In today’s digital era, consumers have a plethora of choices for where to consume music. While algorithm-based services (Pandora, Songza, etc.) are becoming increasingly popular, broadcast radio stations (which still have very human elements) are in a unique position to serve as experts at cultivating good music choices for their listeners. The problem they face however, is identifying which songs will be most popular with their user base. By improving the selection they serve their audience, radio stations can grow their listener base and the ad revenue that comes with it.