Spotify has embarked on many efforts to provide music discovery to its audience, and the latest move in that direction is a playlist collection called Fresh Finds.
Unlike the immensely popular Discover Weekly feature, Fresh Finds are universal playlists for all listeners. The set is divided by genre, with separate lists focused on hip-hop, electronic, vocal pop, guitars, and experimental music. The main playlist contains a sampling from each of those genre collections. All the featured performers are up-and-comers who haven’t been widely discovered yet. The playlists are currently featured in Spotify’s Browse section, and their programming will update with new artists every Wednesday.
It’s that programming that’s most interesting about these playlists. The choices will stem from both sides of the playlist creation coin: human curation and data-driven algorithms. Spotify’s programmers will be following blogs and music sites to get the scoop on the latest new talent, as well as using data from listeners that have been identified as tastemakers who can find the next big act.
“By analyzing the listening behavior of our top tastemaker users, we’re able to predict new breakout artists and filter their hits-to-be into playlists with the most promising new music out there,” said Dr. Brian Whitman, Spotify’s principal scientist.
Many music critics argue that a machine can’t deliver the same success when it comes to recommendations and discovery; it’s even a major marketing premise behind Apple Music. But Spotify’s experiments in algorithmic playlists and in combining them with the human touch have generally been successful. Discover Weekly skyrocketed to 1 billion streams in less than three months; not too shabby for a bunch of computers.