EMI Plays “Moneyball” In The Music BusinessMusic has always been about the sound, the moment, and the numbers. Chance, luck, and talent have always mattered. They still do. But today, with new data and insight, “catching a wave” means not just chancing upon a great ride, but launching tsunami after tsunami.

David Boyle, the senior vice president for customer insight for EMI Music, the world’s fourth- largest record company, has done for music what Billy Beane did for baseball: transform it to a data-driven business.

Except instead of buying runs, Boyle’s goal was to buy hits – from its established talent, and from future stars hidden among thousands of aspiring rock ‘n rollers.

Like Beane, Boyle built winners at EMI converting data to discovery to investment. For established stars, data helped connect the music to the fans, and get people to buy more. Scouring for little known artists, EMI used data to help predict if listeners would love a sound and a song.

EMI’s source for all these predictions? A dataset called the Million Interview Series, built up over a period of several years in the 25 countries where EMI does business. The series contains 100 pieces of data for each of the million interview subjects, everything from where she shops to the new music she prefers.

Every day, EMI generates thousands of data points for each artist and how particular customer segments are responding. It constantly updates the series through the song service Spotify, through which EMI markets its tunes – and in turn collects not just fees, but streams of even more customer data. Tons and tons of customer data based on upwards of 50 different indicators.

EMI’s big data-enterprise started small, with an Excel file on Boyle’s laptop, projecting sales for ten artists, some major. Boyle honed and refined the models and the reports. A few major deals with household-name artists soon proved the small model worked, and Boyle began to scale, ramping up to handle a hundred artists, then thousands. Soon “push” turned to “pull,” as artists and managers demanded the reports at each stop on tours.

A new universal language of data emerged, transforming EMI. “It’s gone from a business that made all those decisions on gut instinct or skills or judgment,” Boyle said, “to a business where data or insight are there in most of the decisions that we make – which artist to work with, which tracks to release as the first single for an artist, how much to invest in marketing, how to invest marketing money.”

The result – EMI has backed relatively unknown artists whose data signals showed strong market potential, all the way to number one hits.

Do they do this better than on hunch alone, like in the “old” days of a few years ago?

“We’ve had major artists decide to re-sign with EMI because of the insight into the kind of projects they should pursue,” Boyle said.

In my next post, I’ll explore how Gilt Groupe is taking personalization to completely new levels with the help of big data.