Pandora launch personalized playlists to beat competitors’ “generic buckets of music”

Pandora are looking to blow away the competition with playlists that aim to be more catered to individual users' tastes than any others out there, TechCrunch reports. "When we think about what the competition’s doing, they’re really putting more generic buckets of music together," says Pandora’s Chief Product Officer Chris Phillips. "And they have a generic name for the playlist, whereas I can go right in [to Pandora’s personalized playlists] and say, “oh this is mood I’m in,’ and it’s spot on."

Phillips doesn't say exactly what competition he's talking about, but as TechCrunch also points out, surely he's thinking of Spotify's Discover Weekly, Release Radar and Daily Mix. "What we’re doing is what we believe is the bleeding edge of deep learning algorithms," he adds.

These playlists will be built by using Pandora's Music Genome, which as TechCrunch writes, "powers Pandora’s other personalization efforts, like its Thumbprint Radio, or the feature where it can automatically play similar music when you reach the end of a playlist you created," along with "more traditional collaborative filtering methods, and in-house editorial curation." "We’ve been building out, for many years, a collection of well over 75 machine learning algorithms and techniques to help drive content discovery and delivery," says Phillips.

Pandora is able to create over 60 personalized playlists, and they're rolling them out "in response to users’ listening activity. For example, if you’re playing a lot of upbeat pop music, you may see a playlist appear called 'Your Party Soundtrack.'" The playlists have already begun rolling out to Premium subscribers.

Pandora's Q4 2017 report counted 74.7 million active users, 5.48 million of which are paid subscribers, compared to Spotify who claim 157 million active users and 71 million paid subscribers. Apple Music has 38 million subscribers, and Amazon's Prime Music and Music Unlimited have an estimated 16 million.

Have you tried out the new playlist yet? Can the "bleeding edge of deep learning algorithms" in fact figure out your personal taste?