African Artists Reached 7 of Spotify’s 74 Daily Charts on 14 November. Across 62 Readings the Mean Fell 15.19 Per Cent and the Median Did Not Move at All

African artists appeared on 7 of the 74 daily country charts Spotify published on 14 November 2026. Across the 62 readings this desk has collected since 13 August, the mean number of charts carrying any African artist fell 15.19 per cent, from 9.13 across the first 31 readings to 7.74 across the last 31. The median did not move at all: 8 charts in both halves. Nor did the number of African artists charting, which went from a mean of 111.32 to 111.87, a change of 0.5 per cent.

All three of those statements describe the same 62 readings, and together they are more specific than any of them alone. African music did not lose its typical Spotify chart footprint. It lost its best days. Every reading on which 10 or more charts carried an African artist sits between 13 August and 2 September, and nothing has come near that level since.

This desk has measured that breadth at single points before, on the first reading in August and again in September. What a 62-reading record adds is the shape between those points, and the shape is not the one a run of single-day readings implies.

Sponsored

What held flat and what did not#

Split the 62 readings into two halves of 31 by position in the series, which cuts the record in late September, and five of the six columns this desk tracks move by more than a percentage point. Only one of those five moves the way a growth story would predict.

The count of African artists charting is the flat one: 111.32 then 111.87, a gain of 0.55 artists. African music’s share of every charted stream on Spotify, home and away together, rose from 2.35 to 2.54 per cent, up 8.0 per cent. Chart breadth fell 15.19 per cent. The share of African chart streams earned outside Africa fell from 1.93 to 1.27 per cent, down 34.0 per cent, and the absolute export figure behind it fell from a mean of 339,092.94 streams to 230,674.58, down 32.0 per cent.

Put the flat column and the falling column together and you get the one derived number that describes what actually happened: African artists per charting chart rose from a mean of 12.5 to 14.51, up 16.0 per cent. The same population of artists, packed into fewer charts.

The mean is doing work here that a reader should see. Swap it for the median on the same two halves and the breadth column reports no change whatever: 8 charts in the first half, 8 in the second, 0.0 per cent. The artist count does the same, 112 and 112. Only the export share survives the swap, falling 33.87 per cent on medians against 33.99 per cent on means. So the breadth result is a statement about the upper tail of the distribution and not about its centre. The typical reading did not get worse. The good readings stopped happening.

The two columns do not merely diverge, they move against each other. Across the 62 paired readings the correlation between the artist count and the breadth count is −0.4903: on readings when more African artists charted, fewer charts carried any of them. Fitted against reading index, the artist column gains 0.05815 artists per reading while the breadth column loses 0.04918 charts per reading. The same exercise on the two share columns gives −0.6468. None of that is a causal claim and the readings are not independent of each other, but the sign is the point, and the sign is negative in both pairs.

The fall is a step, not a slope#

Breadth did not decline gradually across the record. The ceiling dropped once, at the start of September: no reading since 2 September has put an African artist on 10 charts or more. The lower bound stepped down once as well, and later, on 30 September, when 7 charts was recorded for the first time. No reading has gone below 7 since, and no value lower than 7 appears anywhere in the 62 readings.

The distribution says so plainly. Of the 62 readings, 37 recorded exactly 8 charts carrying an African artist, 59.7 per cent of the entire series, and 9 recorded 7, which is the series low and also the value on 14 November. The readings above that are almost all early: 5 readings at 9 charts, the last of them on 12 October; 4 at 10, the last on 2 September; 5 at 11, the last on 31 August; and 2 at 12, on 13 and 24 August.

Re-cut the series and the step shows up as a window effect. Discard every August reading, leaving 51, and the half-over-half fall shrinks from 15.19 per cent to 4.32 per cent. Discard August and September both, leaving 30 readings split into two halves by position, and the fall is not there at all: breadth rises 2.61 per cent, from a mean of 7.67 charts to 7.87.

The export column does the same thing. The share of African chart streams earned outside Africa falls 33.99 per cent across the full 62 readings, 22.97 per cent with August discarded, and then rises 1.44 per cent in the window that starts on 1 October, from a mean of 1.27 per cent to 1.29. Both of the columns that look like a decline in this record stopped declining at the start of October. Only the artist count runs the other way, and it runs the other way faintly: up 0.49 per cent across the whole record, down 1.95 per cent with August discarded, down 2.59 per cent from October onward.

Cut it shorter still. Measure the last 10 readings against the 10 before them, the shortest window with readings on both sides, and breadth is identical: a mean of 7.8 charts against 7.8, a change of 0.0 per cent. Over the same 20 readings the artist count moved 0.36 per cent down and the export share 7.25 per cent down. On the most recent 20 readings, breadth is the column that has stopped moving most exactly.

That is a correction to how this desk’s own recent reporting should be read. The export-share piece published on 8 November reported the export column contracting and the market count contracting with it, and over the window it used that was right. Four readings later, measured on a window that excludes the step, neither column is contracting. The direction of the full record and the direction of the last six weeks are not the same direction, and this desk reported the first without the second.

So a reader should take two statements from this panel and not one. African music’s chart footprint outside its own four markets is narrower than it was in August, by a margin that survives every cut that contains early September. And it stopped narrowing at the start of October. A piece that reported only the first of those, including this desk’s own, describes a trend the data says has already ended.

Which 7 charts, and what is on them#

Four of the seven are the African markets Spotify publishes a daily chart for: Nigeria, Egypt, Morocco and South Africa. They carried 18,394,633 of the 18,576,643 African-origin chart streams on 14 November. The other three charts are Saudi Arabia, the United Arab Emirates and Portugal, and between them they carried 182,010. The remaining 67 charts, 90.5 per cent of everything Spotify published that day, carried none.

The per-market detail is where the thinness becomes concrete. Nigeria’s own chart was 80.54 per cent African-origin on 14 November and Egypt’s 76.52 per cent, against Morocco’s 44.66 and South Africa’s 19.64. Outside Africa, Saudi Arabia’s chart was 11.43 per cent African on 148,354 streams, the United Arab Emirates 2.15 per cent on 15,023, and Portugal 0.41 per cent on 18,633.

Portugal’s 18,633 streams are one artist. Anselmo Ralph, the Angolan singer, is the entire African-origin presence on the Portuguese chart that day, and he charted on no African chart at all, which is why this desk’s own anomaly check flagged his row for review before publication. The flag is correct and the classification is correct at the same time: an Angolan act whose chart presence is in Lisbon rather than Luanda is exactly the exporter this panel exists to catch, and he is also 18,633 of the 182,010 streams in a continent’s entire measured export column on his own.

The counter-reading#

Four things could produce this result without the finding being real, and each was tested before publication.

The denominator could have moved. It did, by one chart, and that is the objection this finding most needed to survive. On 13 August this desk’s panel carried 75 published country charts, 12 of them with an African artist on them and 63 with none. On 14 November it carried 74, 7 with an African artist and 67 with none. The published panel shrank by one chart over the record while the breadth count fell by five, so coverage can account for at most one of the five and African records leaving charts that are still published accounts for the rest. France is the clearest single case: it carried 55,833 African chart streams on the 11 November reading, it is still on the panel today, and on 14 November it carried none. Nothing was removed from under it.

That same 13 August dataset carries one discrepancy a reader should see rather than have smoothed over. It recorded 91 African artists charting and 455,051 streams earned outside Africa on the day, of a 14,925,262 total. The series used throughout this piece restates 13 August as 93 artists and 380,704 export streams of a 15,831,536 total, because origin is applied to every date as currently registered rather than as registered at the time. The restated figures are the better ones and the gap between them is the cost of correcting a roster retroactively, which is the right thing to do and is not free.

The early readings could be outliers propping up the baseline. The five highest breadth readings in the whole series are all in August, and all five sit in the first half, so dropping them is the harshest available test rather than a neutral one. Run on the whole series, the breadth column sums to 523 charts across 62 readings for a mean of 8.4355; drop the single highest reading and the mean moves −0.69 per cent, drop the top three and −1.95 per cent, drop the top five and −3.08 per cent. Applied to the half-over-half comparison, dropping the first half’s five largest readings takes the gap from −15.19 per cent to −10.93 per cent. The fall is smaller than the headline number without those readings. It does not disappear. But the median test above is the more honest version of this objection, and it says the mean fall of 15.19 per cent is entirely a tail effect: strip the tail by using the median and the number is 0.0 per cent. Both figures belong in the same sentence, which is why they are in the same table.

The fall could be an artefact of where the window starts. It is, and that is the finding rather than an objection to it. Breadth and export share are both negative in every window containing early September and both positive in the window that does not. Anyone quoting the 15.19 per cent figure without the 2.61 per cent figure next to it is quoting half a result, and that includes this desk.

The classification could be wrong. Four rows on this run were flagged for an editorial decision before any figure here was written. Three were artists placed by a name match rather than by an identity link, which is the failure mode that once admitted a Filipino singer to this panel as South African: Egypt’s Lege-Cy, Morocco’s Tchubi and South Africa’s Mbuzeni. All three were checked against an outside source and against the markets they actually charted in, all three were confirmed at the country already assigned, and all three were re-registered against that evidence so they are no longer carried on a bare name match. The fourth was Anselmo Ralph. Nothing was reclassified, so no figure above moved. One of the three deserves naming as the weaker case: no biography giving Mbuzeni a birthplace was found, and the confirmation rests on a Zulu-titled maskandi single credited with a South African collaborator, reviewed by a South African outlet, charting only in South Africa. That is weaker evidence than a birthplace and it is recorded in the table as such.

What this does not measure matters more than any of the four. These are charted streams, not streams. An artist with a real and growing audience in London, Paris or Atlanta who sits below the last published position on those charts is invisible here by construction, and most African artists with overseas audiences are exactly there. The breadth column counts charts an artist reaches, not listeners an artist has, and a market can be growing fast for African music while contributing nothing to this count. Spotify publishes a daily chart for four African markets, so the home side of the panel is thin too, and the fifty or so African markets with no published chart cannot appear in the denominator or the numerator. Collection dates are not evenly spaced, so every figure here is a mean of readings rather than a mean of days. The claim should stay narrow: on the charts Spotify actually publishes, African music is reaching fewer of them than it did in August, and it stopped losing ground about six weeks ago.

What this is worth acting on#

For an independent label budgeting an international campaign, the usable number is 67. That is how many of Spotify’s published daily charts carried no African artist on 14 November, and the best reading in the whole record still left most of the board empty: the series high is 12 charts carrying an African artist, recorded twice, on 13 and 24 August. A chart entry outside Africa is not a reachable target on the evidence of 62 readings, and the three charts where it happened are not generic Western markets: they are two Gulf charts carrying Egyptian repertoire and one Portuguese chart carrying Lusophone repertoire. Every opening in the column was a language or diaspora bridge that already existed. If a release plan assumes a domestic hit converts into a chart position in a new market, this panel says that conversion happened on three charts out of 74 on the most recent reading, and that none of the three was a Western market.

The flat artist count is the more useful half for a working artist. The population of African artists reaching a published Spotify chart has been stable across the whole record, a mean of 111.32 in the first half of it and 111.87 in the second, which means the contest is not getting harder in terms of how many acts get in; it is getting narrower in terms of where. The thing that moved against African music in this record was not competition at home, where share rose 8.0 per cent, but the number of foreign charts on which any African record was visible at all. Acting on that means treating the four African charts and the two Gulf charts as the measurable board, and treating everything else as a market you will have to prove with your own platform data, because this panel cannot see it.

This desk will keep the panel running. The number to watch is not whether breadth falls below 7. It is whether the floor at 7 and 8 charts, which has held on 46 of the 62 readings, breaks upward for the first time since 12 October.