56.7 Per Cent of New Entries on Africa’s Daily Music Charts Are Records That Had Already Charted There. On Spotify’s Four African Charts It Is 76.9 Per Cent

Across 337,686 new-entry flags on seven published daily music chart feeds in 42 African markets, 191,632 of them, 56.7 per cent, were records that had already charted on that exact chart earlier in the same collection window. Only 146,054 were records this desk had never seen on that chart before. The measurement is a slot-level count over 86 collected chart dates between 16 August and 10 November 2026: every arrival on every chart was checked against every earlier observation of the same record identifier on the same chart in the same market.

Pooled across the panel, a new entry on an African daily chart is therefore more likely to be a record coming back than a record arriving. That is a statement about the pool and not about every chart in it: the majority holds in 34 of the 42 markets and on 4 of the 7 feeds, and it does not hold in the other 8 markets and 3 feeds. Where this piece says African charts, it means these 337,686 arrivals taken together.

Seven fixed-length daily chart feeds, 42 African markets, 16 August to 10 November 2026. Apple Music albums 63.4 per cent, Shazam 32.3 per cent.

Sponsored

Show the numbers

Chart feed	Returning, % of new-entry flags
Spotify songs	76.90
Apple Music albums	63.40
Apple Music songs	59.20
Apple Music music videos	57.10
YouTube songs	49.10
Deezer songs	47.40
Shazam songs	32.30

Source: This desk's own afw_chart_snapshot and afw_chart_entry collection of published daily African charts, retrieved 10 Nov 2026 · download the data

The desk’s own collection gaps do not explain it#

The obvious objection is that this is an artefact of how often the charts are collected. If a record sits near the bottom of a chart on Monday, the chart is not collected on Tuesday, and the record is still there on Wednesday, a naive comparison against the last stored reading would call it absent and then new. That mechanism is real, and it is measurable.

It accounts for 26,397 of the returns, or 13.8 per cent. The other 165,235, or 86.2 per cent, arrived on a chart that this desk had also collected on the immediately preceding calendar day. For those, the record was genuinely on the chart, genuinely off it the next day, and genuinely back later. That is real chart movement inside a single day, not a hole in the data. Restricting the whole calculation to only those flags where yesterday was collected leaves the finding at 55.8 per cent, against 56.7 per cent for everything.

The gap-driven returns cluster where you would expect. Apple Music’s three feeds each cover 41 markets and cannot all be collected every day, and between 83.0 and 84.6 per cent of their returns sit on a contiguous day. Deezer, Spotify, Shazam and YouTube cover 9, 4, 15 and 24 markets, are collected far more completely, and between 96.0 and 98.5 per cent of their returns sit on a contiguous day.

The revolving door is worst at the bottom, but it is not only at the bottom#

Sorting every flag by how deep into its own chart it landed, as a share of that chart’s published slot count so that a 20-slot feed and a 200-slot feed can be compared, gives a clean gradient. In the bottom tenth of the chart, 36,853 of 61,379 new entries, 60.0 per cent, are records returning. In the top tenth it is 6,679 of 15,395, or 43.4 per cent.

That 43.4 per cent is the number worth sitting with. Even at the very top of an African daily chart, where a genuine arrival is supposed to be an event, closer to half of all arrivals are records that have been there before. The revolving door is not a bottom-of-the-chart curiosity. It runs the length of the chart.

South Africa 69.8 per cent, Burkina Faso 39.0 per cent#

The spread across markets is wide and it does not track market size. South Africa has the highest rate of any market measured: 5,715 of its 8,182 new-entry flags, 69.8 per cent, are records returning. Benin is at 67.1 per cent, Gabon 66.9 per cent and Congo-Brazzaville 66.1 per cent. At the other end sit the eight markets where returns are not a majority at all. Burkina Faso is at 39.0 per cent, Tunisia, which produced more new-entry flags than any other market at 16,253, is at 42.4 per cent, Kenya is at 45.0 per cent and Egypt at 49.9 per cent. Nigeria, at 52.8 per cent, is only just on the other side of the line.

By feed, Spotify is the outlier, and it is the outlier in the direction that matters most to anyone quoting its charts. Of 4,991 new-entry flags across the four African markets Spotify publishes a daily chart for, 3,837, or 76.9 per cent, are records returning. That leaves 1,154 first-time arrivals on Spotify’s African charts across the entire 86 collected dates. Shazam is the lowest at 32.3 per cent of 14,025 flags, and Deezer at 47.4 per cent and YouTube at 49.1 per cent also fall below half. Shazam’s position fits what Shazam measures: identifications of a record someone has just heard somewhere else, not repeat plays of a record already in rotation. It is the one feed here that is closer to a discovery signal than to a rotation signal, and it is the one feed where arrivals really are mostly arrivals.

The counter-reading#

Three things could break this finding, and two of them are testable.

The first is feed selection. Two of the desk’s nine daily chart feeds were deliberately left out, and the exclusions do work. The desk’s own composite feed is assembled from the other feeds and would double count. The iTunes download feed publishes a chart whose length varies between 47 and 99 slots, so a record’s absence cannot be separated from the chart simply being shorter that day. Removing any one of the seven retained feeds moves the pooled share to between 54.8 and 58.3 per cent.

Those exclusions also flatter the number, and by how much is worth stating rather than asserting that the reasoning was sound. Putting the iTunes feed back, despite the slot problem, takes the share down to 54.4 per cent. Putting the composite feed back takes it up to 57.3 per cent. Reversing both and pooling all nine feeds gives 55.3 per cent of 436,922 arrivals. The published figure of 56.7 per cent is the highest of those four, so a reader who thinks the exclusions were wrong should read the finding as somewhere between 54.4 and 56.7 per cent. It stays a majority either way.

The second is market concentration. Removing any one of the five largest contributing markets moves the share to between 56.8 and 57.5 per cent. Removing all five at once moves it up, to 59.0 per cent, which is the opposite direction from the one that would flatter the finding.

What the robustness run on the raw count does show is that the volume is concentrated: Apple Music’s music-video feed alone holds 38.8 per cent of all returning records, and dropping Apple’s three feeds together removes 83.0 per cent of the count. That concentration is a fact about where the desk collects, not about the rate. The rate is what the finding rests on, and the rate survives every removal.

The third objection is the one that cannot be tested, and it runs against the desk rather than for it. A record that was charting before 16 August, fell off, and came back during the window is counted here as first-time, because there is no earlier observation of it. Every such record is a return misfiled as an arrival. So 191,632 is a floor and 146,054 is a ceiling. For the headline to be wrong in the other direction, the service feeds would have to be reporting records as absent while they were in fact still on the chart, on 165,235 occasions, on days the desk collected that chart both before and after. Nothing in the collection suggests that.

What an artist or a label should do with this#

Three concrete things.

Falling off a chart is not an ending, and how much of an ending it is depends on where. A campaign wound down the week a record drops out is wound down on a signal that reverses itself 56.7 per cent of the time across the pool. In South Africa that rate is 69.8 per cent, so a Johannesburg label treating a chart exit as final is making the worst version of that mistake. In Tunisia it is 42.4 per cent and in Burkina Faso 39.0 per cent, where an exit genuinely is more often an exit. The number to use is your own market’s, not the pooled one.

A chart run is not a run. Measuring a record’s chart life as a single continuous stretch of days will understate it, because the stretch is broken. Anyone reporting chart longevity to a distributor, a sync agent or an investor should count distinct days present, not first day to last day, and should say which they used.

And treat “new on the chart” in a press release as a claim that needs a date. It is accurate 43.3 per cent of the time at face value across these feeds, and 23.1 per cent of the time on Spotify’s African charts. The useful question to ask of any chart-entry announcement is not whether the record charted, but whether it had charted there before.

One limit applies to all of it. These are chart slots, not streams and not consumption. A record returning to a chart tells you it crossed that market’s threshold again; it does not tell you how much it was played, and the four services here that publish a stream or view count do so only for some markets. Chart presence is the thing measured, and chart presence is all that is claimed.