my Initial assumptions
I conducted three interviews for this project, each 20-30 minutes in length, each user from a different demographic. I wrote out my questions in advance, but tried not to completely rely on them during my interviews.
My script included questions that began with who, what, when, where, why and how, such as “What type of device do you use?,” “What types of content do you consume on your device?,” "What apps do you use?,” “When and where do you use them?” and “How do you use them.”
“Why do you listen to music?”
— One of the most important questions I asked during my interviews
With my problem and solution set, I began sketching out ideas for how to make an appealing, highly functional streaming music player for users. Using a Sharpie, I put my ideas to paper. In the beginning, to match trends in my interviews, I designed the app to allow users to select music by activity or emotion. "Muse" was the original title of the project.
The onboarding screen includes prompts for "I Feel" and "I Am."
“I Am” Landing screen
The "I Am" landing screen includes a list of possible choices to tap on to create a playlist.
“I Feel” Landing Screen
The "I Feel" landing screen also includes a list of possible emotions the user might feel to select to create a playlist.
the media player
The song player screen includes common functionality and labeling and a favoriting option.
I showed my initial sketches to a user to get their initial impressions. There was a sense of the confusion about what the app would do for them.
simpler start screen
The onboarding screen is simplified to a "Start" prompt.
For new users, the onboarding process begins immediately, prompting users to add songs to their playlist.
When they add them, a pop-up prompts them to choose an icon to signify how the song makes them feel. This process goes on for the length of the progress bar.
After onboarding, the user is prompted to choose how they feel.
So, I simplified the overall scope of the app to traffic in moods only. The change gave me the opportunity to create an onboarding screen to prompt users to "Create Your Mood Board." It also allowed me to simplify the selection of moods from words to icons. I tried to make the start screen speak more to the nature of the how the app is designed to interact with users. I also went away from text and toward emoticons as a way to make selections. This input lead to a refined first prototype in POP.
media player redrawn
When they choose an emotion, a playlist that contains music fitting that category immediately begins to play in the song player.
reclassify a song
If a song doesn't fit they mood they choose, the user has the opportunity to reset the song to a different mood.
A revision of Screen 2. "How Do These Songs Make You Feel?" is more intuitive than "Create Your Mood Board."
types of icons
A revision of Screen 3. Icons are more identifiable as emoticons.
I sketched higher-fidelity versions of each screen for the prototype.
time for play
I tore pieces of paper to build in the functionality into the app.
Using pop, or prototype on paper
I made higher-fidelity sketches and uploaded them into POP on iOS. From there, I made my first prototype.
the onboarding screens
Further iterations were made to the sketches in their higher-fidelity versions. There were a total of six main screens.
the playlist screens
Icons and buttons were more defined from the earlier sketches in this first prototype.
Once the prototype was complete and functional enough to user-test, I approached two different users to ask if they had 15 minutes to try out a new app I was testing. Before testing began, I engaged them in a conversation about their habits and behaviors regarding streaming music. One was an avid streaming user, the other didn't use streaming services and preferred to listen to her own music library. At the start of each test, I asked them if it would be okay to record the conversation for testing purposes only, to which they agreed.
My second user test
My first user test
In observing the users during the testing and in the video recordings seen afterward, I got some very actionable feedback.
- In both tests, users responded with surprise and positivity to the home screen. Warm smiles. One user said, “It looks fun.”
- The language isn’t intuitive, and at times too technical. “‘Build?’ That sounds like work” was one of the comments made in reference to the home screen button you’d press to create a playlist. “Redefine” was also another work which users were slow to respond.
“‘Build?’ that sounds like work.”
— User Tester 2
- Some of the logos in the navigation were confusing. Some of the users didn’t press on buttons because they didn’t look like buttons or weren’t sure what they would do. This happened on the music player screen during the task which ask a user to change the classification of how a song made them feel.
- The feedback screen when changing the mood classification of a song was confusing to the users. They weren’t sure what was happening.
- Both users said they would use the app if available on the market. One user said, “It’s better than Pandora.”
“It’s better than Pandora.”
— User Tester 1
the onboarding screens
Based on testing, the call-to-action messaging was changed, and I cut out options on the "I'm Feelin'" screen.
the playlist screens
Beyond new messaging, a "New Mood" button was added and a selection pop-up on "Move" button made the re-classification process simpler.
What I Learned
During this project, I was very comfortable with sketching, design and prototyping. User research and testing took me out of my comfort zone, but from those challenges, I found opportunities for growth. I think more exposure to interviewing will help me sharpen those necessary skills and methodology. I will opt to use a voice recorder on any interviews in the future, as the process of taking notes during an interview can interrupt the pacing of a conversation.
After testing the prototype with a user and building another iteration, many new questions arose. Would users want to have songs classified under more than one mood? How many choices of moods would users like before it becomes overwhelming to them? Would users want to create their own moods? Could I develop a signature moment, i.e. a branded microinteraction, that would help strengthen the app’s relationship with the user? Answering these and newer questions would drive further momentum on prototyping and testing new iterations of this app.
Overall, I think I identified and solved a problem that the users I interviewed are experiencing. Moving forward, I would continue to test and iterate to make the app better.