While record label talent scouting may have seemed like one bastion of the industry that was safe from artificial intelligence, Warner Music Group clearly feels differently, having recently acquired Sodatone, a startup which uses technology to identify unsigned talent.
Guest post by Bobby Owsinski of Music 3.0
Automation is taking over so many aspects of the workplace these days, but the last place you’d ever think it would happen would be in record label talent scouting. Having a feel for new music and artists that may resonate with large segments of the population is a talent in itself, and one that’s been rewarded with industry accolades and large salaries. That’s why it seems a little far-fetched to think that the traditional label A&R person could be replaced by a machine learning algorithm.
But is the thought really that implausible? Warner Music Group doesn’t think so.
The company recently announced that it acquired Sodatone, a Toronto-based startup that claims to have “established a reputation as the world’s premier A&R insight tool.”
Sodatone, which launched in 2016, claims it has developed technology that combines streaming, social and touring data with machine learning to help identify the new unsigned talent that it thinks may have some future success. Among the data that it measures is the loyalty and engagement level of an artist’s online fans, as well as the potential ‘virality’ of a track to help make its determination.
The company claims that its algorithm has been used with great success by a number of top label and publishing executives, although no specific names are offered.
Every company in the music business is looking for an edge, so in the end it’s no surprise that at least the talent scouting portion of A&R is becoming automated. If a company has a single taste in music, it’s easier to have one algorithm search for new talent rather than have multiple humans (all with different tastes) doing the same thing. Good talent, especially in this area, has always been hard to find.
Still, you have to wonder just how effective an automated A&R bot will be in spotting the next new underground trend that will change the face of music. After all, it happens regularly every decade or so. On the other hand, maybe it will be successful in finding one sooner (and the sooner the better!). We can only hope.