Spotify Affinities
Client: Spotify
Project: In-app Social Feature
Duration: Approx. 2 weeks
Role: UX Designer
Project Overview
Spotify’s legendary music streaming service has a surprisingly timid history with social integrations. Being the leader of streaming music services, they have been primarily devoted to offering their customer base a premium listening experience that enables them to listen to high quality audio, legally, and virtually anywhere at any time. They decided to pursue a social feature that integrates seamlessly with their current mobile application and fortunately I was chosen to spearhead the strategy, design, and implementation for the delightful project.
Research
Q: For how long do you usually listen to music?
A: “Omg 24/7 is that even possible?!” -Hayley, 24.
It’s fair to say that a substantial majority of the world is a bit obsessed with music - at least subconsciously. Because of this, finding participants who use a music streaming service was no difficult chore. However, there were a few things to consider. For example: Spotify’s general userbase are at least somewhat tech-savvy which would place the most broad participant criteria in generation Y/Z territory.
Methodologies
This project research was approached ethnographically and interpersonally. The methodologies used to gain an understanding for the pains and pleasures for mobile social features and behaviors were competitive analysis, secondary research, and user interviews.
Goals
Discover the void between peak social media experience and peak audio streaming experiences
Extract information about what is attractive to avid social media users about social media applications
Discover what might be isolating in regards to audio streaming (hypothetical social pain point of streaming music)
Develop personas and a list of wants and needs for a social dimension to music streaming
Extract hypothetical out-of-box solutions regarding social media, music, and community
Competition
Currently, Spotify’s social features are mainly focused toward allowing the user to broadcast or share their musical experiences in external applications such as Twitter, Instagram, Facebook, etc as well as collaborative playlist creation. A user may decide to share a new musical discovery as an Instagram story or post on Facebook. Also, the user may use Spotify integration to add music over an Instagram story. These features serve to bring social media users from different applications to the Spotify app.
Nevertheless, despite Spotify being the current titan of music streaming platforms, there are many that would like to see more daring features. John Paul Titlow of Fast Company argues that Spotify shouldn’t necessarily copy the ‘stories’ feature that all the major socials have but instead just aim to make their listening experiences more interactive between songs, playlists, etc.
Personas
Being so familiar with creating, producing, and listening to music, a set of provisional personas were relatively easy to evoke from the ethers. Initially, Spotify users would appear to come from every type of background and express nearly every behavior across the board. Therefore, the provisional persona drafts were focused on those most likely to be heavy users: audiophiles.
After organizing the research, two personas were distilled from the pool of information and given shape from the most notable divisions between temperament: social investment, online habits, and inclination to logic or emotion.
Kimberly
Everyone who knows Kim is sure to like her. She’s warm, passionate, and inviting. If you met her she’d be delighted to exchange notes about the best new soul revival band as much as she’d like to ride around the neighborhood blaring trap. It would be a shame to design a social feature for the world’s most respected music streaming service without her in mind.
David
On the other end of the socio-musical spectrum is David. Rather than sharing music as a form of neighborhood currency, David approaches music sharing as a researcher exchanges findings. To some level, he may even seem to flex new discoveries as a competitive advantage among other audiophiles. Driven musical explorers like him are at the other end of the most invested Spotify userbase.
Summary of Findings
Most people fall in between David and Kimberly in regards to motivation for listening to music on Spotify but the common thread between the two personas as was informed by user interviews was the desire to connect with others to gain musical awareness and even establish connections with people using musical affinities as a litmus for social compatibility.
Project parameters
Rather than committing to a large lift, it was best to approach a new feature with only the most important needs for Spotify’s target users: connection based on community. We could then move forward with more ambitious feature sets depending on success/failure.
Feature Roadmap
Ignoring the Obvious
The most immediate response to developing a social function would be to list “chat” as an option. Our project manager decided against it for good reason: it required too much investment and ultimately, what would suggest it would work now rather than before? We had to dig deeper and look for a more sophisticated approach to bringing people together.
Prioritizing Transparency and Affinity
Based on research, one of the key successes of Last.Fm (during their heyday in the mid-2000’s) was their neighbor feature which paired others by musical compatibility as measured by their pioneering audioscrobbling. To recreate this excitement and to make the experience more ‘now-oriented’ we decided to prioritize the ability to see which users were currently listening to the same song as “User1” (that’s you).
Furthermore, with all of Spotify’s listener data, we built Affinities to reveal a percentage of compatibility (affinity score) that would be displayed on the currently listening screen, right under other users’ names as well as other users’ profiles.
User Flow
To test refine the feature and to anticipate what needed to be designed, we imagined a user flow through the app in order to pin down what components were necessary in the preliminary wireframes.
Our feature design thus incorporates the following components in its core flow:
Observation of current listeners (user song-specific)
Quick add
Observation of user affinity scores
New Spotify Affinities profile
Reactions to ‘currently listening’ (Emoji only)
Quick recommend
Audio recommendations/inbox
design: Bringing It All Together
Most Spotify screens are variations of list processes that have been developed over the years to be easily navigable for users. We didn’t want to tamper with that too much so we retained a lot of similar patterns.
Above: mid-high fidelity wireframes featuring annotations of new features. The new profile screen displays personally curated metrics. If a user wants people to know their most played track of the day/week/month or perhaps a carefully crafted playlist, they can. The other aspects of their profile are auto-populated such as “your affinities…”. These features are informed by data collected by Spotify and that which is in common and/or recommended.
Hi-Fi Design
Hi-fidelity rendering of Spotify Affinities with active, functioning features. Time and consideration was placed on making the feature fit seamlessly within Spotify’s already existing UI interface.
Test
Prototyping
Our user flow became the guidelines for our prototype. The animations between screens were determined by what Spotify’s current user interface utilized or what would be intuitively part of the user flow. We wanted the task flow to be linear and move the user to either their inbox or other users’ profiles intuitively.
Testing
Objectives:
Test functionality of feature and navigability within app
Test path efficiency from current song to user-to-user social connection
Test CTA awareness and effectiveness of CTA design in navigation
Analyze results and optimize for faster, more efficient usability
Test Subject:
High Fidelity Spotify app (mobile) proto-type
Methodology:
In-person observation: participants will perform the tasks pertaining to the given scenario on the current high fidelity Spotify app prototype via my laptop.Remote observation: participants will perform the tasks pertaining to the given scenario for this testing phase while screen sharing on Zoom/Google Meet.Participants:
4 total: (3) men; (1) woman
Age range: 24-35
Current, daily Spotify users + (1) SoundCloud user
Wins & Fails
The usability testing went very well. Most users (75%) found the feature delightful, but all found the app relatively seamless and provided at the very least, an effective social component to Spotify. Predictably, most participants had a learning curve that was mainly related to the function of the feature’s CTA on the currently listening screen. While there was no single aspect of the feature that was unanimously liked, all participants had their own favorites. This is actually a major plus. It reveals that the smart profiles feature in Affinities offers a multifaceted experience to different Spotify users and what is broadly satisfying is likely to be successful in the long run.
Conclusion
There were only a few priority revisions after usability testing. Mainly, they were technical issues in the prototype such as problems with the smart animate feature when navigating to other screens.
The challenge with this project resided primarily in forging a new path with only pieces of old maps; chat wasn’t an immediate option but inspiration from Last.FM’s glory days illuminated possibilities for delight via their former social features. Furthermore, as a music lover myself, I needed to check my assumptions at the door. A lot of attention and consideration was paid to those interviewed and the personas created to guide the feature development.
If Spotify Affinities succeeds, users will be ready for more social integration. Test participants are still eager for chat (maybe it will work better after users are coaxed out of their Liked Songs screen and into other users’ profiles via Affinities). Other little add-ons like a quick view under affinity score, and more visual cards on profiles are on the feature roadmap as well. For now, I should probably take a break and do some socializing myself. But then again, there are still a few releases I haven’t had a chance to listen to…