Anyone who works with big data should admire the algorithmic ambitions of the new Netflix series House of Cards. The program stars Kevin Spacey as a Svengali-like political figure, but the real evil genius behind House of Cards is how Netflix concocted every facet of the series based on complex data sets of prior Netflix user behavior. Netflix has so thoroughly analyzed viewer patterns that they now know what viewers enjoy and don't enjoy—before viewers even see the show.
You could say that that Netflix's House of Cards is based on a 1989 novel by British author Michael Dobbs (it is). You could also say that it is based on a 1990 BBC miniseries (it is). But more than anything, the US version of House of Cards is based on deep statistical analysis of years worth of documented Netflix user behavior—which actors we prefer, how long we remain logged in and at what intervals we tend to hit the Pause and Play buttons.
Each of these behaviors produced actionable guidance for Netflix to create a "guaranteed hit." As noted in a thoughtful profile on how Netflix uses big data, The New York Times points out that Netflix "knows what people want before they do."
Let's start with the trailer for this Netflix original series called House of Cards. Did you see the trailer? I did, and I thought it looked great -- but I saw a different version of the trailer than you saw. Netflix made 10 different trailers, and the version presented to each viewer was tailored to whether that viewer favors comedies, suspense films, action films, male lead actors, or female lead actors. By tailoring the preview to each individual viewer, Netflix was able to maximize viewer engagement.
Big data also determined who would be cast in the series. Netflix's data sets told them that viewers tend to watch the films of Fight Club and The Social Network director David Fincher all the way through, rather than pausing or stopping them. This behavior told Netflix that their customers are wild about Fincher's work, and he got the gig to direct House of Cards primarily on the basis of his films' Netflix viewer engagement patterns.
Similarly, Spacey was cast in the lead role not based on his performing talent, but rather on analysis of how his films perform specifically on the Netflix platform.
"Netflix and Amazon know when you stop and start a program, whether you wanted the whole thing, all of that," said Rick Smolan, author of The Human Face of Big Data. "I end up paying $150 for channels full of nothing I want to watch. These guys know what they are aiming at."
The next thing these guys (and ladies) at Netflix are aiming at is another round of original programming, including their reboot of the series Arrested Development and a horror series called Hemlock Grove. Netflix is not worried whether these shows will be a hit with audiences. Their big data bots have already analyzed and determined that these shows will be hits.
This article is part of Allvoices’ series on ad:tech, the largest, longest-running digital marketing and technology event. Check out allvoices.com/adtech for more of Allvoices’ ad:tech coverage. This series is supported by ad:tech.