African music’s chart share did not rise: 61,000,142 frozen India streams moved the denominator

On eleven consecutive collection dates this desk recorded Spotify’s India daily chart as carrying exactly 61,000,142 streams. On the nine dates that followed it recorded exactly 0. Neither number is a day’s listening, and because India’s chart carries no African artist at all, both went straight into the denominator of the African chart share series this desk publishes.

Correct for it and the headline of the past month reverses. Measured against the charted pool as published, African music’s share of Spotify’s charted streams rose across the window, by between 4.475 per cent and 6.457 per cent depending on which single collection date you drop. Measured against the same seventy-four charts with India removed on every date, present or absent, it fell, by between 1.916 per cent and 3.949 per cent. The two measures disagree about the direction on all twenty recomputations. Treat India consistently and the direction is down whichever way you treat it.

What the India chart actually recorded#

This desk still holds the raw chart snapshots for twenty collection dates, 22 September to 17 October 2026. India’s chart appears on every one of them with 200 entries recorded, which is a full Top 200. The Streams column is what broke.

Sponsored

The repeat is the tell. A daily stream count that is identical to the final digit across eleven separate collections is not a measurement. The cumulative total printed against India’s number one entry says the same thing from the other side: it reads 28,342,366 on all twenty dates, including the nine where the daily column reads 0. A chart whose cumulative totals never advance is a chart that is not being updated.

Then the column emptied. Across the panel of 75 country charts this desk collects, the number carrying a Streams column that sums to zero was 0 on every date from 22 September to 4 October, and exactly 1 from 5 October onward. That one is India, on every single date.

Before blaming the collector, the operating rule here is to read the upstream page. India’s daily page was refetched on 17 October 2026 and returns HTTP 200 with the same column header as every other country page and a zero in every Streams cell. Nigeria’s page, fetched the same minute, returns real values: Seyi Vibez alone accounts for 1,890,131 chart streams on the 17 October panel. So this is not a parser reading the wrong table cell. The source publishes zeros for India, and the collector faithfully stored them.

Why it only moved the denominator#

Every entry on all 75 charts was matched by Spotify artist ID against this desk’s origin roster. African streams inside the India chart came to 0 on all twenty dates, which is the whole reason this is a denominator problem and not a numerator one. That is not a gap in the roster either: every distinct lead artist on India’s chart resolves to a registered origin, and none of those origins is African. The African streams themselves barely moved: 18,573,383 on 22 September against 18,045,294 on 17 October.

The pool underneath them did move, and it moved because India left it. On 22 September the pool as published read 715,048,251 charted streams; with India out it reads 654,048,109. That difference is the entire correction, and it is why the share reads 2.5975 per cent as published on 22 September and 2.8398 per cent once India is removed. By 17 October both readings converge on 2.5776 per cent, because India had already gone to zero and the two denominators are the same number.

Put the two series side by side and the as-published line jumps on 5 October, the day India’s column emptied, by more than the India-removed line does. The India-removed line moves that day too, but only by about the amount it moves on an ordinary day. The difference between the two jumps is precisely the 0.2032 points the India chart had been taking off the measured share on 4 October, the last day it reported anything. On 5 October that gap closes to 0 and it stays there.

It is not the panel that changed size#

The obvious alternative reading is that this desk’s measuring instrument shrank and the pool shrank with it. It did not. Across the same twenty dates the number of country charts collected was 75 on the lowest date and 75 on the highest, a range of 0 per cent. Chart entries captured ran from 14,355 to 14,493, a range of 0.961 per cent. Total charted streams ran from 648,909,299 to 825,092,170, a range of 27.151 per cent. The instrument held still to within one per cent while the thing it was measuring moved twenty-seven.

Attributing that pool contraction market by market, from the series high on 27 September to the series low on 6 October, India is the largest single contributor at 34.62 per cent of the fall. The United States is second at 19.76 per cent, Brazil third at 7.19 per cent and Mexico fourth at 7.15 per cent. A country chart that stopped reporting outweighed the real softening in the largest streaming market on earth.

The counter-reading#

Three things could break this finding, and one of them half does.

First, the reversal could be carried by a single collection date. It is not. Dropping each of the twenty dates in turn and comparing the later half of the window against the earlier half, the India-removed series falls on every one of the twenty recomputations, mean -2.6146 per cent. Excluding the one reading closest to zero it is -2.6514 per cent, excluding three -2.7289 per cent, excluding five -2.8109 per cent. The largest single row accounts for 3.66 per cent of the total and the worst scenario moves the mean by 7.51 per cent. Nothing here hangs on one day.

Second, 61,000,142 could be a real number that India genuinely produced on eleven consecutive days. If that were true, the pool as published would be correct and removing India would be the distortion rather than the fix. The frozen cumulative total of 28,342,366 is what rules it out. Daily figures can in principle repeat; a running career total cannot stand still for twenty days while the chart it sits on is supposedly being refreshed.

That objection is worth taking seriously enough to price, though, because it has a testable consequence. There are three defensible ways to treat the India chart, and only one of them produces a rise. Take the same numerator on the two endpoint dates and change nothing but the denominator. As published, which counts India at 61,000,142 on the first eleven dates and at zero on the last nine, the share goes from 2.5975 per cent to 2.5776 per cent, a change of -0.765 per cent, and it is this inconsistency that turns into an apparent rise once you compare half of the window against the other half. Remove India on every date and the pool goes from 654,048,109 to 700,073,676 and the share falls -9.231 per cent. Keep India in on every date at 61,000,142, which is the treatment you would use if you believed that figure, and the pool goes from 715,048,251 to 761,073,818 and the share falls -8.719 per cent.

So the finding does not depend on believing the zeros or disbelieving the 61,000,142. Treat India consistently either way and African chart share fell by between eight and nine per cent over the window. The only treatment that makes it look like growth is the one where the market is in the denominator for part of the period and out of it for the rest, which is the treatment this desk was using.

Third, and this is the half that bites: the pool contraction is not all India. Drop India from the market attribution and 34.62 per cent of the fall goes with it, which leaves 65.37 per cent that is real softening across the United States, Brazil, Mexico and most of the rest of the panel. Drop the top three and 61.58 per cent of the fall disappears. The honest statement is that India corrupted the direction of the share measurement, not that it invented the whole pool movement.

The limit that cannot be fixed: only twenty of the forty-five readings in this desk’s presence series can be audited this way, because the raw snapshots behind the other twenty-five have been pruned. Everything before 22 September 2026 is uncorrectable with what is on disk. This correction applies to the window it covers and no further, and anyone quoting an African share figure from this desk dated before 22 September should treat the denominator as unaudited.

What an artist or a label should do with this#

Stop pricing anything off a share. A share has two halves and you only control one of them. African charted streams went from 18,573,383 to 18,045,294 over this window, a small absolute fall, and the as-published share rose anyway because a market with no African artists on its chart stopped reporting. If you had used that rising share to argue for a bigger advance or a better split, you would have been arguing from an artefact.

The number to carry into a negotiation is the absolute one, in the market you actually sell in. Seyi Vibez’s 1,890,131 chart streams on the 17 October panel do not change when a page in another country breaks. A percentage of a global pool does. When a distributor, a platform or a trade report hands you a share figure and a direction of travel, the only question worth asking is what the denominator did, and whether anyone checked.

For this desk the practical consequence is a method change, stated here so it is on the record: India is removed from the charted pool on every date from 22 September 2026 onward, present or absent, so the same seventy-four charts are compared throughout. The series will read lower and it will read right.