Deezer’s African Number One Turns Over on 80 Per Cent of Days. Shazam’s Turns Over on 10.9425 Per Cent. Same 8 Feeds, Same Window, 7.3109 Times Apart

Deezer’s number one position in the nine African markets this desk tracks turned over on 80.0 per cent of the days it was observed between 16 August and 9 November 2026, a mean unbroken reign of 1.25 days. Shazam’s number one in its 15 African markets turned over on 10.9425 per cent of days observed in the same window, a mean reign of 9.1387 days. That is a 7.3109-fold gap between the steadiest and the most volatile of the 8 chart feeds this desk collects daily across Africa, measured the same way, over the same 12-week window.

Share of observed chart-days on which the title or artist at position 1 differed from the previous day. Sorted from steadiest to most volatile.

Show the numbers

Feed	Churn rate at #1, %
Deezer (songs)	80
iTunes (songs)	64.58
Apple Music (music-videos)	58.08
AFW composite	35.32
Apple Music (albums)	27.46
Apple Music (songs)	26.66
YouTube (songs)	15.25
Spotify (songs)	12.95
Shazam (songs)	10.94

Source: This desk's own chart collection (afw_chart_snapshot / afw_chart_entry), retrieved 9 Nov 2026 · download the data

Sponsored

This desk runs a daily collector against 8 source-and-chart-type combinations that between them cover African markets: Spotify’s songs chart in 4 markets, Apple Music’s songs, albums and music-videos charts in 41 markets each, Shazam’s songs chart in 15 markets, Deezer’s songs chart in 9 markets, iTunes’ songs chart in 17 markets, and this desk’s own composite ranking, which blends whichever feeds are available per market, in 42 markets. Every one of those feeds was captured across the same calendar window, 16 August to 9 November 2026, 85 dates on the calendar though Apple’s three chart types were actually collected on 83 of them in every market, a 2-day gap the other feeds did not have. Pulling every row at position 1 across 17,702 snapshot-days and counting a change whenever the (title, artist) pair differs, case-insensitively, from the day before gives market_days, changes, churn_rate_pct and mean_reign_days for each feed, pooled across all of that feed’s markets.

The gap is not a coincidence of which markets each feed covers. Nigeria is the one market all 8 feeds publish a chart for, and the ordering holds with or without it: iTunes and Deezer churn fastest, Shazam and Spotify churn slowest, in every market they share. What differs structurally between the feeds is what each one is actually counting. iTunes is a paid-download chart, the only one of the 8, and a download chart moves the moment one new purchase spike outpaces yesterday’s cumulative count: its number one changed on 64.5848 per cent of the market-days observed, and the position was held by 716 different titles across 1,409 market-days. Shazam counts recognition queries, people holding a phone up to a song they do not already know, which is a slower-building signal: its number one changed on only 10.9425 per cent of market-days, held by 121 different titles across 1,252 market-days, a mean reign of 9.1387 days, more than a week.

This desk’s own composite feed sits in the middle at 35.3238 per cent churn, and that is explained by its construction rather than by African listening itself: the composite is built by taking whichever underlying feed is available in a market and ranking from it, so it inherits volatility from every feed that touches it. It is more stable than iTunes, Deezer or Apple’s music-videos chart, all of which it sometimes draws from, and less stable than Spotify or Shazam, the other two it sometimes draws from. A single blended African panel number, on its own, obscures which instrument actually produced the figure on any given day.

Apple Music’s own three chart types, collected from the same 41 markets on the same dates, split apart by more than 30 percentage points depending only on which content type is ranked: 26.6574 per cent churn on songs, 27.4599 per cent on albums, and 58.0815 per cent on music-videos, which churned through 676 distinct number ones, more than any other feed, in fewer market-days than Apple’s own songs chart. The population and the collection method are identical across all three; only the thing being counted changes.

That instability in the pooled average is itself the counter-reading. Averaging churn_rate_pct across all 9 rows and then dropping Deezer alone moves the mean by 14.67 per cent; dropping the three fastest-churning feeds, Deezer, iTunes and Apple’s music-videos chart, moves it by 41.77 per cent; dropping the five fastest moves it by 55.3 per cent. A single “average churn across Africa’s charts” figure is not a stable description of anything: it is mostly a description of how many paid-download and video feeds happen to be in the average that week. The honest statement is the spread, not the mean. A second check, recomputing every churn figure under a stricter string match that strips punctuation and feat. credits before comparing titles, moves every feed’s churn rate by less than 1 percentage point in either direction, so the spread itself is not an artefact of cosmetic title formatting day to day.

The limits worth stating plainly. This measures position continuity, not the stream or download volume underneath it, so a steady number one is not necessarily a bigger number one, only a longer-held one. The comparison uses exact title-and-artist string identity rather than a stable catalogue identifier, which slightly overstates churn on any feed where the same recording is occasionally printed with a different feat. credit or punctuation from one day to the next; the sensitivity check above bounds that risk at under 1 percentage point, which is small against the 7.3109-fold gap the piece reports. The window is 12 weeks, short enough that a single unusually long or short reign inside it can move a feed’s mean materially, and the feeds cover between 4 and 42 markets each, so a feed with more markets had more independent chances to register a change on any given day even though churn_rate_pct is already normalised per market-day to correct for that.

For an artist or an independent label in Lagos, Nigeria, deciding where a single promotional push goes, the practical reading is that “reached number one” is not one claim, it is 8 different claims with 8 different shelf lives. Nigeria is the one market every one of the 8 feeds covers, so an act there can, in principle, top any of them. A number one on Shazam or Spotify is, on this window’s evidence, a credential that is likely to still be true in a press release written a week later. A number one on Deezer, iTunes or Apple’s music-video chart has a better-than-even chance of being overtaken before that release goes out, and should be dated to the day it happened rather than presented as a standing claim. A label stacking multiple “charted at number one” credits from different feeds in the same campaign should know those credits do not carry equal weight, and a reader checking a press release’s number-one claim against the date it was made is checking something this data says is worth checking.