After a technology meltdown in 2015, losing all my beloved iTunes, many of which I’d ripped myself after uploading CD after CD in 2004, I gave into the whole music streaming app craze, which seemed ubiquitous in my social circle. For what felt like years I’d been regaled about this Pandora entity—but it seemed impure, unfair, dare I say creepy—suggesting what you might like to hear based on one random song, and forcing advertisements upon you in order to keep listening.
After seeing the convenience of Spotify as used by a friend, who paid for premium, I opted for the latter. The interface was more approachable, and for a flat, affordable monthly fee, I had access to everything, sans commercials.
While I really didn’t take the time to “follow” all my favorite artists and curate playlists, based on what I was listening to or “saving”, I began to notice something: the radio feature, based on one song of your choice, ala Pandora, was pretty on point. That gave me the incentive to then try the “Discover Weekly” feature, again, something I’d normally overlook, which now was hitting me every Monday with a hearty, eclectic list of interesting, and mostly wonderful tracks. As a DJ and a complete music nerd I found myself impressed with these selections—many of which were old, obscure, world-music, you name it. I consulted a fellow music nerd/friend/confidante about it, and by the following Monday he was calling me from 5,000 miles away to tell me how “Discover Weekly” was both blowing his mind and making his week.
As a self-proclaimed Luddite, I found myself particularly enchanted by what seemed like little elves in a factory spying on everyone’s music listening history and actively researching from the most extensive library imaginable using magic, telepathy, and round-the-clock work. How do they KNOW?
Taking into consideration some of the concerns of Boyd & Caplin, the Spotify conundrum is just that. Other streaming platforms, apps and social media have more obvious or predictable algorithmic tendencies—ones that are specifically designed to manipulate anything from the socks you buy to the person you elect in office. This is not to say that Spotify doesn’t control what songs show up where and what artists might get more exposure than others, but in this illuminating article by Adam Pasick, he seeks (and succeeds) to debunk the methodology behind this. What we learn through his research is that it’s 1) very dependent on other people, 2) pretty fair/progressive (in that, a fancy curated playlist made by a company or someone with a lot of followers is scanned and picked apart just as much as some guy in his mother’s basement who only listens to 5 albums) and 3) a little bit random.
“Spotify also creates a profile of each user’s individualized taste in music, grouped into clusters of artists and micro-genres—not just “rock” and “rap” but fine-grained distinctions like “synthpop” and “southern soul.” These are derived using technology from Echo Nest, a music analytics firm that Spotify acquired in 2014, which learns about emerging genres by having machines read music sites and analyze how various artists are described.” (Pasick)
It also can’t be an accident that a fresh playlist awaits on Mondays—when we might need it the most.
Caplan, Robin, and Boyd, Danah. Who Controls the Public Sphere in an Era of Algorithms? Data&Society (2016). Online.
Pasick, Adam. The Magic that makes Spotify’s Discover Weekly playlists so damn good. Quartz Media (2015). Online.