When they inevitably write the oral history of how a crime flick involving four of Hollywood’s most prominent Black women went from social-media meme to movie, it will be the beginning that will prove most controversial. But most historians will likely point to a stray post on Tumblr, offering a wry caption to a 2014 Paris Fashion Week shot of a glamorous Rihanna and a bookish-looking Lupita Nyong’o from user elizabitchtaylor: “They look like they’re in a heist movie with Rihanna as the tough-as-nails leader/master thief and Lupita as the genius computer hacker.”
Whether that idea wended its way into Twitter, or if that plotline popped into someone’s mind apropos of nothing, is unclear. But what happened next is fairly straightforward. By late April, the prolific Black Twitter community got caught in the collective consciousness of the joking meme, and one user tweeted the idea at the Oscar-winning Nyong’o, who said she was in if Rihanna was—who then assented three days later. Emboldened, another Twitter user asked Ava DuVernay, a director whose Oscar-nominated work on movies like the Martin Luther King Jr. biopic Selma has made her one of the top names in the business; she was in. And then another asked Issa Rae, the creator of the YouTube web series Awkward Black Girl to write the screenplay; she responded with a gif of a cat, furiously typing.
And then almost exactly one month later, after a reportedly furious negotiation at the Cannes Film Festival, Netflix announced that it was actually going to make that film. And so an active community of the social-media network was able to turn a meme into reality, a moment that has been met with exultation and surprise. The people’s collective imagination giving birth to an actual thing in the world: what could be a more incredible thing?
The truth is, though, that there is only one thing surprising about the movie’s creation: how it took it so long. If anything, the whole process was merely a humanized illustration of how so many movies and TV shows are already being made: through data points and algorithms.
That’s not really new. The hunches and gut instincts of powerful studio executives have, for decades, been about connecting bankable stars with populist ideas, judged through simplistic metrics like box-office revenues or “quad segments” of audiences—male, female, over 25 and under 25. And film studios have long used focus groups to judge response; there have been countless tales of how focus groups forced changes to the endings of entire movies or the happy-ever-afters for potential love stories, from Blade Runner to Romeo Must Die to Scott Pilgrim vs. the World. But the use of actual data, rather than gut instincts and hidebound conventions, has grown by leaps and bounds in recent years. Test audiences and focus groups have been buttressed by big-data analytics. Social-media reaction and IMDB scores are being closely examined and forecasted. A 2015 article from the MIT Technology Review even did some simple data-mining to find the factors that drive successful movies (it found that established directors, more than big-name stars, were the primary driver of profits).
It is, in part, why we are trapped in a world of sequels, prequels, revivals and franchises; with huge budgets, distracted and fragmented audiences and plateauing sales, movie studios can ill-afford mistakes, and data gives them advice, even if it can tend to be boring and conservative. Reliable intellectual property like comics and toys and established fables, after all, offers some semblance of a guarantee of built-in audiences. It is, for instance, why the massive Sony leaks revealed that the studio was planning on stitching the ill-fitting Men In Black and 21 Jump Street franchises together—it was another effort to find a large, rich place in the middle of both profitable franchises’ Venn diagram. But even these rules are changing; look no further than the collapse of D.C. Comics’ former cinematic juggernaut, or the recent and epic $175-million boondoggle of Guy Ritchie’s King Arthur, which had stirred execs’ dreams of spawning a lucrative Marvel-like Camelot universe but wound up making a pathetic $14.7-million in North America in its first weekend, instead.
But for as much as the major studios are beginning to harness data to influence artistic decision-making, Netflix has been among those at the vanguard, which has mastered algorithm-driven content creation—an art form reduced to a science. With nearly 100 million users around the world watching countless hours’ worth of TV shows and movies riddled with metadata, Netflix owns incredibly rich data sets, with more generated every second by the users themselves, the kind that makes film studios’ efforts look puerile in comparison. After all, Nielsen ratings only track select TV watchers, ticket sales can’t tell too much about whether audiences finished the movie or left for a washroom break, and analysis of social-media responses self-selects for people who actually want to share their thoughts. Netflix, on the other hand, knows exactly the point at which you pause, rewind or fast forward, when and where it is you watch content, and what terms you’re searching for; it can even assess how brightly lit or sound-mixed a movie or show should be, based on in-the-moment characteristics, to make incredibly granular decisions on what shows and movies it wants to make.
The 2013 hit House of Cards was their first major success in mining that data, tapping into what they know to build a Frankenstein’s monster of a show that was engineered to be a hit before anyone had even seen it. Its data told them that Kevin Spacey was a very popular actor among its user base, and that movies directed by David Fincher were too, and knew that people watched the original BBC House of Cards series on the platform. The rest was just a linking calculation away. “Netflix’s data indicated that the same subscribers who loved the original BBC production also gobbled down movies starring Kevin Spacey or directed by David Fincher,” wrote Salon’s Andrew Leonard. “Therefore, concluded Netflix executives, a remake of the BBC drama with Spacey and Fincher attached was a no-brainer, to the point that the company committed $100 million for two 13-episode seasons.” It was seen as an expensive risk at the time, a huge amount to invest to beat out the usual suspects. It then became, at the time of its release, the most streamed piece of content in 41 countries including the United States.
“Because we have a direct relationship with consumers, we know what people like to watch and that helps us understand how big the interest is going to be for a given show,” Jonathan Friedland, Netflix’s chief communications officer, told the New York Times. “It gave us some confidence that we could find an audience for a show like House of Cards.”
So making a movie helmed by Rihanna and Nyong’o, directed by DuVernay and written by Rae—a formula spat out by the urgings of the Internet itself—is a no-brainer. Of course it’s going to be at least a little bit popular: the Internet has spoken, just in a way beyond its typical tongue of data sets. Heck, there are decent odds that, somewhere in some back room, a machine has already spit out this cinematic combination.
If anything, it boggles the mind that the studios didn’t go all in on this film; no matter their feelings or capabilities in tapping into Big Data, a passionate Twitter demand for these specific people is a clear, human, data-driven symbol of built-in audience. Why wouldn’t they take them up on it? How could they let themselves get beat by Netflix, again?
With this Rihanna-Nyong’o heist movie, the real highway robbery may well be whatever Netflix paid to make it happen. It seems likely that the math will bear it out.