/ Product Design Masters Course Case Study, 2020
/ UX Research
/ Product Design
As part of my course projects at Parsons, I had the task of designing a new feature for a mobile app of my choice. I worked on creating a feature named "Scheduled Playlists" for Spotify, with which you can schedule any playlist to play at specific times of day and join other live public playlists made by favorite artists and friends.
8 weeks (October to December 2020)
Ideation, user research, interaction and experience design, prototyping, information architecture
The social disruption caused by COVID-19 has increased levels of loneliness and depression, especially in young adults.
Research has found that there has been an increase in depression among young adults (aged 18-23) during the pandemic, with loneliness accounting for much of this increase. Per the Journal of Adolescent Health, targeting loneliness is warranted given this may have important downstream effects on mental health.
Remote work is here to stay on a semi-permanent basis, and it is blurring the boundaries between work and personal life.
The surge in remote work brought on by COVID-19 has blurred work-life balance, creating the dilemma of when exactly work time begins and ends. The shared work and play environments have disrupted existing lifestyles and daily habits by erasing prior environmental cues tied to our habits.
Despite the problems posed by social distancing and the pandemic, there are still opportunities created by the current circumstances.
People are online now more than ever. Are there more ways to feel other’s presence besides intentional virtual hangouts?
Many have access to a private work space, allowing freedom and control over their environments, including what they listen to.
What role can music play in our lives during these times?
How can we use music to reinforce current or build new habits as a means to adapt to our new environments in the context of COVID-19?
The Design Process
Defined jobs to be done
What are Spotify's jobs to be done? The main job it should do is allow users to listen to music. Some secondary jobs are categorizing songs into playlists or albums, and sharing or discovering music or podcasts through its AI-based algorithm or friend suggestions.
App audit , competitive analysis, & deep dive into user flows
I conducted a heuristic analysis for Spotify, and it was concluded that Spotify met usability standards consistently, though could use some improvement in its flexibility and efficiency of use through more intuitive swiping or tapping gestures. I also did a competitive analysis on other music streaming apps such as Apple Music, Youtube Music, and Amazon Music. Finally, I outlined the user flows for each of the jobs to be done to map out areas of opportunity.
I conducted 8 generative interviews with Spotify users and non-users, with the purpose of understanding user pain points and desired improvements or features. I found that the most common pain points were:
The custom-made playlists are often hit or miss
The playlists don’t venture too far out from one’s taste, which may be a downside for people wanting to experiment
Users are unable to find low-key songs, covers or unreleased music on Spotify
Users need to tap “like” on a song to save it to listen later, which mixes these songs with songs one actually likes
I initially started exploring the problem space of the "Liked Songs" playlist becoming a messy area of songs that are saved for later and actually liked by the user. I began to ideate a couple of solutions, such as by creating a staging area prior to the liked songs, or by allowing the user to sample a couple seconds a song to assist them in rapid judgment of liking the song.
I decided to change directions and rethink the feature after my generative interviews, and was particularly inspired by one person who mentioned:
“ I always play the same playlist while driving, and when I hear this specific part of the song I know it’s time to change lanes”—Interviewee
Research on habit forming
After hearing those words from the interviewee, I started thinking more about the emotional, social, and functional jobs of Spotify. I became highly interested the relationship between music and habits, and so I read Atomic Habits by James Clear to learn more about habit forming and cues that lead to habits.
Survey on habit forming
With the insights I gathered from the book, I created a survey with the purpose of evaluating common cues and exploring motivation rituals for building habits. I also asked people to rate their individual habits to get a look at current behaviors and areas for which music could play a role. Lastly, I asked about apps used to maintain/build habits and also asked about the alarm app to determine whether I wanted to use this framework.
Analyzing Survey Findings
I gathered survey results from 13 individuals and made sense of my findings by looking for data trends and organizing the information in affinity maps.
My top three insights from my research findings were as follows:
Music can be a motivation ritual
Some already use or imagine music as part of their motivation ritual for habit making because it is a pleasant, non-aggressive reminder.
Independence vs. Accountability
There are two main divergent preferences in habit making: some prefer accountability from humans whereas others prefer approaching it independently.
Users are interested in the social spaces of music
Due to social distancing and isolation, interest in livestream events, especially in music, has increased among Spotify users.
At this point of the design process, I had a vague idea of what I wanted to build towards — a feature that would take advantage of the possible behaviors and habits that could be triggered by music, in the setting of the pandemic. I started off my ideation phase by creating user personas to define use cases.
Primary use case:
An adult who has started working remotely since quarantine is struggling at drawing the line between work and life. Work hours have bled into his person life and he wants to mark the boundary so he can relax better in the evening.
I then conducted a competitive analysis of apps that have been successful at helping users build habits in spaces other than music. My goals were to understand trends in UI and UX that were used to help motivate users engage in a repeated behavior.
After looking at competitor apps, I sketched out different ideas and directions in which I wanted to take this feature, and narrowed it down to two divergent solutions.
Solution Idea One
A music alarm
Users will want to freely choose their own songs or playlists and set them to play at their designated times of the day. This allows for user independence and variety, while creating an automatic environmental cue that may be used for habit forming.
Solution Idea Two
Users will want to view curated songs or playlists that will play at set times, and sign up for them as they wish. This reduces decision-making but also increases accountability as a user can join a playlist that is being listened to by others at the same time.
I started off by drawing lo-fi sketches with pen and paper, exploring layouts for both of my two ideas. I grabbed inspiration from existing UI trends for alarms and live streaming apps.
Idea 1 Sketches
Idea 2 Sketches
I then turned my sketches into lo-fi wireframes in Figma, which I then arranged together for prototyping.
In order to assimilate my designs further to Spotify's look and feel, I increased the fidelity of my designs to mid-fi. I then prototyped my designs including move-in and move-out animations similar to Spotify's because I wanted to test whether the new feature ideas could be imagined in the context of the existing app.
In addition, I considered naming this feature "Soundtracks", as inspired by one of Spotify's famous phrase "Soundtrack your life". I thought the reference to a soundtrack could accurately represent the idea of having background music that matches the rhythm of your day or your mood.
Mid-Fi Prototype for Idea 1:
Mid-Fi Prototype for Idea 2:
Given I had two very different ideas for my feature, user testing was crucial to deciding the direction of this project. I conducted 9 interviews with male and female Spotify users aged 23-33, working in tech, healthcare, law, and design.
My goals for user testing were:
Explore reactions to my two divergent ideas
Understand user behaviors in the context of music, habit forming, and socializing
Determine usability issues to fix in order to increase satisfaction with the feature
Synthesis & Iterations
Key Findings for Prototype 1:
Initial homescreen is confusing on how to create a soundtrack because it looks too similar to the homepage
There should be a way to play the songs before assigning to a soundtrack
Users want to select songs from multiple different areas
Creating a time for a playlist to play is too regimented for some
Want freedom to control the order of songs in playlist
Need to clarify scenarios of what will happen after soundtrack ends
Perk for habit building is the specificity of the time and song selection for user’s intentions
"Soundtracks" may be associated with something more passive, less intentional
I gathered my user testing findings into affinity maps that captured three main categories: feedback on specific points of the user journey, suggestions on specific smaller features within the larger feature itself, and the users' overall impressions
Key Findings for Prototype 2:
Users keen to finding playlists and songs recommended by friends and artists they trust
Enjoy listening with friends together at the same time, as it creates camaraderie
Make scheduling options clearer
Chat may not be very useful to most, and can even lead to a feeling of being “spammy”
Make a personal calendar of soundtracks
Keep general list of soundtracks separate, not under your library
Overall more preferred than prototype 1
Also seen as effective for habit building when there is accountability or fear or missing out
Users would want to see both features combined
“Soundtracks” also not appropriate for some, feels more like “radio”.
In summary, the main insights I gathered from my research were:
Joining live playlists with friends felt more exciting
Most interviewees preferred prototype idea 2 because they enjoyed listening together with others at the same time, and this connection was a key highlight.
But creating one’s own schedule seemed more practical
Several interviewees expressed wanting both feature ideas combined because the music alarm idea seemed practical for habit forming in daily life.
The term "Soundtracks" was not the best descriptor for this feature
There was confusion on whether “Soundtracks” were different from any other playlist, so users thought it would be clearer if one could schedule
the playlist any time they want, instead of calling it a soundtrack. Introduction of a new term seemed hit or miss.
Make it social... but not too social
I discovered there was a fine balance between private vs social spaces on Spotify—users liked the social component but didn’t want it to bleed into areas like the chat because it’s a listening app, and not a video or social media app.
User Journey Mapping
After making iterations based on action items drawn from the key insights above, I merged my two ideas into one and upgraded my designs into hi-fi prototypes. I then tested them on 3 additional interviewees and drew out user journey maps to ensure I was understanding the user's emotions and frustrations throughout the main user journey.
One problem I ran into at this stage of the process was validating where to locate this feature in the mobile app. I initially had included the feature under the playlists section of the app in "Your Library", but then I realized that this feature should be more discoverable from the homepage and should also have a distinction from other playlists. After deliberation and feedback, I added a section for "Scheduled Playlists" in between the Playlists and Artists tabs and made the schedule accessible through both the homepage and "Your Library".
For the final iterations, I focused on smoothening out microinteractions, thinking through different states of the feature for each use case, and further polishing the visual design.
New Feature on Spotify:
A feature where you can:
Schedule tracks to play automatically & match the rhythm of your day
Join scheduled public playlists made by your favorite artists or friends and listen to them together in real time.
Feature on-boarding example
Use Case Scenarios:
Signing up for a live playlist
Signing up for a playlist that plays in the future
Scheduling an existing playlist in your library
Deleting a scheduled playlist
Checking activity tab
Playlist starts playing at designated time and user receives notification
Why would Spotify want this?
If this feature was, in fact, a possibility, and artists could create their own playlists on Spotify, I imagine the company would want this feature because:
It allows creators to feel more connected to their following and vice versa.
The input of human-based recommendations into an algorithm-based app could strengthen business-consumer relationships.
Overall, it gives Spotify a competitive advantage, potentially attracting more customers.
Social isolation is difficult and there is no substitute for human connection
This feature idea is only one positive way to approach the problem space of making up for the lack of physical human connection, but it is not a be-all end-all solution. Having music to match your daily habits may be effective for some, but cannot replace human cues that may have existed before, such as morning coffee runs or exchanged glances with coworkers at the end of the work day.
How can we extend the live playlist experience further?
I would like to think more about enhancing the live playlists for a richer experience that perhaps does more with augmenting the listening experience rather than what one sees on screen.
Things I would like to improve...
My current proposed feature is a combination of multiple smaller features. If I had more time with this project, I would want to narrow down my ambitions by going more in depth into one or two main features, and nailing down both the experience and visual language to be cohesive. I would also like to experiment further with the information architecture to make sure that this feature is placed under the most intuitive information categories.