Billboard Revises Country Chart Methodology

[koganbot]

Billboard has a squib today, describing its new way of computing the Country Songs chart. The chart used to be airplay alone, as compiled by Nielson; now it includes streams and downloads (though I presume there's no way to tell where and whom the downloads are coming from — that is, whether they're coming from the country audience or not). Also: "With digital download sales and streaming data measuring popularity on the most inclusive scale possible, it makes perfectly logical sense that the radio portion of the new chart calculations include airplay from the entire spectrum of monitored formats." Don't know if I'm interpreting that sentence correctly. Does it mean Billboard is now counting the airplay a country song receives on noncountry stations as well as on country stations? (In this case, is it always clear what a country song is? What about an alt-country track that gets a lot of play on Triple A but almost no play on mainstream country?*)

Guess who has a song that's number 42 on Mediabase's current country airplay chart, that was number 21 on last week's Billboard Country Songs chart, and that has just jumped to number 1 on the new Country Songs chart as a result of the change in methodology!

(I wish Billboard would tell us the actual numbers for a song: e.g., this is the number of spins (or the audience size), this is the number or streams, this is the number of downloads, this is how we weight them. I know that the difference between number 1 and number 10 is usually greater than the difference between 11 and 20, which is about the difference between 21 and 40, etc. But it would be useful to see data. Paradoxically, if your business is based on the claim that you're the best source of information, you have to limit access to that information.)

Btw, here's a track that I'm embedding 'cause I happen to like it, not 'cause it illustrates anything. It's currently 33 on the Mediabase country airplay chart, still rising but probably not much more, not in the Country Songs Top 25, and I have no idea how its streams and downloads add up, or compare (it's had 426,000 streams on YouTube for three weeks, which is solid but not amazing; for example, Kristen Kelly's "Ex-Old Man," rising equally slowly at number 30 on Mediabase, has 539,000 YouTube streams over six weeks):

[UPDATE: This particular embed is no longer up on YouTube, and I can't remember what it was. Might well be the Kix Brooks song I listed as CURRENT MUSIC down below, but I don't know.]

*EARLIER UPDATE: Okay, Triple A is so small that it's not going to make much of a difference: Mumford & Sons "I Will Wait" is number 1 on Triple A with an audience size of 2.584; if it had the same-size audience on the country chart, and you added the two together, that'd only raise its rank from something like 42 to 38 on a combined chart. But if you were to decide to call "I Will Wait" a country song, and were to add together its Triple A audience, its Hot AC audience (3.168), and its Alternative audience (12.180), you'd get a hefty 17.932. Since most country songs aren't scoring big on noncountry formats, "I Will Wait" would be at least in the mid 20s on an airplay chart for country songs that combined all radio formats. Now, I assume "I Will Wait" is not getting any country airplay and isn't getting classified as country. But let's say it got a bit more than a smattering, enough to give it a relatively small audience on the country stations, say about what it's got on Triple A: 2.584. Well then, presumably you'd have to put it on your country chart, and reasonably high. For an interesting parallel, the audience on country radio for "We Are Never Ever Getting Back Together" (in its eighth week) isn't much higher: 2.803, and look where it's charting. (Of course, you can choose to give the noncountry charts less weight, which is probably what Billboard is doing, if my interpretation that it's including them at all is correct.)

More general Billboard article about chart changes here.

Comments

[dubdobdee.livejournal.com]

No wisdom to add, just to say that I enjoyed the precision of the phrase "no way to tell where and whom the downloads are coming from" -- why not "are going to"? because what we're talking about is the information that may or may not be going the other way

(2) becomes (1)?

[koganbot]

The precision (assuming there is any) is that a "download" involves two entities*, (1) the entity making the track available for download, and (2) the entity doing the downloading. Of course, a (2) can become a (1), and often does. In any event, both those entities are necessary for there to be a download, and therefore the download is coming from both. Only malware, adware, and spyware are strictly one way. With downloads, a lot is unclear about both. Ditto streams. But I was more concerned with (2) than (1), and was writing more quickly than precisely.

My concern is this: I'd think that the point of the genre charts ("country," "hip-hop," "hip-hop and r&b," "adult contemporary," and so forth) is to inform us what's going on in a particular market or submarket. Airplay is obviously an uncertain measure, since some tracks sell well without a lot of airplay. Yet sales are often a worse measure, since most people who tune in (this is especially true of AC) don't buy the song, but that doesn't mean they don't listen for particular songs and don't get put into reach of advertisers by those songs. Also, there are still people who buy albums because particular songs are on them, and singles sales don't measure this.

But with physical sales you sometimes can have at least some idea where the transaction was located (the album or single was lifted from the country bin, let's say, or was purchased at a store that mostly serves the country audience). And a song played on a country station is presumably getting heard by an audience whose members often listen to country stations, hence a country audience. (Not that someone can't belong to more than one audience.)

Whereas with downloads and streams, except in the rare instance where the dl or stream is posted on a site that has a specialty audience (say CMT or Pitchfork), the particular download or stream doesn't carry this information. We want to take downloads and streams into account (owing to the inadequacy of airplay as a measure). But if we're trying to measure what's going on in a particular market (in the way that "country" is a market), we risk losing that information if we overweight downloads and streams and if we weight all airplay equally. It is useful to know that a track (say, by Blake Shelton) that's aimed for the country market also sells beyond that market. Something selling beyond the market tells you something about the market, and that should be part of the tally. But if it's overweighted in the tally, then you're losing information about the market. This is because you're starting to measure how well stuff that just happens to be in the particular market is doing overall, which then will blur your picture of the market. "Gangnam Style" jumped to number one in hip-hop by the new measure, and I'm skeptical that that's a true measure of how it's doing as hip-hop.

I don't know how Billboard figures this stuff out. Do they just guess?

*Or three, if we count the thing being downloaded.

[brak55.livejournal.com]

It wasn't just Country. They updated most of the Song charts including R&B and Rock. The Rock Songs is really different as all 12 of the songs on Mumford and Sons' album Babel are in the top 25.

[koganbot]

Dave wrote:

But "popularity in digital outlets" isn't necessarily democratic — my guess is that the audience for downloads and legal streams is mostly white and/or relatively affluent.

Working with kids illustrates this — kids in suburban/affluent public schools that I work with have all kinds of paid, automated digital access to music material. Kids get iPod Touches in second grade and iPhones in fifth, and they often have "allowances" at their discretion by the time they're graduating elementary school. By contrast, kids in public schools in the city seem to show a lot more variety in how they find, download, purchase, etc. their music — they listen to the radio more, they certainly don't have "digital allowances," and though they're fairly likely to have mobile devices by late elementary school, it's not as widespread as the suburban schools.

And maybe more importantly (if more anecdotally — so take with a grain of salt, as it's just based on informal firsthand observation) there's not really a culture of entitlement around those devices and the centrality of the devices in media consumption as a status symbol. That is, the cool kids all know the songs and the dances, but there's no particular social capital in being the person who owns the album, or has all the songs in their phone, etc. Would love to actually test this to see if there's a difference there.