I really didn’t imagine, when I started this topic two weeks ago, that I’d end up making a three-post series about it. It was just supposed to be a silly little thing to make a week’s content on the blog, you know? Now it’s almost 6% of my output for the year. You didn’t come here to read about this, did you?1
Last week I showed you some graphs to demonstrate that the UK albums chart is different from when I was a kid, in that it has a lot more older stuff, and in particular a lot more greatest hits albums. I speculated that this might be the result of streaming, and asked you to think about how you’d count streams on an albums chart, because it’s not easy to come up with a way of doing it.
Fortunately, you can find the album charts rules on Google, so you can see exactly how the Official Charts Company do it—section 9, “Album Streams”, is the important bit for our purposes. You can tell they’ve really tried, bless ’em. The basic conversion is that a thousand streams of tracks from an album are considered equivalent to one sale of the whole thing, but they’re aware that that’s not perfect—you could game it by releasing an extremely long “album” of every song you’ve ever recorded, for one. So only the 16 most-streamed songs count. You could end up with an extremely popular song shooting its album to the top of the charts even if nobody listens to the rest. So they “neutralise” (their word) the two album tracks with the most streams to the average of the rest of the tracks,2 meaning this only becomes a problem if there are three huge songs at the same time (at which point it doesn’t seem unfair to credit the album with some of the success).
I think the bit that explains what’s going on here is a clause in Rule 9.3, which I’m going to quote in full:
Where a track appears on more than one album by an artist, streams of this track will be attributed equally to each studio album that has not charted in the top 40, a maximum of one studio album that has charted in the top 40 (the album to first chart in the top 40, or other studio album nominated in advance by label), and a maximum of one greatest hits album (the hits title with the highest sales DUS [Defined Universe Sales3] for that given week, or other hits title nominated in advance by label).
So suppose you stream the song “Dancing Queen”, which is ABBA’s most-streamed song on Spotify (2.2 billion streams). That appeared on ABBA’s fourth album, Arrival, which got to number 1 in 1977. That means that, for any other studio albums to be eligible (without displacing this one), they can’t have got into the top-40—although I think that’s irrelevant, because as far as I can tell this is the only studio album it was on. It has, however, been on several greatest hits albums, starting with 1979’s Greatest Hits Vol. 2. But only one of those is allowed to count; presumably, based on its chart position, that’s ABBA Gold.
So two albums are eligible for “Dancing Queen”, meaning each gets half the credit for streams of the song. Therefore, every 2,000 streams of “Dancing Queen” count as one sale of ABBA Gold. Well, not quite, because of that “neutralisation” thing—the top songs are recalculated every week, rather than being based on all-time streams, but it seems a fairly safe bet that they’ll be pretty well aligned. So that means extra plays of “Dancing Queen” and “The Winner Takes It All” are unlikely to boost ABBA Gold’s chart position.
But then we get down to “Gimme! Gimme! Gimme! (A Man After Midnight)”. Sorry, did you say “Some Maths After Midnight”? Well, why not. Now, this can only be an indicative example, because the current UK charts methodology hasn’t applied for all that time and doesn’t apply worldwide, but let’s do it anyway to get a sense of scale. “G!G!G!(AMAM)” is not as popular as “Dancing Queen”, but it’s still had 1.3 billion streams on Spotify. Again, ABBA Gold only gets credit half of them,4 but that still would be equivalent to 650,000 sales of the album. That might not sound like much, but that’s just from one track—and the total number of actual sales of the album (physical or download) worldwide, ever, is “only” 5.6 million.
So I think that’s why greatest hits albums are doing so well. Unless a song has been on multiple studio albums (and even then, unless many of them haven’t done that well), half of a song’s streams count towards a compilation album once it exists—and so albums specifically designed to be full of popular songs are naturally going to do better than ones with more “deep cuts”. And because those songs tend also to have enduring popularity, the line-up of artists popular enough to make the charts with their greatest hits isn’t going to change a lot.
Does that explain it? Yes. Is it satisfying? No, not really. Like I said last week, it means the albums chart isn’t super reflective of what’s popular right now, and it’s denying newer artists the chance to chart that these older bands had. Then again, maybe they don’t care any more—like I also said last week, who pays attention to the albums chart?
Who indeed?
- Actually, what did you come here to read about? It’s not like this blog has a theme. The tagline is literally “A blog about stuff”. ↩︎
- Excluding any above the limit of 16. ↩︎
- The “Defined Universe” is the collection of all places for which chart data is collected, apparently. ↩︎
- But only because the song was included on a re-release of Voulez Vous, ABBA’s sixth album in 2001. Until then, it had only been released as a single and on greatest hits albums—meaning, had that persisted, every stream would have counted towards ABBA Gold. ↩︎


Leave a Reply