The Modern Approach to Discover Music

Overview Steps

Domain research
Define problem
Define product purpose
Prioritize strongest to least important
Simplify roadmap
Increase impact


User Experience
Product Strategy
User Interface
Visual Design
Frontend Development (swift)


1 Designer
2 Engineers


After leaving SAP as a Sr. UX Designer in 2015, I joined forces with my friend, AJ Asver, founder of Just Spotted (acquired by google) and Julien Gaillard (Senior Data Scientist at Twitter). We all shared the same common problems with music discovery.

We started with a high level discussion about what was out there right now and we knew there were several existing music apps out there but we felt there was a missed opportunity in some very key areas for music discovery, sharing and most importantly, in the relationship we have with music.


SoundCloud and Spotify seemed to be the primary source for streaming music from our experience. Songza, Pandora and iMusic were next in the charts when you wanted a more curated radio-like experience. They were all different in some ways and it felt like they were all scattered features under multiple names but they all had one common experience, no emotional connection. Each app made me feel like I was searching an endless list by a machine and I was wasting more time browsing than enjoying myself. The list goes on. There were several missing features we saw across multiple apps. We began discussing and brainstorm solutions on whiteboards and define our next steps.


How do we understand the value for each user?
How can we learn the emotional complexity of the user and how do we monitor that so we can provide better recommendations?
Do people want to choose their mood or not?
How can we improve engagement?


Primary: Delight users with music they will enjoy that enhances their current emotions or experiences.
Keep users coming back with >50% success rate with recommendations.
Understand what music users like to listen to, when and where.

Secondary: Find a way for this app to provide distribution / revenue for new artists.


Interface: New songs delivered to user throughout the day in their track stream
Only two screens in app: Track stream where new music appears everyday.  Looks like a CD/Vinyl stack or carousel. Archive where tracks that a user hearts are stored. Songs can only be skipped by flipping to the next song. All music plays are done on user’s existing music apps (Youtube, Spotify, iTunes, Google Music etc.) Start with Spotify?
Users can provide feedback on songs by doing some kind of gesture while the song is playing. Swipe right to “heart” the song and keep it in their library. Swipe left to remove from feed.


From napkin drawings to moleskine wire-framing. I make sure I get everything down on paper before transferring them to Sketch or Illustrator. I executed journeys, wireframes, prototypes and design specs.


Wireframes were essential and important to make quick iteration and test new features. Then I was able to build a working prototype that we were able to get in front of people to get some early user feedback. Throughout the project I worked closely with engineers who implemented all my designs.

User flows

After defining our user cases we started building our user flows. We started with white boarding but eventually moved to Sketch to flush them out in more detail. Building on our own gave us a lot of freedom, but it also meant we were going to have to really prioritize the features and cut out whatever we thought was unnecessary. So we paired the features and flows down to the most essential key areas

Beta Demo

Lessons Learned

I had an incredible experience working on this project, it really showed me the value of working with a 1:1 ratio of designers to engineers. Building something from the ground up, with limited resources really forces you to evaluate the things you need most, and to solve problems as simple and efficient as possible.

Building a product from scratch can be more complex and overwhelming than most people think. For starters, you start off with no pay and from day one you're already burning cash that you don't have so it's wise to build your product as quickly as possible and launch! I also understood the value of team building and direct communication. When you work with a large company it's easy to hide your emotions or work in your own pace, but when you work with a small team, it's not that easy. On top of the daily constraints, you also have to balance your personal life and its not easy when you work with a small team. Always sync up with your team if you have any blockers or any questions. I was shocked when I realized big companies make you socially awkward or people don't know how to communicate when there's an issue. So much time can be saved if we respond to any concerns immediately, both personally and professionally.

Try to sit right next to the engineer(not across from you or in the same building but literally right next to him/her), clear, and direct communication is key when you work in small teams. You will save so much time rather than emailing back and forth. Make sure you have daily stand ups with your team so everyone is in sync. Productivity will increase significantly. I promise.

The Office