner for that.
- Select Artists / Genres of music to include in the category
- Confirm the selection
- New Category has been created.
Low-Fidelity Design
Run these designs on your own device (Opens in new tab)
These low-fidelity designs helped me visualize the virtual interface and the placement of the action buttons that resemble the app's existing UX.
Introducing
YouTube Music's custom categories
You can now run this app on your device
There is a need for a mechanism that lets people create their own categories instead of playlists and define the rule about what type of tracks/genres/artists they want to add to that particular category.
These new categories will auto-generate playlists containing tracks based on the people's selection. It will update itself through the app's recommendation engine based on the user's listening habit.
The choices
Users can create custom categories to define the rules about what type of music they want to add to that category.
Currently, selection rules can be defined as
- Genre selection / Multi-select
- Composers or Singers selection / Multi-select
- If the track in the category contains lyrics or not / Toggle
- If the user wants the app to update their recommendations / Toggle
Ofcourse, these rules can be enhanced as well, providing a full-control to users and their listening preferences.
Upon successfully creating the category, it will add itself to YouTube's own defined category list.
Final / High-Fidelity Designs
You can now run this app on your device.
These are the final designs that were built upon the low-fidelity design.
For this project, I am focusing on building this feature for YouTube Music, but similar concepts can be replicated for other apps; the UI and the interaction pattern would be similar to those apps.
How it works
Step 01
Users can start the journey by clicking on Add New on the top bar of the application.
Alternatively, The User can go to the newly created My Categories section in the Explore section to start the journey.
Step 02
Now, the user will select the type of music they want to include in the category.
They can select between Artists and/or Genres, which would be multi-select.
This function works similarly to the YouTube Music current selection that it offers when a person installs an app for the first time.
Step 03
Other settings include the name of the category.
Users can now configure some more details about the particular category. Users can define their custom banner for this category; if the input is not provided, Google's recommendation system can take over and determine a banner for this category based on what type of music it belongs to.
Apart from these, we have specific settings which users can use to finely tune their music selection:
- Exclude tracks with vocals
Through this setting, users can opt-out of the music containing lyrics. - Improved music recommendation
Which will use Google's recommendation engine to further enhance the app's listening experience.
….and you are done
Congratulations, A new custom category has been created.
Users can access this category through the top bar.
Wait! There's more to "explore."
As an extension of the feature, I've added the My categories section in the "Explore" section of the application, where the user can see all of their created categories and choose to add/delete categories.
That is where the user can edit the Category settings, including Banners, Name or the selection of artists and other settings.
Areas affected
These areas needed to be changed/updated to accommodate the newer features.
- Home Screen
The CTA and the created categories will be accommodated on the top-strip of YouTube Music.
2. Explore Screen
In the Explore part, a new option will be added, "My Categories", which will contain all the user-created categories.
What If….?
We likely need to address some of the scenarios to handle all the moving use-cases.
- What If the user wants to change the rules defined for a category?
Users can also choose to edit the rules defined by the category anytime, and updated rules shall be reflected on the app in the form of newly updated Pseudo-playlists when the app loads next time. - What If a user adds custom music to any active pseudo-playlist created by the custom category?
Users can add custom music to any active pseudo-playlists, an ongoing feature of all music players.
But if the track doesn't lie in the Genre/List of composers defined by the user earlier for the category.
The rules will be updated, and they will be modified — on the run — to accommodate this new selection. - What If a user deletes a custom category while a pseudo-playlist from that category is playing at that instant?
It will not affect ongoing playlists instantly; category entry will be deleted from the Top-strip of Home page and from the "My category" section.
Users can still save the ongoing playlists as custom playlists, which will not be updated dynamically upon changes to the list.
Closing Thoughts
This project was an exercise for me to enhance my design skillset, learn new things, and solve problems.
I've learned
- How to effectively use surveys,
- What care goes into designing one,
- What factors do we need to consider when sharing or expecting answers from someone, and more?
I've also learned
- What care goes into creating an extension of the feature, rather than creating one from scratch,
- What guidelines do we have to follow and
- We have to keep this in mind while creating an experience for the user.
It would be really awesome if these companies saw this case study and made this feature a reality; that'd be really cool.
If you can implement this functionality or know the people who can, Please connect with me, and I can elaborate even more on the features and…can develop a better UX for these (I am a User Experience Designer, after all)
It would certainly solve my problem since I am a user of these applications.
Thank you
…for reaching here, and If you liked what you've just read, there's more where that comes from.
And, if you have any constructive criticism on my writing and what can I do to improve, please share your thoughts as well. You can view my work on my website: www.ankitpassi.com
