← Back to portfolio
Discovery research · SoundCloud · Dec 2022 – Mar 2023 · case study 04 · 5-minute read

SoundCloud: charts that let niches be found

SoundCloud is full of niche scenes, but its charts only had broad genres. I proposed a small, testable change: let creators and fans propose and vote on new subgenre charts.

My roleSolo, self-initiated innovation case study (a 55-slide deck), written to support my applications to SoundCloud's creator-growth and music-discovery teams
WhenDec 2022 – Mar 2023
StageDefine and develop
MethodsChart and search audit, competitor comparison, PESTLE, minimum viable change, agile build–launch–test plan
The headline

A whole scene with no chart

0 charts

There was no grime chart and no way to filter for women in grime. A grime artist couldn't stand out in the charts, and a grime fan couldn't browse the best of the scene.

The problem

A discovery problem on both sides

When a subgenre has no category, creators can't stand out and listeners can't dig in. Fewer plays mean lower royalties, so niche artists find it harder to turn a passion into a career. I called this the symbiotic discovery problem: creators and fans lose out together.

What I proposed

Start small: open-design charts

  1. Propose and vote

    Creators and fans suggest niche chart categories and vote on them, using Instagram Story tags where the scenes already talk.

  2. Add the label and the chart

    Each winning subgenre gets a label and a playlist with two filters: all-time best and this month's hot releases.

  3. Test before building more

    Launch, measure and only then join the new labels up to search, the home feed and generated playlists.

Why a chart first? It's the cheapest change that tests the idea. If niche charts don't get used, nothing else built on the labels is worth the engineering time.
The bigger picture

A learning community, hiding in plain sight

Longer term, I saw SoundCloud as a space where musicians and other creators connect and co-create releases before they go out to a wider network of partners. Subgenre labels decided by the community would be its foundation. Where the labels mean something to a local culture, they could also shape local charts.

What I took from it

Ship the smallest change that tests the biggest assumption

Then → Now

Then: I wrote it as a job pitch. Now: I'd lead with the evidence and the test plan, and keep the vision short.

Then → Now

Then: Categories decided by the platform. Now: categories proposed by the community. The same idea runs through my later creator-economy work.

Then → Now

Then: One genre example. Now: I'd test across several scenes before generalising.