Misix Library

The Data-fication of the Music Industry

August 19, 2015


  • With Nielsen SoundScan, the music industry can judge the popularity of artists based on irrefutable data.
  • Like SoundScan, Shazam also offers industry insiders a data-driven approach to tracking the rise of upcoming artists.
  • These services have dramatically affected the landscape of the Billboard 200.


Years ago, way back in the early 1990s, Billboard magazine still constructed its famed charts of the most popular songs and albums using a strategy that now seems almost unfathomable: It asked people.

Here’s the problem with people. They lie. They make mistakes. Sometimes, they’re just plain wrong. For decades, the music industry tolerated those flaws because there wasn’t anything better. Then, as is usually the case, technology came along to save the day — two days, actually. One of them changed how Billboard measured artists’ popularity, and the other changed how those artists get discovered in the first place. The impact they made on what you put in your ears is significant.


Day one: March 1, 1991

Let’s get back to the Billboard charts. Under the archaic system in place prior to 1991, someone from a radio station or record store manually filled out a report of that week’s most played or best-selling music. For decades, Billboard relied on that system, which was as flawed as you can imagine:

  • There was the time-honored practice of payola — record companies bribing radio stations and/or individual DJs to play a specific song.
  • Financial self-interest tainted results from store owners, who had no incentive to promote an album they didn’t have in stock.
  • Both sources simply stopped reporting specific songs or albums when the label stopped promoting them.

The trouble with relying on people to deliver data, particularly ones with skin in the game, is that when they can use the system to make money, they probably will. Record labels did because getting songs in and out of the charts quickly meant they could sell more. Record-store owners did because they needed to move product. DJs did because someone with more money than them wanted to give them some.

The solution to this problem was Nielsen SoundScan, which collected its first batch of data March 1, 1991. By getting numbers directly from cash registers — a notoriously unemotional and dispassionate source — SoundScan made it possible for Billboard to take a step toward true accuracy. The impact was immediate, particularly in the country segment. Even more particularly in the Garth Brooks segment.

“The very first week SoundScan came online, his then-current album, ‘No Fences,’ shot into the top 10,” Slate chart columnist Chris Molanphy told National Public Radio’s On the Media. “It’s widely perceived that the advent of truly accurate counting allowed the industry to perceive just how popular he was for the first time and promote him accordingly.”


Just think: no SoundScan, no Chris Gaines.


Billboard isn’t foolish enough to think its system doesn’t need occasional improvements, either. In November 2014, it added streaming services to its formula for the Billboard 200, responding to a rising medium that according to the 2014 Nielsen Music U.S. Report delivered more than 164 billion songs to users in 2014.

Of course, this update had its own Garth Brooks in the guise of Disney Channel graduate Ariana Grande, whose “My Everything” went from No. 36 under the old system to No. 9 using the reformulated model in which 1,500 streams (or 10 downloads) equal one album. Why those figures? Well, the download criterion is pretty obvious, as most albums contain in the neighborhood of 10 songs. As far as the 1,500 streams, we have a reasonable theory:

  • Spotify says its per stream payout is “between $0.006 and $0.0084.”
  • The midpoint of that range is $0.0072.
  • $0.0072 x 1,500 = $10.80.
  • The average price for an album in iTunes’ Top 10 as of Jan. 6 was $10.39.


Day two: July 10, 2008

Apple launching its App Store makes this date notable enough, but we’re more concerned with one of its initial residents: Shazam.

The music-identification service had been available since 2002, when its users needed to do something called “dialing” to access it. After holding their phone aloft to capture whatever caught their ear, they received a text message telling them the song’s title and artist. With its rebirth as an app six years later — and subsequent updates that added GPS and background auto-tagging capabilities — Shazam altered how new acts get discovered and future hits get market tested.

On the discovery front, there was a time when agents relied on word of mouth to ferret out the next big thing, traveled to wherever these promising youngsters played and used good old fashioned instincts to determine whether they were as good as advertised. Shazam has made this approach as quaint as sending a telegram.

Word of mouth has been replaced by data that reveal songs gaining popularity. Interactive maps make it possible to spot upcoming local acts without crisscrossing the country. And who needs instincts when you have the crowdsourcing efforts of more than 100 million active mobile users telling you what people want to hear? Warner Music Group decided that approach was good enough to collaborate with Shazam for a label imprint on which it could sign artists with the data to back up their talent.

There’s still the small matter of turning those artists into worldwide phenomena. Might as well use data for that, too, right? In an article last December, The Atlantic explained how the path to superstardom starts in cities like Victoria, TX, where Republic Records essentially focus-tested a single from one of its signees. By checking the song’s Shazam tagging numbers in the days following initial radio play, Republic could find out whether it made an impact on listeners.

There’s no statistical evidence that shows Shazam’s completion percentage is any better than the scouts still pounding pavement. But the service has offered acts to watch the last couple years, a few of which hit big:

  • 2012 — Frank Ocean, A$AP Rocky and Lana Del Rey.
  • 2013 — Haim and French Montana.
  • 2014 — Sam Smith and Vance Joy.

Does that mean 2015 hopefuls like Echosmith and Sam Hunt should make room on the wall for their platinum records? Well, probably not. For starters, Shazam’s .233 average (seven out of 30 acts to watch) is more Dave Ross than Dave Winfield. Then there’s the fact that streaming’s recent rise helped knock another 9% off album sales in 2014 (257 million units vs. 289.4 million in 2013), so the music industry isn’t exactly awash in million-selling records these days.

But hey, thanks to technology, at least it knows those numbers are accurate.