100 million dollars. That’s the value Music Xray expects to be adding to the recorded music industry in 2018, driven by its new Artist Investment program, currently in limited private beta. In the past few months, Music Xray’s confidence in its ability to predict successful songs has become strong enough, that they are beginning to invest in the careers of artists with high potential songs. It’s the culmination of years of data collection and industry crowdsourcing within its closed ecosystem, combined with artificial intelligence and music analysis software that picks likely winners and matches song candidates to opportunities.
Prior to submitting their songs to the professionals behind thousands of industry opportunities, artists put their songs through a process Music Xray calls Diagnostics in which five professionals and 20 potential fans evaluate them. The results get plugged into Amazon’s Machine Learning Platform and compared to hundreds of thousands of songs that have gone through the same process over the past several years. Music Xray knows which of those songs ended up receiving offers from the industry on the site, which ones were rejected multiple times, and how potential fans reacted. Once the calculations are made, Music Xray generates a “Selection Prediction Score, ranging from 0% likely to be offered a deal all the way to 99%.
Artists can then make informed decisions about whether or not they should continue to submit to opportunities. Submissions cost $16 on average so in essence, Music Xray is discouraging artists with low-potential songs from continuing to submit while keeping the pipeline to the industry largely unclogged.
Recently, Music Xray has begun offering to cover the cost of 20 submissions for artists with songs likely to succeed. In exchange, Music Xray gets a cut of the deals the artist lands.
“There is no other approach like this anywhere,” states Mike McCready, CEO of Music Xray and seasoned hit song identifier. “This program finds that rare sweet spot where emerging talent can get a boost. Taking the risk on fronting submission fees makes business sense for us because we’re basing the decision to invest on empirical data.”
To identify future winners, Music Xray has created a sophisticated filter that unites human and machine intelligence. On the human side, industry decision makers evaluate thousands of songs and acts each day. Any musician can have their music considered for almost any industry opportunity--from major label deals to placements in TV shows--with the guarantee of being heard and getting a response, something unheard of in the traditional music business. The Selection Prediction Score Music Xray generates on the first transaction assumes the song will be shown to the professionals behind at least 20 opportunities. Music Xray is over 90% confident in their predictions.
And that’s where things get interesting.
“We’ve reached the point where we can make strong, meaningful predictions,” notes McCready. “We have a critical mass of artists and industry professionals interacting to generate enough reliable data to add big computing power and make sense of it. Until we had the site’s 8 or 9 core features, which took years to build, and until that last piece, Selection Prediction, was plugged into Amazon Machine Learning Platform, we couldn’t do this. Once all the elements were in place, we had a sustainable model for growth and for supporting artists who had yet to break into the mainstream.”
The returns on these investments are generating a growing revenue stream for the company, while providing an invaluable service to some of the industry’s largest players of a continuous supply of needles from the haystack of millions of songs and acts.
“We’re putting our money where our mouth is,” says Jeff Durand, Music Xray’s CTO and principal tech architect. “We crunch the numbers and if we think a song or band is destined for a deal based on our own math, we might decide to make the investment.” Music Xray aims to invest in a diverse array of sounds and genres, to maximize success.
If tracks don’t fare well in the site’s analysis, Music Xray doesn’t lead artists on; the site purposefully sheds over 80% of the users who join each month. McCready says, “We’re a filter. And that requires that we filter a lot of the music out.” So far, the company has a great win rate: 13 out of the 13 early artists chosen for investment have been selected for deals.
Industry users of Music Xray include players like Atlantic Records, Roc Nation, Epic Records. Supervisors for TV shows like The Voice and Dancing With The Stars have gotten on board.
“It’s the clever application of very diverse technologies that make Music Xray work. It’s rewarding work because we’re solving complex problems and making the industry a better and more transparent place,” concludes Durand.