uPlaya’s David Meredith on How Hit Song Algorithms Can Actually Help Artists

The words “hit song science” tend to set musicians on edge.

Rather than inspiring curiosity about a new piece of technology, the idea of a hit predictor, or some kind of perfect pop song analyzer, usually conjures up images of soulless homogeneity, of computers issuing ruthlessly effective and perfectly empty radio hits, of true creativity and artistry being wiped off the face of the earth as if it were a computer virus.

But according to David Meredith, the CEO of Music Intelligence Solutions, the company that runs uPlaya, that’s actually the exact opposite of what it’s for.

“With the explosion of creativity that’s happening these days, we need more sophisticated tools to help the content creator get the recognition they deserve,” Meredith explains, “and for the fan to filter through it.”

He’s not wrong. According to Tom Silverman, there were over 81,000 albums released last year that sold fewer than 100 copies. According to Meredith’s research, there are over 10 million songs available on the iTunes Music Store that nobody has ever bought. A decade into the 20th century, artists are done celebrating the fact that they can make a professional-sounding recording on their own. Now they’re wondering how to get those recordings noticed.

“We all need a way to curate this massive amount of stuff, and the tools aren’t there for it,” Meredith observes. “There’s a need for actual evaluation.”

The evaluation part is what tends to make musicians squirm the most: if a computer’s going to evaluate my art, how is it going to do it? But Meredith and his researchers have done a lot of homework. In building up uPlaya’s two services, hundreds of millions of songs were analyzed at a multitude of levels. “We can look at timbre, pitch, chord progression, melody, vocals, harmony, everything about the song [including the song's language],” Meredith explains.

uPlaya can analyze a song’s variables two different ways. The first, called hit potential technology, takes stock of past and current trends to determine the likelihood that your song, given a requisite amount of marketing money, will become a hit. It is, according to Meredith, supposed to be predictive rather than reactive, and the songs that score high enough are given awards that denote a strong likelihood for success.

An example of uPlaya's Music Universe.

The second, and more important type of analysis is called the Music Universe, or as Meredith likes to think of it, “Google Earth for music.” This type of analysis provides users with a visual display not only of what songs their music resembles, but of the likelihood that their song will appeal to fans of established hits.

This helps artists on a number of levels. It can help artists find kindred spirits or help them market their music to bloggers. It can help convince a label manager to try and introduce an artist’s song to a new audience.

It can also help smaller artists looking to score licensing deals. If a music supervisor wants a song like Bruce Springsteen’s “The Rising” for a soundtrack, but he can’t afford to pony up the $200,000 fee that song commands, he can jump into the Music Universe and find that song you wrote that pushes all the same emotional buttons. This theory is actually being put into practice by the Second City. The legendary comedy troupe is releasing its first branded feature film, Buzzkill, and they are letting Variety choose songs for its soundtrack using the Music Universe.

Meredith’s also encouraged by the fact that others see uPlaya as a critical tool for A&R. “There was a really interesting music blog who said that within five years, your Hit Song Science score is going to be more important than how many plays you’ve had on Myspace,” he says.

Of course, there will be dozens of things more relevant than Myspace plays in five years. But Meredith is confident that his company’s technology will be valuable to the entire industry, and for years to come. The artificial intelligence technology that powers uPlaya’s services is used for poring over mountains of data at incredibly high speeds. It’s been used in the financial services industry. It’s been used to optimize power grids and communication networks, “anywhere you need to sort through massive amounts of data and make informed decisions.”

The music industry might not change as quickly as financial markets do, but it is a very complex, dynamic business. The musical, financial, and cultural definitions of a hit song, like most cultural phenomena, are changing. But even as personal tastes, economic ecosystems, and models shift beneath everybody’s feet, one thing is never going to change. “Everybody on that continuum, they all need feedback,” Meredith says. “They all need attention on their songs. We see people using it on all levels.”

Tagged as: David Meredith, Interviews, Technology, uplaya