Streaming music’s quest for the magical, mythical mixtape

cassette tape 300wSpotify’s newest product, the Discover Weekly playlist, is streaming’s latest attempt at perfect music programming. What is perfect programming? Online music services vie to deliver the right song to the right person at the right time, and, just as important, the right flow of music to make the stream sticky. Just as in radio, music services want to prevent listeners from tuning away. Allowing song and playlist skipping helps build Time Spent Listening, but it is increasingly easy to skip off to another platform entirely, like going up the radio dial.

The mixtape metaphor is invoked sometimes, to bring the crackly romance of 70s and 80s technology into the antiseptic gleam of data-driven music. What’s really appealing about the ideal mixtape is a combination of knowledge and effort put into the best of them. Great mixtapes are kept for years, stretched and shredded with use.

Subscribers to music services can make their own playlists. Hooking into the shared playlists of other members can be a mixtape experience also. (With dimensions impossible to achieve in the analog world: One of my favorites in Spotify goes on for 302 hours.)

But in a crowded industry, where music services offer a paralyzing choice of millions of tracks, they compete on the basis of knowing YOU better than the others, a delivering perfect experiences.

To some extent, the competitive fulcrum hinges on data vs. humans. Although streaming music is inherently a data enterprise, sentiment has swung over the human-powered curation. Apple’s launch of Beats 1 brings old-style radio to new-media reach, and celebrity hosting by Zane Lowe, the ex-BBC radio jock who was the ideal mixtape creator for millions of people. It is telling that Beats 1 is the lead product for Apple Music. But aside from that, the jukebox platform of Apple Music inherits much of the human curation which informed its predecessor, Beats Music.

Merging the old values with new conveniences is appealing, and the point of Dash Radio, which is expanding its platform of radio-style hosting and programming with over 60 pureplay stations.

Streaming the right music to the right person at the right moment hooks into the user’s mood and activity. Songza got that ball rolling with its Concierge service, matching its pre-built playlists to the specifics of each user’s day. many other platforms (e.g. Rhapsody, iHeartRadio) have run with that concept.

The 8tracks service is perhaps most purely aligned with the original mixtape experience. It is a crowdsourced collection of personal streams shared across the platform — the playlists are actually called mixtapes, and the creators are called DJs. Users can upload their owned recordings or explore the built-in 8tracks library which is heavy on indie labels. The listening experience is personal and eclectic.

On the data side, music intelligence company The Echo Nest was a giant presence in the industry, providing playlisting brains for dozens of music services, until it was acquired by Spotify in March of last year. (The Echo Nest continued serving its external clients post-acquisition, but some of those services ended their contracts.) The merger cemented Spotify’s commitment to data and algorithmic curation.

That commitment has been core to Pandora’s programming also. Founded as a technology project to develop the Music Genome, Pandora analyzes music uniquely, identifying elements of music that might cross genre boundaries. Nearly 80-million people listen to Pandora each month, trading catalog size (Pandora has fewer than five-million tracks) for uniquely precise programming.

Spotify’s new Discover Weekly playlist is rolling out to members this week, and to our ears is a decided success. We are as impressed as we were with Play It Forward, Spotify’s personalized end-of-2014 playlist gift to members. In each case, the playlists are unique to each user, which can be accomplished only with data-crunching.

If the perfect online listening experience depends on some blend of algorithms and people, it is interesting to observe that the two market leaders (Pandora for Internet radio; Spotify for music jukeboxing) are pushing their data advantages.

Brad Hill