Napster's Listener Network.



The subscription music app Napster, formerly Rhapsody, is a revolutionary way to browse and search through millions of songs at your fingertips. you can play anywhere, anytime. We were always searching for ways we can make it easier for our users to find new music.





Art Director, UX Designer | 2016

3 Hero Screens of the Final App

Problem



Subscription services rely on their customers to stay engaged. When a user cancels early after signing up, it becomes a high cost for the company. We wanted to find out why users cancel and come up with a solution to keep them engaged and likely become a lifetime member.

My Role & Process


Full-time Art Director, UX, Product team. My direct impact was Research, Product Design, and Marketing.

Project Length: 1 Year
Team: 3 Members
Role: UX Design, Art Direction
Tools: Sketch, Adobe CS,
Invision,

  • one | Information Gathering Created with Sketch.

    This is where I go and find our best and worst customers. How can I leverage the behaviors of users who are happy with the product?
  • Two | Build a Persona Created with Sketch.

    Customer interviews and personas were developed including journey maps.
  • Three | Solution Created with Sketch.

    An algorithm was developed for a new music discovery feature. A new section was created in the user's core experience of the app.
  • Four | Conclusion Created with Sketch.

    After completion, I tested with 4 developers and 3 designers for qualitative feedback

Information Gathering & Research





The data was the first place to look for clues. Find the users who were not only loyal customers but people who loved music. We realized the users who had a higher play count were spending more time on the App. We also saw that the users who canceled spent less time.

After conducting 10-day phone interviews, we came to the conclusion that discovery was most important to everyone. We also recognized that the users who played new songs more tended to cancel less. They also worked harder for their music. These members were "Feature heavy." They searched, they played more public playlists. They played more radio. We concluded they were more independent than the other members. They spent their free time looking for music while the others searched less and relied on regular radio or Pandora.

From our interviews, we created three types of users:

Influencer

  • Played at least 20 new songs per week
  • Spent 7+ hours per week discovering new music on the internet, with friends, and physical music retailers.

Casual Listener

  • Played at least 5 new songs every three days
  • Spent less than one hour per week discovering new music on the internet, with friends, and physical music retailers.

Passive Listener

  • Played less than 1 new song per week
  • Spent 1 hour a week discovering new music on the internet. Listens in the car 8+ hours per week.

We realized that the Influencers were a unique group. All of the members equally cared about music, but the Influencers were obsessed with looking for it. The Passive Discoverers said they'd listen to music more, but they didn't have the time. It was interesting because all of the members, we tested, said they had "busy lives." The Influencers just found the time.

"I look for my music everywhere!
Sometimes, I read about bands on my phone in meetings."
~ Music Influencer
Mapping the Journey

Here is an example of a Journey Map for the Influencer. One of the behaviors that stood out from the other users was this group took the time to research music independently. This later became an asset for our ideation.

Solution



We concluded that we had to find a way to expose the music from these hard working Influencers to the more passive users. If we could find a way to match them together, we could provide more music and generate more plays!

1. Find a Match
Match users who were listening to the same songs, albums and genres. In this example, match all users who listen to Bob Marley.
2. Find the Similarity
Next, we want to find a second common artist between an Influencer and a Passive Listener. In this example, these two users listen to Kendrick Lamar and Bob Marley, the other users who don't listen to these two artists are eliminated from the match.
3. Matched
We kept repeating the process until we find a match of at least 80% of the same songs played. We have our match!
4. Provide Music
Now we had songs that were potentially likable by our passive user. Since these two members were matched, we assumed they would enjoy the other songs the Digger was listening to.
5. Pass New Songs
We now had new content to offer our Passive user.

Implementing the feature was surrounded by the user. Profile pages were the center of the user's experience.

Home Screen (Left) Trending Screen (Right)

Conclusion



The project was implemented into the platform and was received with great success. Retention increased by 1.6% after 3 months for the pilot program.

Product Marketing Video
"Listener Network is a smart blend of human curation and machine learning, and uses Rhapsody's existing Music Intelligence Engine to power the entire experience."
~Contributing Writer, The Next Web
Microsoft UI Fabric | A.D. | 2016
Coming Soon

I'd be happy to have a thorough conversation about my projects.
if you have a project you need help with, contact me at:



paul@paulrileycreative.com


Inspired by music, fueled by coffee, love to cook, and spend time socializing with incredible people.