Re-imagining how we share music on Spotify — a UX case study
I don’t remember the last time someone sent me a song. Perhaps it’s just me, but I have had this feeling that we’re sending each other less music. Gone are the times when we used to check out each other’s CD collections, better yet what music our friends had on their iPods. Music sharing needs a new awakening, and it starts with us
I’ve often pondered what can be done for the music world. In 2018, sales of digital download purchases were down by 27% and album sales were down by 41%, continuing a general downward trend since the boom of online music sharing. Vinyl sales, which in recent years have once again become popular, have increased in revenue — but lagging way behind other forms of distribution. It is clear that streaming music has become a primary source of how we access music. But what are we listening to and how are we discovering it?
Disclaimer: This is a speculative project. I am not affiliated with Spotify or any other music service. As a freelance UX Designer and a musician, I took on this challenge because I’m passionate about music and design.
Phase 1: Research
When Spotify launched in 2006, it was clear that the music industry needed some help. Today streaming accounts for 75% of how we listen to music. By the time you’re reading this, Spotify will have roughly 198–210 million users worldwide. Currently, 83 million of them are paid subscribers with numbers expecting to reach 95–100 million by early 2019. With almost twice the number of subscribers than Apple Music, Spotify will remain to be the leader in the paid music streaming market. It is clear that Spotify is dominating the music world, but how are consumers feeling about Spotify and how is it ultimately impacting music, and how we access it?
“Today streaming accounts for seventy five percent of how we listen to music.”
I knew these were tough questions to ask, so I started by zooming out, to take a look at the big picture first. I analyzed the market and performed a competitive analysis to get a broad view of key competitors in the field.
Though the analysis didn’t reveal much about a problem space this time, it provided insight on Spotify’s position in the global market. The analysis made me realize Spotify has a huge responsibility for the music community and economy, around how we interact with and share music, just like MTV had in the ’90s.
I wondered, how are we sharing music with each other? Taking this as my hypothesis, I conducted a 4-week design challenge to figure out how a social feature could be implemented into Spotify, and ultimately proposed a solution.
The Band-Aid attempt
In 2015 Spotify removed the messaging feature due to low engagement. It was hidden somewhat in the menus, with users citing that it took many clicks to get to and was not simple to use because of the unnecessary navigational burden placed on the user. I actually liked the feature and used it quite often with several friends — one which I had a crush on and ended up marrying two years ago.
The now-removed messaging feature served two functions: to facilitate messaging and sharing music with friends and family. It served as an enabler to express our thoughts and feelings with one another through music. Many expressed frustration when this feature was removed, though it seems many weren’t aware of it in the first place. To me, this signals an opportunity to get it right a second time around.
Before trying to figure out what to build — as tempting as it may be to think of, it was most important to start with users. Users always come first as they are ultimately the ones that will use the product. This might sound cliché or obvious, but a staggering number of companies see this mere fact as a UX problem, rather than a company problem. The fact is, you can’t just tack on UX by the end of a product launch cycle. It’s like trying to drive with your eyes closed. Though somehow it’s being done all the time. — No, I don’t mean people are driving around with their eyes closed.
“You can’t just tack on UX by the end of a product launch cycle.”
To start, I was trying to figure out pain points users have when sharing music on Spotify. According to Spotify, 72% of Spotify listeners are millennials, with 26% 18–24-year-olds, 29% 25–34-year-olds, 16% 35–44-year-olds, 11% 45–54-year-olds, 19% 55 and older. I scheduled interviews with mostly millennials since they represented the largest user demographic.
I was able to interview several participants for my research. Interviewees were screened on the basis that they were regular users of Spotify. I spent around an hour with each participant, uncovering little bits of information from each real user about who they are and what their experiences are through using Spotify. I compiled a lot of notes and synthesized the findings into a single uncovered problem: users have to leave Spotify in order to share music with each other, disrupting their experience by forcing them to leave the app. I found that this led to sharing music less between users.
As also uncovered from research, I was able to identify another common pain point amongst users. 75% of them said that the curated playlists that Spotify offers such as Discover Weekly were not really instrumental in addressing their needs to find new music. Spotify does intelligently predict what users would like to listen to however it doesn’t allow discovery outside of their music taste. — One user describing this similar to an “Echo Chamber”. Pretty much if you play several EDM tracks, chances are you’ll be hearing EDM for the rest of your life. Nobody wants that as their everyday routine. I mean if you’re into that, it probably shouldn’t be the only music you listen to. One participant stated: “I often ended up listening to the same songs rather than discovering any new music.” During the interviews, 100% of participants claimed that they like to discover new music from friends and would like to check out their friends created playlists.
On to generating new ideas…
2: Define
As I found out a common pain point amongst users, I focused on how users share music with one another on Spotify and then conducted a second round of interviews with a narrowed down scope of focus. One user commented, “I don’t share music on Spotify because it’s not clear how it works”, making it further clear about our problem space. Copying and sharing the link by pasting it somewhere else is not really sharing, it’s a workaround to a non-existent solution.
I needed to zoom out and get a clear view of our users and what they had to say. For this, I made an empathy map to organize the research into categories that would inform the user flow and our persona. I used post-its, which allowed me to paint a broader picture with clear cues as in what to build. Using a wall helps, as it allows you to take a glance at what the users are saying in a memorable way. Also writing short simple statements with a bold marker helps us digest information to clear to see, in little bits.
3: Ideate
The empathy mapping exercise is a great tool to understand your users. It helps us organize thoughts and patterns uncovered from our interviews into the creation of our user persona. This persona represents the target needs and aspirations of our user. Though fictional, the results are entirely uncovered through research, giving further purpose and clarity to our problem space.
After research, I had uncovered a problem space and knew some of their pain points. I found out what my users weren’t getting. I wasn’t formulating any solutions yet, but only gathering more evidence to support my claims from our users that there was a problem that needed addressing.
How might we…
In order to organize our findings into actionable statements, I used a technique called HMW — short for How Might We statements. HMW questions reframe uncovered research insights into question format sentences. This allows conversations around our problem statement, enabling Rapid Brainstorming Sessions to generate as many ideas as possible. A good example of a How Might We for Uber would be: “How can we get users from point A to point B?”
Once I had about 40–50 how might we statements. I narrowed down my top four. These were:
How might we allow users to share music without leaving Spotify?
How might we help users view songs sent by their friends easily?
How might we make it easier to discover music from friends?
How might we make it easier to pick and share music with friends?
The HMW exercise allowed me to focus on a specific problem statement. I then generated ideas using a technique called Crazy 8ths. This technique works well to generate ideas rapidly, blindly in a sense, where with each iteration the solutions start becoming more clear and take shape and form. I’m not uploading what these look like because of the rough nature of them.
Applying findings synthesized from our empathy map and rapid brainstorming sessions, I needed to start visualizing proposed UX ideas.
Storyboarding helps visualize our ideas into how it fits into our user’s daily lives. As stated from the Nielsen Norman Group: “Visual representations of UX stories capture attention, provide clarity, and inspire us to take action.” By applying our learnings into this format it directly informed the flow our users would take when sharing music with their contacts.
Once I had the missing pieces of how things would fit in the flow of things, I was ready to build a user flow and start building my hi-fidelity wireframes. I decided what needed building was a way of sending music directly in Spotify to contacts without leaving the app. A new notification system would let the users know that they received a new track/playlist and other useful information, saving those files automatically into a playlist of music sent from friends. Currently, users have no way of searching for a previously sent song. When they get recommendations from friends they leave the app and click on a link from Social Media or their messaging platform, going to listen to the song, only for it to vanish from memory days later, unless manually saved.
“It’s all in the story. For which it tells the story of our everyday user.”
User Flows inform your design…
The Storyboarding exercise informs the journey our user would take when sharing a song with a friend. Once this was clear, I could determine the amount of design work necessary to build the new features.
The user flow was created to inform our design and show how the new features would be implemented to the existing platform:
4: Prototype
Hi-Fidelity Wireframes were created to accommodate the new screens from the user flow. Screenshots of the app were taken and the additions were made using Spotify’s existing design guidelines.
All I needed at this point was a working prototype of features to be added so I can test it with real users and gather feedback through usability testing. I took screenshots of the app and tapped the screen with my magic wand for a working prototype of the user flow from before.
Click here to view and interact with the prototype.
I reached out to my circle and contacts to prepare for usability testing. It takes time to gather participants for testing, so by the time I start working on the designs, I will have scheduled participants for the next day ensuring no hangups in completing testing.
5: Test
All the work up to this point boils down to this moment. The moment of truth! We ultimately want users to be able to accomplish tasks we give them to do during our Usability Testing. Users were given the task of finding music sent from a contact, playing it, and then sending the newly discovered track to another friend using Spotify, without leaving the app.
As result, participants were able to complete all given tasks with 100% accuracy, citing ease of use and instant familiarity of user interface elements.
Conclusion
Spotify offers a great service. I appreciate the ease of having access to the “record store” everywhere I go at a reasonable monthly fee. I also love vinyl records. Perhaps what drew me to collecting vinyl in the first place was my bittersweet bond to Spotify. I had missed the hand to hand, and social interaction around listening to music.
The know-it-all algorithms of deep machine learning that allows Spotify to curate custom playlists is enjoyed by many and is becoming the primary way of how we find new music to listen to. But it’s not an all in one solution, and should not be. If Spotify manages to address how we interact with music socially, it will make sharing easier and more memorable in a meaningful way.
Thank you! If you enjoyed reading this, please 👏 and share!
