Guinea-Bissau’s iTunes Download Chart Had 2 Number Ones in 70 Days. Across 17 African Markets, 90.94 Per Cent of Download Slots Never Reach Streaming

Across 17 African markets where this desk collects both of Apple’s daily song charts, 79,190 of 87,082 iTunes Store Top Songs slots, 90.94 per cent, were held by a record that never appeared on the same market’s Apple Music streaming chart on any date in the window measured. Both charts are published by the same company, for the same countries, on the same days. One ranks paid downloads and the other ranks streams, and in Africa they almost never describe the same record.

The measurement is this desk’s own daily collection of both of Apple’s public feeds, 16 August to 6 November 2026, which gives 1,153 paired market-days across Botswana, Burkina Faso, Cabo Verde, Egypt, Eswatini, Gambia, Ghana, Guinea-Bissau, Kenya, Mauritius, Mozambique, Namibia, Niger, Nigeria, South Africa, Uganda and Zimbabwe. Every iTunes slot inside those paired days is checked against that market’s full Apple Music songs chart on every date the store holds, matching on a normalised title plus credited artist. A slot counts as never-charting only if it is absent from the market’s streaming chart on all of those dates, not merely on its own day.

17 African markets, 1,153 paired market-days, 16 August to 6 November 2026.

Sponsored

Show the numbers

Market	Never-charting, %
Niger	97.57
Namibia	96.95
Cabo Verde	96.31
Mozambique	96.07
Kenya	95.67
Gambia	95.29
Uganda	94.66
Guinea-Bissau	91.72
Ghana	91.03
Mauritius	90.82
Botswana	89.77
Zimbabwe	88.85
Egypt	88.02
Eswatini	86.78
Nigeria	86.72
South Africa	82.17
Burkina Faso	75.66

Source: This desk's own daily collection of Apple's two public chart feeds (afw_chart_snapshot / afw_chart_entry): the iTunes Store Top Songs RSS feed, which ranks paid downloads, and the Apple Music Top Songs feed, which ranks streams, retrieved 7 Nov 2026 · download the data

The gap is in every market, and it is not a chart-tail effect#

No market on the panel comes close to agreement between the two charts. Niger is the widest at 97.57 per cent never-charting and Burkina Faso the narrowest at 75.66 per cent, with South Africa next at 82.17 per cent, Nigeria at 86.72 per cent and Egypt at 88.02 per cent. The obvious objection is that a download chart’s lower reaches are thin and would have nowhere to land on a streaming chart, so the same test was run on the top 10 of each day’s iTunes chart only: 10,048 of 11,530 of those slots, 87.15 per cent, still never reach the paired streaming chart. The records being bought most in these countries on a given day are, with few exceptions, not the records being streamed most. That panel figure does hide a real split, and it runs against the finding: in the two largest download markets the top-10 gap is much narrower, 66.03 per cent in South Africa and 66.09 per cent in Egypt. The further up the chart and the larger the market, the more the two instruments agree, and nowhere do they agree on most of it.

It goes deeper than individual titles missing a chart. Of the 79,190 never-charting slots, 70,261, or 88.72 per cent, are credited to an artist who appears nowhere on that market’s Apple Music chart across the entire window, under any title. This is not a catalogue cut slipping below a threshold while its artist charts normally. These are whole artists who exist on the download chart and nowhere on the streaming one.

The named cases show what that population looks like. Foo Fighters’ “Sean” held the iTunes number one position in Guinea-Bissau on 61 separate days in this window and in Gambia on 26, without once entering either country’s Apple Music songs chart; the same band’s “Saint Cecilia” led Ghana’s download chart on 17 days. Vangelis’s “Conquest of Paradise” took 10 days at number one in Cabo Verde and Tracy Chapman’s “Baby Can I Hold You” 10 in Niger. The clearest African case is Nigeria, where Tems held the download number one with “Me & U” on 14 days and never appeared on Nigeria’s Apple Music streaming chart at all.

In five of the 17 markets, the download chart barely moves#

Before reading that gap as a hidden audience, it is worth asking whether the download chart in these countries is moving at all. It largely is not. Measuring day-over-day carryover, the share of each day’s chart that was also present on the previous collected day, Guinea-Bissau’s iTunes chart returns 99.89 per cent. Two distinct records held its number one position across 70 paired days. Niger returns 99.42 per cent, Cabo Verde 98.68 per cent, Gambia 98.57 per cent and Eswatini 96.47 per cent. These are not charts responding to daily purchasing. They are near-frozen lists.

The same markets’ streaming charts churn normally. Guinea-Bissau’s Apple Music chart carries over just 26.44 per cent of its entries day to day against its download chart’s 99.89 per cent. At the other end of the panel the download chart is genuinely active: Nigeria returns 13.88 per cent carryover, Egypt 14.01 per cent and South Africa 26.58 per cent, all below their own streaming charts.

The obvious suspicion is that a near-frozen chart is this desk’s own collection failing, serving the same cached payload back every day. It is not. Every snapshot was fingerprinted twice, once on its titles in position order and once on its titles alone with position discarded. A stale payload would repeat the same ordering; a thin but live storefront re-ranks the same few titles. Guinea-Bissau produced 8 distinct orderings from 80 snapshots but only 5 distinct chart memberships, so the feed is updating and re-ranking while the same small set of titles keeps clearing Apple’s reporting threshold. In 10 of the 17 markets, including Nigeria, Egypt and South Africa, every snapshot returned a distinct membership, which is what a working collector looks like. The staticness is a property of those storefronts, not of the collection.

The counter-reading: frozen storefronts do not explain the gap#

If the never-charting share were an artefact of dead storefronts, it should collapse in the markets where the download chart actually moves. It does not. Grouping the 17 markets into carryover bands, the five near-frozen markets average 93.53 per cent never-charting, the nine middling ones 91.05 per cent, and the three actively churning ones 85.64 per cent. That is a spread of 7.89 percentage points across the entire range of the variable, and the association across all 17 markets is weak: Pearson r of 0.35 and Spearman rho of 0.38. Staticness widens the gap slightly. It does not create it. The honest statement is that this desk tested the most plausible alternative explanation and could not make it account for the finding.

Nor does any single market carry the total. Namibia, the largest contributor, is 7.9 per cent of the never-charting count, and removing it moves the total by the same 7.9 per cent. Dropping the top three takes 22.55 per cent out and dropping the top five takes 36.43 per cent, which is close to what removing five of 17 rows should do.

The remaining threat is matching error, since two feeds crediting the same recording differently would read as absence. Two stricter passes test it. Stripping “feat”, “ft” and “featuring” credit tails from both titles and artists moves the figure from 90.94 per cent to 90.92 per cent. Discarding the artist string altogether and matching on title alone, which is the most generous possible match and will wrongly merge different recordings that share a name, gives 90.48 per cent. The whole of the artist-credit risk is therefore bounded at 396 slots, 0.45 percentage points, against a finding of 90.94 per cent.

Three limits belong on the page rather than in a footnote. First, this chart store begins on 16 August 2026 and holds nothing earlier, so “never” can only mean never inside this window; a back-catalogue record that charted on Apple Music in July and then fell off is counted here as never-charting, which inflates the figure by an amount this store cannot measure. That matters most for exactly the kind of legacy repertoire named above. Second, Burkina Faso has only 36 paired days against 64 to 77 for every other market, and it is an outlier on two further measures, so its 75.66 per cent should be read as the least certain row on the panel; removing it entirely leaves 76,809 of 83,935 slots never-charting, 91.51 per cent, which moves the headline up rather than down. Third, an earlier build of this dataset computed the never-charting count as the same-day complement while describing it as a window-wide test. That overstated it by 3,162 slots, 3.63 percentage points, and the figures here use the corrected window-wide test throughout.

What an artist or label should do with this#

Treat the download chart as a readable second demand signal in three markets and ignore it in five. In Nigeria, Egypt and South Africa the iTunes chart turns over fast enough to reflect actual purchasing, and there the finding has a direct use: a record’s absence from the streaming chart is not evidence that nobody is paying for it, as 14 days of Nigerian download number ones for a title with no Nigerian streaming chart presence demonstrates. In Guinea-Bissau, Niger, Cabo Verde, Gambia and Eswatini, a download chart position is close to meaningless as a measure of this week’s demand, because the same records hold the same positions almost every day. An artist deciding where to push a paid release, or a label reading a distributor’s territory report, should know which of the two the market in front of them is. Two different things are being separated here and the distinction matters commercially: the feed itself is live in 10 of the 17 markets, which returned a distinct chart membership on every snapshot, but only 3 turn over fast enough for a position to mean anything about this week.