Every entity in Rvrb is built from public sources, cross-referenced, dated, and graded. Here is exactly where the data comes from, how we rank it, how often it refreshes, and where coverage is still thin.
The public feeds every figure on the site traces back to.
The dot beside a fact, source, or chart position tells you how well it’s backed.
One aggregate ranking for Indonesia — and the only ranking we publish.
score = avg over present sources of 100·(101−rank)/100, weighted 0.40·Spotify + 0.35·Apple Songs + 0.25·Apple AlbumsEach source’s best chart position for an act is normalised to a 0–100 scale, then averaged across the charts it appears on, weighted by reliability. An act needs at least one public chart to appear; the quality dot (● 3 sources / ◐ 2 / ○ 1) reflects how many corroborate the position. Movement is today’s rank against its recent baseline. It is built purely from public chart positions — nothing about plays, sales, or editorial judgement feeds it. See the live chart →
Every figure is stamped with its snapshot time. See what’s fresh right now →
Biographies are drafted from referenced sources and reviewed by a human before they are published — never auto-posted. Each fact is dated, so you can see how current it is.
When a source goes unreachable we hold the last good value and mark it cached, rather than show nothing or a zero. A cached figure is honest about being cached.
What we don’t cover well yet — said plainly.
Indonesia first. We go deep on Indonesian acts before widening. A foreign artist may appear only where they connect to an Indonesian release, festival, or collaborator.
Listener counts are partial. Monthly-listener figures come from public data; a handful of acts aren’t matched to a source yet and show no count rather than a guess.
Chart history is young. Our canonical daily chart record begins 5 May 2026. Anything that reads on long-run chart presence — a year of history — will not have a full window behind it until mid-2027. Until then, a short record is presented as a short record, never stretched to pose as a long one.
Coverage is filling. Credits, line-ups, and back catalogue are still being completed. Gaps are expected — and a correction is the fastest way to close one.
Empat pilar (Jangkauan 30, Resonansi 25, Relevansi 25, Realisasi 20), dibandingkan dengan 497 artis Indonesia pada norma beku music-id-2026.3. Skor terbit hanya jika kelengkapan data ≥ 60 %; per 30 Oktober 2026, 4 artis memenuhi syarat. Setiap skor memuat rentang ±, kelengkapan, dan jejak audit input mentah.
score = Reach 30 + Resonance 25 + Relevance 25 + Realisation 20One hundred points, split four ways. The weights are fixed and published; they are not tuned per artist, per genre, or per anything else. The same four pillars run for music and for sport — only the inputs inside them differ.
The score is free, and it is the same number for everyone. There is no paid plan that buys a different score, an earlier one, or a better one. If Rvrb ever sells anything, it will be analysis around the number — never the number itself, and never a way to move it. We also hold no stake in the artists, labels or festivals we measure, and no payment has ever changed a figure on this site.
Most artists show no score, and that is the design. A score is published only when enough of its inputs are actually present — a 60% completeness threshold, fixed and public. Below it an artist shows Belum diskor · data belum cukup with the data we do hold underneath. Rather than fill the gaps with zeros or estimates and print a number that looks precise, we print nothing. A missing signal is never counted as a zero, because “we have no data” and “we measured nothing” are not the same statement. The count above is live: it is the number of artists whose most recent computation cleared that threshold.
Scores are ranked against a frozen Indonesian reference set. A 0–100 score is a percentile, and a percentile is only meaningful against a fixed population. Ours is the norm vintage named above: Indonesian artists only, frozen on a date, and never re-ranked against whoever happens to be active today. When the population is revised, that is a new vintage with a new name, and the change is published rather than applied quietly. Every score arrives with its inputs, their sources and timestamps, the version of the weights, and the coverage behind it — if a figure cannot show its working, it does not ship.
What counts as attention is fixed with the vintage, too. Two terms measure an artist’s share of the mentions received by the whole reference set — over the last week and the last month — and Momentum is the 30-day trend of that share. For music-id-2026.3 the mentions are counted from the international news index GDELT and the domestic Indonesian music press. Adding a source to that basket makes a new vintage; it is never added to the old one, because it would move every published percentile with no change in the world.
A score that is not recomputed within 7 days is labelled old. The page says Skor lama and shows the date it was computed, rather than presenting a September number as current in December. An old number presented as current is a wrong number.
Questions about a score? Contact us →
The method is settled and every published number can be traced. What is not finished is the data behind it: domestic Indonesian press has counted towards attention share since September 2026, but its sentiment is not yet served, so the Resonance pillar’s sentiment term is measured for only a handful of artists; and follower series exist for roughly two thirds of the first hundred. Both are being filled at the source, not estimated here.
We would rather show you nothing than a number you cannot check. Unscored is always acceptable; a wrong number never is.
Corrections are part of the pipeline, not an afterthought. Every fact here is meant to be editable. Send a correction →