The data that can predict a pop hit

This article was taken from the June 2015 issue of WIRED magazine. Be the first to read WIRED's articles in print before they're posted online, and get your hands on loads of additional content by subscribing online.

These days, musicians go from uploading a YouTube cover version to signing a record deal in a matter of months, making it tricky for talent scouts to keep up. Alex White, CEO and cofounder of New York-based music-analytics company Next Big Sound, however, already knows who'll be top ten up to a year in advance, by sifting the potential hits in a complex ranking process.

The company uses data from social media sites -- such as number of Vine loops played, or Spotify streams -- to come up with a score for individual artists from 1-100. This measures the likelihood of that artist hitting the Billboard 200 (the US albums chart) within the next 12 months. "Iggy Azalea crossed 80 per cent likelihood in July 2013, and in April 2014 she debuted her album at number three on the Billboard 200," says White. Social sources such as SoundCloud, Instagram and YouTube combine with traditional ones, including radio plays and TV appearances, to produce a ranking that updates daily, based on real-time data. Next Big Sound is just one in a cluster of monitoring tools. Others include royalty tracker Kobalt (WIRED 05.15), which allows artists to monitor when and where their music is used.

And as social networks rise and fall almost as fast as the stars using them, Next Big Sound's ranking algorithms constantly change, too. "We run blind training sets to re-assess the weighting of each service," White explains, "Five years ago, MySpace would have a totally different weighting than today, so it tracks where consumers are finding their music."

This article was originally published by WIRED UK