The biggest companies in the world struggle each day to surface content you want to see from the vast reservoirs of existing information. Google has invested billions of dollars in its search algorithm, designed to bring you the best 10 results for anything you care to search. Facebook has invested a similar amount in their News Feed algorithm to surface the posts from your friends and connections that you’re most likely to engage with.
Any industry with millions of pieces of content and a single person at the other end of the search box will struggle with the problem of surfacing the right content to the right person at the right time. Music is no different: each streaming service (with the exception of Prime Music) has 30 million or more songs.
No one could possibly listen to this much music in their lifetime, and the number of available songs only continues to grow. How can streaming services possibly help you find music that you want to hear?
Apple and Spotify have tackled this problem from opposite ends. Spotify placed its bet on algorithmic curation when it acquired the Echo Nest at a valuation of $100m. But Apple is betting on human curation.
In fact, judging by the amount of times Jimmy Iovine has declared the superiority of human curation, and their investment of $3 billion into Beats — a service dedicated to human curation as its value proposition — and that Apple Music employs hundreds of music editors to build music playlists, Apple is betting big.
Apple focuses so hard on curation because music discovery is uniquely suited to a human touch. There aren’t so many new songs published each day that a team of human editors will find it impossible to listen to a meaningful portion of them. On the other end of the stream, music listeners care about taste, have a finite number of songs they’ll take a chance on, and good music is so widely appreciated that personalization plays only a minimal role. Music fans are willing to listen to someone with taste who says “You’re gonna like this.”
Music discovery can be served with an algorithm, but algorithms have limitations. One reason Apple focuses so hard on human curation is a limitation called the filter bubble. When algorithms use information about what you already like to surface similar information or songs, you get homogenized results. That’s why a 90’s Pandora station will nearly never play a surprising song.
When Facebook tested their News Feed algorithm for the filter bubble effect, they found that people, due to choice of friends and choice of what to click on, see more posts from their own political affiliation and a reduction in posts from the opposite political affiliation.
In music, the filter bubble means you lose the chance for serendipity, for a wonderful song to take you by surprise. Human curation surfaces and suggests music that an algorithm won’t.
Apple Music is clearly better, right? Not so fast.
The problem is that Spotify isn’t totally algorithmic. They use human curation, too. They use an Echo Nest invention called Truffle Pig to help surface tracks that their human editors can then curate into playlists.
Earlier this year they rolled out new features like matching songs to your running pace, and heavier human curation. As early as 2013, Spotify offered human-curated playlists and has only continued to expand it. Even Spotify’s primary method of algorithmic curation is decidedly human — Spotify uses collaborative filtering, which finds other users who listen to music like you, then suggest music they listen to but you don’t.
Finally, one of the best parts of Spotify is the social network built around friends, experts, and artists. Spotify already has some of the best user playlists out there, and it also features playlists curated by music experts like Filter Magazine and Pitchfork.
Jimmy Iovine is openly contemptuous of algorithms…
“The only song that matters as much as the song you’re listening to right now is the one that follows this. Picture this: you’re in a special moment…and the next song comes on…BZZZZZ Buzzkill! It probably happened because it was programmed by an algorithm alone. Algorithms alone can’t do that emotional task. You need a human touch. And that’s why at Apple Music we’re going to give you the right song [and] the right playlist at the right moment all on demand
. . . which is why it’s even more surprising is how heavily reliant Apple Music is on algorithmic curation. The For You tab appears almost entirely surfaced by algorithms, and most of the radio stations are reminiscent of Spotify’s identical radio feature. In fact, outside of some human-curated playlists, most of the Apple Music service is driven by algorithms.”
Apple has focused on human curation as a primary selling point, but unfortunately for Apple, Spotify is far ahead. In the future, look for both services to focus on a blend of human and algorithmic curation.
For now, curation is a big selling point for Apple, but it doesn’t mean much for the artist.