1. Main parameters

For each public, listed release the platform counts the number of distinct buyers who completed a purchase in the past 30 days, then picks the release with the most distinct buyers. A release must have at least three distinct buyers to be eligible. When two eligible releases have the same distinct-buyer count, the one with the higher sales revenue over the window wins; if they are still tied, the one created more recently wins.

The 30-day window is rolling: a purchase that succeeded 31 days ago no longer counts. Free streams, plays, follows, and likes do not count at all. Multiple purchases by the same buyer count as one buyer, so volume from a single account (or a handful of accounts) cannot lift a release.

A release that wins the SQL ranking can still be dropped before it reaches the hero slot if the artist has not completed Stripe payout onboarding (so a purchase would fail at checkout) or if the release has been flagged for conformity review. When that happens, the slot falls back to the freshest verified release and the “Top this month” badge is hidden, replaced by a “Fresh release.” tagline. So the presence of the badge is itself a signal: if you see it, the release was picked by distinct-buyer ranking, not by recency.

2. Why this criterion

SynthCamp's mission is sustainable artist income on AI-assisted music. The number of distinct buyers is the most honest signal of an audience genuinely choosing a release: it measures reach across real people, not how much any one wallet spent. Ranking by raw revenue instead rewards a single big spender, and is cheap to game by having a couple of accounts buy a release back and forth at the price floor. Counting distinct buyers (and requiring at least three) makes that manipulation far costlier, because each fake buyer needs a separate, funded account that clears payment and abuse checks. Free signals (listens, follows, likes) are excluded because they are the easiest of all to inflate.

The 30-day window keeps the surface fresh and gives newer artists a real chance against entrenched lifetime totals. The eligibility filter exists because pointing the hero at content that cannot be purchased (Stripe missing, conformity flag) would frustrate the listener and embarrass the artist.

3. Relative importance of inputs

Distinct-buyer count over the 30-day window is the primary and only ranking input; sales revenue is used solely to break ties between releases with the same buyer count, and a release must reach three distinct buyers to be eligible at all. There is no boost for follow count, no diversity bias by genre, no novelty adjustment, no editorial override, no per-viewer personalisation, no paid placement. The full source of the ranking is in the public repository.

4. Your options

Since this ranking is the same for every viewer, there is nothing to opt in or out of. There is no personalisation to disable, the badge is always present when an eligible release exists, and there is no setting that would change which release appears for you specifically.

If a personalised surface ships in the future, the controls will live in account settings and this page will grow a section explaining how they work.

Update cadence

The ranking is recomputed at every page request, with no cache. Buying a release changes its position on the next home page load. A short cache may be added later for performance; if it is, this page will note the cache duration.