Search it, find it, buy it

Bandcamp is an online record store with search as its top sales-driving feature. I defined new search engine functionality that made our search feel more seamless than ever by solving our users’ biggest pain points, allowing music fans to quickly find and purchase what they’re looking for.

My Role
  • Product Design
  • User Research
  • Bandcamp
  • Summer 2022
Bandcamp Search on desktop and mobile web
Incremental visual updates with a powerful new backend enable music fans to find exactly what they’re looking for.

Bandcamp is where passionate music fans go to directly support artists. As the top sales-driving feature, the search bar is where many of these sales begin. But while fans may know what they want, they can’t always find it.

A good search shouldn’t be noticeable, so why are people noticing it?

As the sole designer on the Search & Discovery team, it was up to me conduct research and comb through existing data to understand which parts of search were working and which parts weren’t.

Examples of how users formulate their search queries
With the Data team’s plate already full, I took ownership of shifting through thousands of queries to discern any helpful usage patterns.

Search insights

  1. Fans were getting unexpected matches to their searches as their searching behavior evolved with modern search engines.
  2. Artists were frustrated that their new music didn’t appear in search as soon as they released it.
  3. Ambiguous search terms had fans digging through pages of results.
Most commonly failed search queries are combination queries, filler words, character & word substitutions, and items not in catalog
I put together a list of our most failed search queries to identify the biggest pain points our users faced within search.

Let’s get technical

After analyzing a dizzying amount of search query data, I learned that the way users constructed their search terms was simply too advanced for our existing search engine. So we rolled up our sleeves to make some major improvements to the backend. I worked closely with developers to define and tweak the behavior of newly supported query types in order to ensure that the search experience felt as seamless as possible.

Users attempt to narrow down results with a more specific query but it only makes the results worse
Journey of how a user might adjust their search query in an attempt to get a more relevant result.

Adapting to how fans search with combined search terms

As users search, they’ll typically add more search terms hoping to narrow down the results, but in our case it actually used to make matters worse. I defined a set of supported combinations to address one of the most commonly failed searches. Now fans can combine search terms like artist + album, track + album, and artist + label to find catalog items that were previously inaccessible via search.

Old search results with irrelevant items New search results with highly relevant matches
Before & After Justice’s album Woman is finally searchable. Combining multiple search terms like artist + album now return vastly more relevant results.

Leveraging new release hype by making speed a feature

When an artist publishes an album, fans flock to find it whether it’s to purchase, wishlist, or listen. But if that album doesn’t appear in search, it’s like it doesn’t even exist. In order to take advantage of the buzz around release time, we had to prioritize reducing the lag time between an artist publishing a release and the album appearing in search. By treating speed as a feature, we also eliminated artists’ top complaint about search: dude, where’s my album?

4 hours in the worst case scenario to appear in search Less than 10 minutes in the worst case scenario to appear in search
Before & After Even in the worst case scenario, the lag between publishing and appearing in search was significantly reduced and in most cases new releases appeared in search almost instantly.

Less digging with new filters

For the truly ambiguous search terms, users could still have to dig through pages of results to find the right one. I designed a simple filter based on existing design patterns that enabled fans to get to the right result quicker.

With filters, Only Human by KH goes from #26 overall to #1 within tracks, making it easier for DJs everywhere to make a beeline to their intended tracks.

Better search results meant better engagement

The new search engine was a success! After we launched, we saw an increase in the engagement rates of actions like click on result, play music, and add-to-cart — even in periods where overall site engagement was down. The new search backend laid the foundation for many of the UI updates in to come in the future.

Next Project

Bandcamp Discover