OVERVIEW
The Challenge: Create a mobile app that will allow users to to have a perfect music experience.
Outcome: A mobile app that facilitates a music service which adapts to each user.
MARKET RESEARCH
I conducted market research to become familiar with the main music apps that are already on the market. I performed an analysis to understand what these apps were tackling and what they were lacking, what they had in common and their differences.
| SERVICE | PROS | CONS |
|---|---|---|
| PANDORA | - Gives music recommendations based on the music you like - The app learns as you use it - There are pre-made stations based on moods, activities, decades, etc. - Easy to use and intuitive interface |
- Limited free version (no music recommendations, lower audio quality and no desktop client) - Only available in the US - You need an account to use it |
| SPOTIFY | - Gives music recommendations based on your listening history - There are pre-made stations based on moods, activities, decades, etc. - Playlist creation and collaboration - Easy to use and intuitive interface - Works in multiple devices |
Limited free version (lower audio quality and advertisements) You need an account to use it |
| LIVEXLIVE | - You don’t need a user account to stream music - Multiple pre-made stations |
- Limited free version (lower audio quality, limited skipping number and advertisements) |
| GOOGLE PLAY MUSIC | - Personalised music recommendations based on genres you like - You can upload your own music |
- Displays advertisements - You need a Google account to use it - You need to buy music in order to listen to it |
| PLAYLIST | - Over 40 million songs to choose from - You can make your own playlists or access pre-made playlists |
- Must log in with Facebook or mobile number - Works only on iPhone |
FINDINGS
- There is a great number of music streaming apps in the market with similar characteristics
- Most music apps require an account to use the service
- Most of them have a limited free version
- Many have pre-made lists and stations based on different subjects
- Many give music recommendations, making discovery easier
INTERVIEWS
To understand better the mindset of the target users I interviewed 7 people of different ages and backgrounds. After completing the interviews I wrote down all of the responses and began to assemble them based on topic.
By organising the interview responses, I found out trends and patterns that showed insights about the user experience. Through this process I singled out some user needs and observations:
- Users give great importance to music recommendations by friends and family
- Users enjoy the custom music recommendations on some of the services they use but they would like to see an improved version of the recommendation system which correlates more with their music tastes
- Users see music as both a social activity and a solitary one, depending on the mood and the situation
- Users agree that they usually prefer to listen to music from certain genres and decades, although in most cases not exclusively. At the same time, most of them have some music genres they would never listen to
- Users enjoy finding out new music and, in some cases, will even pay to discover new artists
- Users don’t appreciate that the advertisements that come with the free version of some services shows music that they would never listen to due to the dissimilarity with their music tastes
PERSONA CREATION
After this, I used all of the data I collected during the process to create two personas: the leisure music listener and the invested music lover. I found that both users have different needs and motivations.
INSIGHTS
I wrote down a number of insights and questions to come up with a list of solutions which would give me an idea of the service that I would be creating.
| INSIGHTS | NEEDS | QUESTIONS |
|---|---|---|
| Users give great importance to music recommendations by friends and family |
Users need a social side to their music experience |
How do we improve the social side of music services? |
| Users enjoy the custom music recommendations on some of the services they use but they would like to see an improved version of the recommendation system which correlates more with their music tastes |
Users need an improved music recommendation AI |
How do we improve the music recommendation AI? |
| Users agree that they usually prefer to listen to music from certain genres and decades, although in most cases not exclusively. At the same time, most of them have some music genres they would never listen to | Users need to be able to choose which genres and decades they want to listen to and which ones they don’t want to see |
How do we organise music so it doesn’t show genres the users don’t want to see? |
| Users don’t appreciate that the advertisements that come with the free version of some services shows music that they would never listen to due to the dissimilarity with their music tastes | Users who don’t pay for a subscription need targeted advertisements which do not disturb their listening experience |
How do we target music advertisements based on the user’s music preferences? |
Next, I tried to come up with a number of solutions and features for each one of the questions:
How do we improve the social side of music services?
- Connection with social media services to share with friends the music the user likes and listens to
- A community who enjoys music and shares their views and knowledge
- Option to ask friends or the community for recommendations based on different subjects: similar to an album, similar to an artist, within the same music genre, etc.
- Collaborative playlists
How do we improve the music recommendation AI?
- Other than just based on what the user listens to, give the option to answer questions about likes and dislikes to improve the service
- Make it learn from what the user listens to and how their music taste evolves, deleting suggestions based on music they don’t listen to anymore
- Create an improved music database and an AI which understands the similarities and differences between artists and genres
How do we organise music so it doesn’t show genres the users don’t want to see?
- Possibility to create custom playlists
- Option to block certain music genres, decades, etc.
- Option to have your music organised within different systems and categories
How do we target music advertisements based on the user’s music preferences?
- Add an option on the settings to allow targeted music advertisements if on the free version of the service
- Give an option to choose which music genres the user wants to be advertised on and which ones they don’t
INTERACTION DESIGN
After this, I created a site map for the app that outlined the navigation and information hierarchy.
I then completed the task flow and user flow to imagine the ways a user could use the app to discover new music. This helped me make sure that all the information was organised in the most intuitive way for the user.
TASK: Find artists similar to Joy Division.
WIREFRAMES
I started creating digital wireframes of some of the main screens. During this exercise I examined how the content could be structured to satisfy user goals.
USABILITY TESTING
I asked two participants to test out the prototype’s usability. The participants were given three scenarios with a task to complete:
- Find artists similar to Joy Division
- Find a saved album
- Find a playlist from the 80s
Both users were able to complete all tasks.
VISUAL DESIGN
I then moved on to visual design. I started brainstorming and looking for inspiration for a brand identity, colour scheme, logo and name for the app.
I produced a style guide with all the UI components and patterns.
Finally I applied the style to create the high-fidelity wireframes.
Please view the final prototype in this link: PROTOTYPE