Spotify Line-In
While at Spotify, I worked with a newly formed team to figure out how to allow users to help identify problems. From the Line-In FAQ page: "If you see something amiss or just plain ol’ missing, now you can help us correct it or add it…” and “By suggesting edits… you'll help artists connect with more fans, and help other people discover new music by fueling our personalization engines.” Line-In was positioned to help all personalization efforts across the company.
When Line-In was released it was featured in numerous publications such as Billboard, Variety, TechCrunch, and Quartz.
Role & Contributions
Lead Designer in charge of concepts for how to crowdsource edits to Spotify’s metadata
Worked with Product Owner and Engineering Lead to define scope and feasibility of MVP
Presented ideas and statuses to VP, Design of the Tribe
Worked with User Research Team for initial MVP testing
Worked within predetermined Spotify Web Style Guide as much as possible
Early Concepts
An early idea was to get greater participation from Spotify Free users by incentivizing them in multiple ways.
A second concept tapped into the idea that the Top Fans were the best experts.
A third desktop idea let you opt-in to make changes based on your listening habits…
…and then ran you through a small wizard-type interaction to gain feedback.
A fourth option directly led into the MVP version…
…where you could report an issue and get a structured feedback form with a comment box.
MVP Version
The team and Senior Leadership had discussions around some of the early concept work, feedback from users, and thoughts from the curation team, and decided there was a bigger opportunity available. The team and I decided we should take cues from the power of editing a wikipedia page and giving feedback to search results on google to rethink our next iteration.
The MVP became a website that a user could visit either on their own or launch via a “suggest an edit” menu option in the Spotify desktop client. Users could help Spotify clean up its metadata catalog by adding suggested edits, upvoting or downvoting another user’s suggestions, or adding new content including aliases and tags, external websites, artist roles, and more.
A sample artist page
While not all artists would have a page as robust as Drake, the Artist page was built to degrade gracefully as sections would slide up or collapse when not needed or empty.
Some features of the artist page:
A photo grid showed an overview for easy viewing while each photo could be opened larger to view controls to allow a user to mark it as accurate or not.
Tags a current user added were shaded in yellow to indicate they were not yet pulled into the system. The only control was an 𝖷 to allow a user to remove what they added but not upvote their own additions.
Sections that were verified were marked with a 🔒 lock.
Users could navigate directly to albums from the artist’s page to make further suggestions or edits.
Track & album Pages
Track and album pages had a slightly different set of options. Explicitness was a unique option that tracks offered that allowed users to select different types of explicitness from a dropdown. Artist Roles allowed for users to identify who did what and help give credit where it was due.
A sample track page
A sample album page
Line-In Landing Page
The focus of this page was to get people searching for artists, albums, or tracks as quickly as possible as well as showcasing the latest edits to the system.
Line-In Search Results
Search results had to offer a variety of starting points for more ambiguous searches. Users could dive deeper from any one of the sections if their exact search wasn’t shown.
Inline action interactions
Early iteration of the “Add New…” control
Each section had a control that would allow editors to add additional metadata that they found relevant.
Upvote/Down Vote Control
Each tag allowed a user to upvote or downvote its contents. The idea was that after a vote hit a certain threshold, the system would either continue to persist the tag or remove it from the system.
Wrap Up
Line-In was slated for the UX Research team to pick up right after my contract ended. Line-In was later released in March of 2018 and was very close to what you see above, but had incorporated another project I had worked on, Spotify’s Track Rating Tool. Line-In has since been retired and, as you can see, a lot of users were pretty disappointed to see it go.