African-origin artists’ share of Spotify’s daily chart streams has averaged 1.919 per cent across the eight collection dates between 25 August and 4 September 2026, down 9.2 per cent from the 2.113 per cent it averaged across the six dates before that, 13 to 24 August. This is measured across every entry on every daily chart Spotify publishes for the 75 markets this desk tracks, crediting each entry’s streams to its lead artist’s country of origin, dividing African-origin streams by that day’s total collected chart streams, and averaging the resulting daily ratio, unweighted, within each period.
Unweighted daily ratio of African-origin chart streams to total chart streams collected, across all 75 tracked markets. Fourteen collection dates; gaps between points are missed collections, not zero days.
Show the numbers
| Date | African share of charts, % |
|---|---|
| 2026-08-17 | 2.19 |
| 2026-08-13 | 2.15 |
| 2026-08-22 | 2.12 |
| 2026-08-20 | 2.09 |
| 2026-08-21 | 2.07 |
| 2026-08-24 | 2.06 |
| 2026-08-27 | 1.97 |
| 2026-08-31 | 1.93 |
| 2026-08-25 | 1.93 |
| 2026-09-02 | 1.92 |
| 2026-08-28 | 1.92 |
| 2026-08-30 | 1.92 |
| 2026-09-04 | 1.92 |
| 2026-09-01 | 1.85 |
Source: kworb.net Spotify daily country charts, artist origin from MusicBrainz, retrieved 4 Sep 2026 · download the data
What actually moved#
The total pool of chart streams this desk collects each day is not fixed. It ranged from 708.7 million to 780.97 million across the 14 dates measured, a swing that has nothing to do with African music and everything to do with how much the whole panel streamed that day. Share, not the raw African stream count, is what strips that noise out, and share is what moved.
| # | Period | Dates | Collection dates | Mean African share, % | Change vs Period 1, % |
|---|---|---|---|---|---|
| 1 | Period 1 | 13-24 Aug 2026 | 6% | 2.11% | 0% |
| 2 | Period 2 | 25 Aug-4 Sep 2026 | 8% | 1.92% | -9.20% |
Source: Computed from this desk's african-presence-over-time dataset, 14 collection dates between 13 August and 4 September 2026, retrieved 4 Sep 2026 · download the data
The drop is not a single bad day carried across an average. Every one of the eight dates from 25 August onward sits below every one of the six dates before it: the highest reading in the second period, 1.969 per cent on 27 August, is still lower than the lowest reading in the first, 2.056 per cent on 24 August. That is a level shift, not one collection date dragging a mean down.
The African-origin population behind this measurement is dominated by one artist on any given day: Asake, the largest single African-origin contributor to the 4 September panel, with 1,938,116 daily streams inside Nigeria alone, 13.6 per cent of that day’s entire 14,264,672-stream African total. The share reported here weights every stream equally, by summing them all before dividing, so a swing in one dominant artist’s daily volume can move the ratio on its own even when the count of artists charting does not change.
It did not happen because fewer African artists were charting. The artist count actually rose across the first period, from 92 names on 13 August to 100 on 24 August, and has held in a 98-to-101 band ever since, essentially flat. What changed is how many markets those artists reached on a given day: markets_charting ran 11 to 12 across the six dates through 24 August and 9 to 11 across the eight since, including three dates, 25 August, 1 September and 4 September, at the low end of the whole 14-date series. More artists are charting somewhere; fewer of them are charting in as many places at once.
The counter-reading#
A period comparison invites an obvious objection: 25 August is one boundary among 13 possible ones in a 14-date series, and a different cut point might tell a different story, or no story at all. It does not. Splitting the series at every one of the other 12 possible points, not just the one used above, produces a decline every single time, ranging from 4.6 per cent (splitting after 2 September, which leaves only the final date in the later block) to the 9.2 per cent reported here (splitting after 24 August). There is no split point anywhere in this series that produces a rise.
| # | Split after | Dates before | Mean before, % | Dates after | Mean after, % | Change, % |
|---|---|---|---|---|---|---|
| 1 | 2026-08-13 | 1% | 2.15% | 13% | 1.99% | -7.54% |
| 2 | 2026-08-17 | 2% | 2.17% | 12% | 1.98% | -8.97% |
| 3 | 2026-08-20 | 3% | 2.14% | 11% | 1.96% | -8.35% |
| 4 | 2026-08-21 | 4% | 2.13% | 10% | 1.95% | -8.09% |
| 5 | 2026-08-22 | 5% | 2.13% | 9% | 1.93% | -8.97% |
| 6 | 2026-08-24 | 6% | 2.11% | 8% | 1.92% | -9.20% |
| 7 | 2026-08-25 | 7% | 2.09% | 7% | 1.92% | -8.09% |
| 8 | 2026-08-27 | 8% | 2.07% | 6% | 1.91% | -7.85% |
| 9 | 2026-08-28 | 9% | 2.06% | 5% | 1.91% | -7.15% |
| 10 | 2026-08-30 | 10% | 2.04% | 4% | 1.91% | -6.62% |
| 11 | 2026-08-31 | 11% | 2.03% | 3% | 1.90% | -6.64% |
| 12 | 2026-09-01 | 12% | 2.02% | 2% | 1.92% | -4.91% |
| 13 | 2026-09-02 | 13% | 2.01% | 1% | 1.92% | -4.63% |
Source: Computed from this desk's african-presence-over-time dataset, 14 collection dates, 13 August to 4 September 2026, retrieved 4 Sep 2026 · download the data
The 24/25 August boundary used in the body is not arbitrary, either. It sits on the single largest day-to-day drop anywhere in the series, a fall of 0.129 percentage points between 24 and 25 August against a next-largest day-to-day move of 0.095 points, and it is one of only three points in the series where every reading on one side sits below every reading on the other. But it is also the split that produces the largest magnitude of the 13 available, 9.2 per cent, so the honest statement is the range, not the single most striking number inside it: the decline holds everywhere between 4.6 and 9.2 per cent, whichever boundary is used.
Two limits stand outside any test. First, collection is irregular: 14 dates span 23 calendar days with gaps of one to four days between them, so a real short-lived event on an uncollected date would not appear in either period’s average. Second, a chart entry is a sample of consumption, not a full accounting; an artist’s streams below a market’s charting cutoff, or on services this panel does not cover, are invisible to this measurement regardless of which way the charted share moves.
What this means for artists and independent labels#
Do not budget against the higher band. Any projection built on African music’s share of Spotify’s global daily charts sitting near 2.1 per cent, the level held through most of the first three weeks of this desk’s tracking, has been wrong for eight straight measured dates and eleven calendar days running. The current, sustained level is closer to 1.9 per cent, and that is the number to plan around until the series says otherwise.
Watch reach, not headcount. The artist count charting at all barely moved, 98 to 101 versus 92 to 100, while the markets those same artists reached per day fell. An independent label tracking its own roster’s international footprint should count markets charted per week, not just whether a track is charting somewhere, because this data shows the second can hold steady while the first quietly narrows.
Re-check this measurement, do not re-quote it. This desk will keep extending the same series. A reader citing the 1.919 per cent figure in three months should pull the current african-presence-over-time dataset rather than repeat this number, because a longer series is the only way to tell a sustained level shift from a longer cycle this 14-date window is too short to see.
