Find, don't search: The principles behind music search UX

At AIMS, our mission is to solve search for the music industry. That means we spend a lot of time thinking about the technology, but just as much about how it should feel to use. And the more we build, the more we come back to something that sounds obvious but rarely gets applied: the goal of music search is not to search. It's to find. That distinction has shaped everything about how we approach UX, and it comes down to four principles.

1. Design for finding, not for searching

There’s a difference between a “search strategy” and a “find strategy”. A search strategy focuses on how you interact with the tool: what you type, which mode you select, how you configure your query. A find strategy focuses on the outcome: you need the right music, fast. Everything else should get out of the way.

The UX should absorb complexity rather than expose it. The underlying tech might be doing five different things at once. The person using it should feel like they did one thing: found the music they needed. This changes the design decisions you make at every level. You stop asking “how do we build a great search experience?” and start asking “what’s the fastest path from an idea to the right track?”

2. Input-agnostic by default

A music search might start with a keyword, a reference track, a YouTube link, or a detailed prompt describing exactly how a scene should feel. The traditional approach gives each of these its own interface, each with its own rules.

A better approach is a single point of entry that figures out what you’re doing based on what you give it. Paste a link? The system knows it’s a reference search. Type a sentence? Prompt search. The intelligence should live in the system, not in the user’s decision-making.

3. Always tell people where they are

Simplifying the input doesn’t mean leaving people in the dark about what happened. When a system interprets and runs your search behind the scenes, it needs to communicate that back to you. Clearly, but without turning it into a tutorial.

The key is making this communication visual and intuitive rather than instructional. Think Icons, patterns, and consistent visual language that builds familiarity over time. If someone notices that a certain icon keeps appearing next to the results they like best, they’ll start seeking it out naturally. That’s orientation done right.

This also matters for onboarding. When a new team member starts using search tools for the first time, good orientation compresses their ramp-up. The interface teaches them as they go.

4. Offer choices that open doors, not add steps

Simplicity doesn’t mean stripping out every option. It means being intentional about where you add them. A well-designed search adapts to how someone actually works, rather than forcing them into a fixed format. The test is simple: does this option expand what the user can do, or just slow them down?

If someone pastes a YouTube link, the system can run a similarity search based on the audio. But it could also analyze the visuals and find music that matches the mood of what’s on screen. Offering that choice gives the user a new capability. That’s a door opener. Compare that to an interface asking you to pick between prompt search and keywords search before you’ve heard a single result. That’s making the user do your engineering team’s job.

A good rule: if someone has to understand the underlying technology to make the choice, you’re asking too much.

Great tech needs great UX 

The technology behind music search matters. But technology only works as well as the experience around it. That's why at AIMS we care about both, and keep pushing to build search that gets out of the way and lets you focus on the music.