This is a ~5 hour design challenge I did recently as part of an interview process.
Finding new music is both easier and harder than ever. While there are seemingly limitless libraries available for streaming at any time — it can be incredibly overwhelming.
Research & Initial Thoughts
I started by looking into how Apple Music and Spotify structured recommendations, following artists, and players.
My favorite pattern so far for discovering music has been the combo of Spotify’s Discover and long press preview features. Being able to quickly get through previews in a list that’s already curated for me was often a quick and valuable experience. Spotify has removed this feature sadly, but I always found it very useful.
Apple Music tried to push following artists heavily during onboarding, but from a more social feed based approach. I’m definitely interested in connecting with artists, but don’t find the additional social channel very useful. It also isn’t used much by artists themselves. Spotify allows artists to highlight things, but it isn’t heavily promoted.
Questions/Thoughts
Whenever starting a project now I like to do a dump of everything I’m thinking through and questions I might have. When I haven’t done this I always forget something I wish I hadn’t.
Ideate
After looking into other patterns I started thinking about the most enjoyable or useful patterns I’ve used in the past. I explored using a combination of Spotify’s long press preview pattern and indicators for how likely one would be to like a track or album. These indicators could be potentially populated by:
Popularity among people you follow and artist recommendations
Number of plays by following X recommendations by artists you follow
Exploring recommendation indicators, long press preview, and collections
Exploring how to add or dismiss
Initial Library Import/Export explorations
While I found the concept of indicators to help to get more context more quickly very interesting, it felt overly complex and cluttered. I wouldn’t completely disregard the pattern, but I think it would take more time to make it a good experience.
Define
The more I thought about creating a music platform with recommendations I started to realize the gravity of it having to compete with Apple Music, Spotify, and the countless other competitors. So instead of creating a product focused on being your streaming service and music discovery tool all in one — I decided to focus on solving the core problem and building on the environment that exists.
There were a few problems I identified that I thought could be solved by a separate tool:
Managing libraries when using multiple streaming solutions
Listening to a song doesn’t mean I’d recommend it, but it gets recommended to followers
Issues with shared accounts
Other people in a house might play music off of the main Apple Music account on an Apple TV
Those plays change the main accounts recommendations to something the account owner might not be interested in
There could be a way to recommend music with more intentionality. Also, if this product is where all of your recommendations live then it might be the best‘source of truth’ for your library. If you have multiple services it would be amazing to link these into this product and export wherever you want. While I’m not positive of the technical feasibility of some of these links, it would be great to leverage other services like Ticketfly to import data from shows you attend or use Amazon vinyl purchases as recommendation data as well.
Job/Goal statements
To make sure I was thinking through the perspectives of users trying to accomplish specific goals I wrote out job/goal statements to cover the problems I wanted to solve for.
Explore
I started exploring different information architectures — especially around the makeup of the tab bar. This helped me to think about how much architectural weight was given to each piece and how they interact with each other. Seeing them all together helped me to compare different approaches.
From here I started to explore visualizations for players and ways to quickly add something to your library or dismiss it. I was really interested in combining long press preview with the concept of a collection that you could drop tracks, playlists, or albums into. The collection could live as a temporary home for items until you either add them to your library or get rid of them.
Research & Initial Thoughts
Questions/Thoughts
Ideate
Define
Job/Goal statements
Explore