How Streaming Services Are Remaking The Pop Charts
How Streaming Services Are Remaking The Pop Charts
3:51pm Jun 05, 2015
For roughly half a century, the BillboardHot 100 — America's hit barometer — underwent constant change as it accommodated all the new ways Americans consumed popular music.
And yet, in a larger sense, for the first 50 years or so, it didn't change much at all.
From its founding in 1958 to the last half of the '00s, the Hot 100 measured, broadly, just two things: radio airplay and retail sales of the most popular songs in the country. Sure, every couple of years, Billboard had to change the chart's formula to account for new radio formats; the transition from AM to FM stations; the advent of 12-inch singles, cassingles and CD-singles; computerized sales-and-airplay tallying; and finally, the invention of the retail download store. But in all that time, radio was still radio, and the fundamental act of paying a couple of bucks to own a copy of a song stayed the same. In other words, for about 50 years, the Hot 100 was the gin-and-tonic of charts: two basic ingredients, one tasty cocktail.
In the last decade, however, the Hot 100 has become more like a martini, a three-ingredient cocktail. If radio is still the gin, and digital sales are the vermouth, the totally new ingredient — the olive — is streaming. Of course, a single stream of a song on Spotify or YouTube counts for the chart far less than does a sale of a single, or a play on a Top 40 radio station. But according to Billboard, depending on data volume in any given week, streaming in aggregate now makes up 20 to 30 percent of the Hot 100's data. That's a big olive.
This means every time you queue up anything from Maroon 5's "Sugar" to Death Cab for Cutie's "Black Sun" on a streaming service, you are voting for a hit. (Not every streaming service, however — more on that in a minute.) The prominence of streaming on the Billboard charts — comparable if not quite equal footing with radio and sales — is appropriate, given how important streaming is for the recording industry, and how (relatively) novel it is as a form of music consumption. Unlike radio, with streaming the consumer gets to choose what she hears. And unlike music retail, after she plays the song, she doesn't keep a copy.
Actually, what streaming most resembles is a very old form of music consumption, the jukebox — instant gratification, no retention. Futurists have been hyping the Internet as a "celestial jukebox" for years; arguably, in the age of Spotify and YouTube, this mission has been fulfilled. Indeed, when you play a music video on YouTube or Vevo, the site is now set up to autoplay another, related video immediately. The hits just keep on comin'.
Sixty years ago Billboard actually tracked jukebox plays — jukeboxes had their own standalone chart and were factored into the Top 100, a 1955 proto-version of the big chart. But by the time Billboard officially launched the Hot 100 in 1958, jukebox growth had stagnated and Billboard stopped tracking the category. Thanks to streaming, however, the playing of songs you don't keep is factored into the charts once again, bringing the Billboard charts back, full-circle, to where they were in Elvis Presley's heyday. (Maybe now that technology has improved, Billboard can re-add data from actual jukeboxes.)
Billboard dipped its toe in the streaming waters in 2007, adding nascent streaming services like AOL Music and Yahoo! Music to the Hot 100 formula years before Spotify arrived in America. At the time, it felt like a beta test for the future of the chart — these early services had modest numbers of users and, in aggregate, accounted for only about 5% of the Hot 100's points.
Streaming only began having a major impact on the chart once Spotify finally reached America in 2011 (three years after its European debut). In March 2012, Billboardlaunched its On-Demand Songs chart, a new ranking of song plays at services ranging from the well-established Rhapsody to the now-defunct MOG — but Spotify made up a major share of the chart, which was also factored into the Hot 100.
In early 2013, less than a year after On-Demand Songs launched, Billboard's tracking of streaming music got a massive boost from the launch of a more comprehensive Streaming Songs chart. Just over a month later, tracking of music video plays on YouTube and Vevo were added to the formula for the first time. The new methodology launched with a bang — or perhaps a hip-thrust — instantly generating a No. 1 hit, Baauer's "Harlem Shake," that might never have made even the Top 10, let alone No. 1, without video play. At YouTube, "Shake" was a juggernaut, generating more than 100 million views in its first week between the official video and the thousands of fan videos that employed the song. Whereas at on-demand audio services like Spotify, "Shake" was more of a lagging indicator, never getting past No. 15 on Billboard's On-Demand Songs.
The On-Demand Songs chart has now been in existence for more than three years, Streaming Songs for more than two. Each feeds into the Hot 100 like Russian nesting dolls. On-Demand Songs still does not include video play, focusing only on music subscription services like Spotify, Rhapsody, Tidal, Rdio, Beats, Google Play, Amazon Prime and others. (These same services also feed into the magazine's recently revamped Billboard 200 album chart, which since last December has combined sales of albums with aggregate streams of tracks.) The more data-heavy Streaming Songs chart takes the On-Demand chart's data and layers in YouTube and Vevo video play (including unofficial user videos containing at least 30 seconds of an official recording; if your adorable home movie of your kids includes "Uptown Funk!" in the background, that counts). It is this Streaming Songs chart that feeds into the biggest one, the Hot 100 — serving as the third leg of that chart alongside Radio Songs and Digital Songs, Billboard's main airplay and sales charts.
As thorough as Billboard's streaming data might appear, it is not all-encompassing. One major celestial jukebox that's not a factor in any of these charts is Pandora, firmly established as America's top music streaming site with, by some measures, more than half of all US music streaming users. There are two good reasons for this Billboard non-inclusion — one logistical, the other definitional.
The logistical reason is that Pandora does not share its data with Nielsen SoundScan or Billboard. This is perhaps unsurprising, given Pandora's currently combative relationship with the music industry over royalties. But the definitional reason why Pandora is an odd fit for the charts is it's not actually an on-demand streaming service. Pandora is essentially a programmable radio station — users input a seed (artist, song, genre, etc.) and Pandora delivers a stream of music like that musical element, but not any specifically requested song. (Indeed, Pandora's current royalty model is premised upon you not getting to pick the particular song you want.)
Were Pandora ever to begin reporting to Nielsen, one imagines the sheer variety of its musical selections might make it the ultimate long tail, pumping up the plays for more obscure acts by amounts that would barely register for the charts. On the other hand, a service as large as Pandora would certainly have some measurable impact on its most-played tracks. Perhaps, someday, Billboard will expand the Hot 100 cocktail with another ingredient. Still, even with the data we do have, we can begin to define what a "hit streaming song" looks like. If Billboard based the Hot 100 entirely on streaming music — rather than the combination of radio, sales and streaming the chart actually comprises — the last couple of years on America's flagship pop chart would have looked quite different.
For example, in a 2013 parallel universe where radio and sales didn't exist, Robin Thicke never would have had a No. 1 pop hit. (And then whom would the estate of Marvin Gaye sued?) Instead of Thicke's inescapable radio-and-iTunes smash "Blurred Lines," the 2013 Song of Summer would have been Miley Cyrus's trollgazy hit "We Can't Stop." Between 2012 and 2014, a series of international viral videos would have gone to No. 1: Korean YouTube king Psy would have had two U.S. No. 1 hits, "Gangnam Style" and "Gentleman"; Norwegian comedians-turned-popstars Ylvis would have had a U.S. chart-topper, as would French singer-songwriter Soko. Lady Gaga's semi-flop album Artpop would have been partially redeemed by a No. 1 hit. Beyoncé's "Drunk in Love" would have hit the top — it would have been her first No. 1 in five years — as would Nicki Minaj's "Anaconda," which would have been her first No. 1 ever. Perhaps most appealing, in the summer of 2014, Canadian reggae band MAGIC! never would have reached No. 1 here with the trifling "Rude."
Notwithstanding my feelings on "Rude," I wouldn't actually have wished for this version of Billboard hit enshrinement. The above-mentioned streaming hits got huge thanks largely to YouTube, which tends to reward visuals over sonics — whether it's callipygian displays or fleshy oversharing. By contrast, the On-Demand Songs chart — which is actually a more limited chart, insofar as it does not include video views — more closely tracks the most-purchased songs at iTunes and most-played songs at radio. Of the songs that went to No. 1 on the Hot 100 over the last three years, 76 percent also reached the top on the On-Demand chart.
Even though it has less data baked in, On-Demand Songs is actually a fun chart to follow; it's only based on streaming audio sites, and so it's like a slightly livelier version of the Hot 100, whose No. 1 spot has turned over very infrequently lately. In a different parallel universe where the Hot 100 was based only on on-demand music services — no YouTube, no radio, no iTunes — Daft Punk's "Get Lucky," the mostacclaimed song of 2013, would have hit No. 1 (on the actual Hot 100, it peaked at No. 2 in a now-infamous Summer Song battle royal with "Blurred Lines"). Indeed, several near-miss, coulda-been Hot 100 No. 1s topped the On-Demand Songs chart, including Maroon 5's "Payphone," fun.'s "Some Nights," Ellie Goulding's "Lights," Justin Timberlake's "Suit & Tie," Drake's "Hold On, We're Going Home," Jessie J, Ariana Grande and Nicki Minaj's "Bang Bang," Hozier's "Take Me to Church" and Sam Smith's "Stay with Me" (like "Get Lucky," a future Record of the Year Grammy winner). The On-Demand chart looks like an awfully good proxy for the Hot 100, with somewhat better variety — appropriate, given how much more genre-blind Spotify users have shown themselves to be.
At this writing, there's actually one more near-miss song that's reached No. 1 on the On-Demand chart but only No. 2 on the Hot 100: "Trap Queen" by Fetty Wap. This small distinction in chart position rather understates the importance of streaming to the success of this hardcore-yet-romantic song — arguably the most street-oriented, least pop hip-hop track to chart this high on the Hot 100 in years.
"Trap Queen," on the other hand, reached its peak positions through shareability and sheer pluck. In an article noting the remarkableness of this all-rap Top Two, Billboard reporter Gail Mitchell revealed that more than half of the chart points for "Trap Queen" on a recent Hot 100 were from streaming — more than double its chart share from digital sales, and triple its radio airplay. This makes intuitive sense; ever since Spotify arrived in America, I have expected streaming to provide hip-hop with a chart boost. Trends in hip-hop shift quickly — late last year, another Top 10 hit by a new rapper, Bobby Shmurda, actually got its chart start via a viral dance shared on Vine — and streaming offers rap fans the chance to sample more new sounds, cheaply or for free.
Ultimately, "Trap Queen" couldn't get past No. 2 on the Hot 100, as a video-fueled new single by Taylor Swift leapt over it into the No. 1 spot last week. But even though "Trap Queen" now goes into the history books as a runner-up hit, we may well remember Fetty Wap (and Bobby Shmurda) as the bridge to the next phase of the Billboard charts — the moment when streaming music came of age as a pop force.