Confident music discovery, helping listeners reach a worthwhile first play, faster.
Atlas is a music-discovery concept built around one question: how do you help someone reach a confident first play, faster? Listeners skip, second-guess, and lose promising tracks because they can’t tell what’s worth trying before they commit.
Over eight weeks I led research, design, and testing to reframe discovery from an endless browsing surface into a short sequence of confident decisions: starting with intent, evaluating faster, and capturing options without pressure.
Interviews on how listeners decide what is worth playing, and why fast discovery rarely means confident discovery.
Streaming platforms make music instantly available. But availability isn’t the same as confidence.
When I spoke with listeners, the friction wasn’t finding music, it was deciding whether something was even worth trying before pressing play.
That pattern wasn’t about algorithmic failures, it was about signal failures.
Even strong recommendations break down if users can’t evaluate them well enough to act.
Atlas started by focusing on one central question:
Comparison labeled “Fast Discovery ≠ Good Discovery”, fast-discovery surfaces (infinite scroll, autoplay playlists, algorithmic feeds, personalized surfaces) paired with their failures (skip within seconds, unclear relevance, distrust recommendations, lose promising tracks later)
Early research pointed to something subtle but consistent:
After talking to listeners, Atlas became less about expanding discovery surfaces and more about strengthening evaluation signals.
Instead of treating browsing as an exploration space, I began treating it as a decision window.
Insights-to-direction diagram, four research insights (skipping creates fatigue, mood drives discovery entry, discovery windows are short, people trust context not algorithms) each mapped to a design direction
From those signals, I defined three priorities that guided every later decision.
Moving forward, Atlas became a system designed around confidence before pressing play.
“Three Priorities That Guided Every Decision”, three columns (Start With Intent, Evaluate Faster, Capture Without Committing) each with supporting bullet points
Reframing discovery from a browsing surface into a sequence of confident decisions.
Most streaming apps start with a playlist, a feed, or a search box. Atlas starts with a question:
Playlists lean on trust in whoever curated them. Algorithmic feeds go further and remove the decision almost entirely, trading transparency for speed. When listeners can’t tell why a track surfaced, discovery starts to feel arbitrary and they fall back on skipping or replaying the familiar.
Atlas moves the first decision earlier. Instead of filtering results after the fact, it asks for context up front, so discovery begins by setting direction rather than reacting to a feed.
“Discovery Models” comparison of Playlists, Algorithmic feed, and Atlas across what each starts with, how it works, and when it breaks down
Traditional discovery loops look like this:
browse → browse → browse → …maybe play.
That works when listeners already trust the recommendations. When they don’t, browsing turns into hesitation disguised as exploration.
So I reframed discovery as a sequence of decisions instead of a place to wander. Most interfaces treat pressing play as the moment evaluation begins; Atlas treats it as the moment evaluation ends. That shift moves the work earlier: instead of committing to a full track to find out if it fits, listeners compare options, read relevance quickly, and decide with less risk.
Atlas became less about navigating a catalog and more about supporting a confident decision before play.
Side-by-side flow diagrams, Traditional Models (open app to browse feed to skip to browse to maybe play, “hesitation disguised as exploration”) versus the Atlas Model (open app, curiosity, intent, results, test, decide, play, save)
Once discovery became a decision flow, the interface had to make options easy to evaluate. Five features carry that weight:
Moderated usability tests validated the guided decision flow and surfaced targeted refinements.
Testing Atlas with users gave me key insights on how to move closer towards their goals with music discovery:
Testing confirmed the core direction of Atlas, but it also revealed where discovery needed stronger clarity, flexibility, and evaluation support.
I focused the next iteration cycle on helping users understand recommendations faster and move through discovery with less hesitation.
I introduced several targeted changes:
Outcomes of Atlas and what designing for the moment before play taught me.
Atlas improved the moment where discovery typically breaks: the decision to try something new.
The result is a more confident path from curiosity to commitment:
A person holds a phone displaying the Atlas Discovery Mode screen
This project changed how I think about discovery.
I started Atlas assuming the problem was navigation, but through research and testing I realized the real friction happens earlier than that.
It happens in the moment someone asks:
Is this even worth trying?
Listeners aren’t overwhelmed by a lack of music. They’re overwhelmed by uncertainty.
They don’t want more options. They want clearer signals.
Designing Atlas shifted my focus from organizing content to orchestrating confidence.
Closing reflection panel for the Atlas case study