If you're old enough to remember renting from a video store, you probably remember a scene like this:
CUSTOMER: I want to rent a comedy.
CLERK: What kind?
CUSTOMER: I don't know. Maybe one with that guy? The one who does that thing with his face? You know.
CLERK: [bursts into flames]
The rise of Big Data has made recommendations much more precise, usually based on the principle that you'll probably like the same things that people who like the same things you already like also like. But Netflix, whose streaming services depends on accurately recommending new titles from the ones they have available -- since the chances are good the one you actually want is not -- has gone beyond science to a kind of black art: In addition to "Popular on Netflix" and "Friends' Favorites," my page currently offers me a selection of "Critically-acclaimed Emotional Independent Dramas," "Witty Independent Comedies," and "Dramas Featuring a Strong Female Lead." (It's true -- I do like dramas and comedies.)
The Atlantic's Alexis Madrigal started to wonder how Netflix generates what they call "alt-genres," and just how many of them there were. The answer, as it turns out, is nearly 77,000, including "Cult Evil Kid Horror Movies," "British set in Europe Sci-Fi & Fantasy from the 1960s," and "Time Travel Movies starring William Hartnell."
None of these is the answer to "Hey honey, what should we watch tonight?" But by "tear[ing] apart content," as Netflix's V.P. of content Todd Yellin puts it, they're able to analyze on a cellular level what people want to see. "The data can't tell them how to make a TV show," Madrigal writes, "but it can tell them what they should be making. When they create a show like House of Cards, they aren't guessing at what people want."
This is, in its way, not so different from how Hollywood has approached movies for years: "Horror's big right now," or "People are tired of science-fiction." But Netflix takes it much, much further.
As Netflix is still only dipping its toes in content creation, its hard to know exactly how this kind of data crunching will affect the future of the industry, especially once other companies try to get in on the act. The genre-creation widget at the top of the article offers an absurdist take on how oddly specific the results might get, but it might not be too long before a data-collection powerhouse joins forces with (or acquires) a media conglomerate, at which point things could get really interesting.
The only semi-similar project that I could think of is Pandora's once-lauded Music Genome Project, but what's amazing about Netflix is that its descriptions of movies are foregrounded. It's not just that Netflix can show you things you might like, but that it can tell you what kinds of things those are. It is, in its own weird way, a tool for introspection....
The human language of the genres helps people identify with the recommendations. "Predicting something is 3.2 stars is kind of fun if you have an engineering sensibility, but it would be more useful to talk about dysfunctional families and viral plagues. We wanted to put in more language," Yellin said. "We wanted to highlight our personalization because we pride ourselves on putting the right title in front of the right person at the right time."