The Big Data TV: Data Analytics, Algorithms, and Netflix’s Original Programming

Curator's Note

 One of the topics about Netflix’s original programming that I find to be very important but lack in discussion is Netflix’s use of big data in its original programming decisions. Netflix has always been a data-driven company as the Netflix Prize it launched in 2006 testifies. Netflix believes that with the help of big data it is able to understand what its users want and is making big bets on its original programming. Netflix’s confidence in big data (which has been expressed by people at Netflix including Reed Hastings, the CEO, and Ted Sarandos, the Content Chief Officer) was well demonstrated when it outbid major cable companies such as HBO and AMC and secured the licensing rights of House of Cards, agreed to license new episodes of Arrested Development (a show that was cancelled by Fox in 2006), and ordered a second season of Orange is the New Black even before it made its debut. 

Netflix willingly committed $100 million for 26 epsodes over 2 seasons of House of Cards without seeing a pilot of the show because it knew many of Netflix users were fans of David Fincher films, watched movies starring Kevin Spacey, and watched the original British version of House of Cards. So purchasing the rights to stream a US remake of the BBC show featuring Kevin Spacey and directed by David Fincher made perfect sense for Netflix to start its original programming. Netflix also used big data to promote its original series differently to different types of Netflix users. Creating six different versions of trailers of House of Cards, Netflix presented a tailored cut to users based on their past viewing behavior. 

As Netflix delves deeper into big data for original programming decisions the issue of what all the number crunching and data analytics would mean for creativity and originality in television production (and of course, the issue of privacy) will be debated. Will Netflix big data creativity bring more of the same stories to be told differently (after all, House of Cards is a remake of a BBC miniseries and Arrested Development is a reboot of Fox’s cancelled show) and make us miss opportuntities to encounter new and different things as Andrew Leonard warns? Or will Netflix show us new possibilities of storytelling in the era of big data by proving a harmonious blend of algorithm and creativity through its original programming?

 

 

 

 



Comments

Aaron Dickinson Sachs's picture

Same old formula, more data

Thank you for the post, you raise an interesting question. Before I launch into my own thoughts on the topic, just a quick Go Hawkeyes! since I too am an alum.

The use of viewer data in Netflix’s own production decisions is interesting, and definitely worthy of additional investigation. And I think your questions at the end are solid. In some sense, I think it will do both. On the one hand, Netflix’s use of data to produce House of Cards seems like simply a new version of the same formula that film and TV production companies have been employing for years. I remember being an intern for film and commercial director Simon West during preproduction for Tomb Raider: The Movie. Most of what I did actually involved script coverage—reading scripts that were submitted to the director and reporting on whether they seemed worthy of landing on the desk of someone who was actually being paid for their time. A key convention of script coverage is the tag line summary, which must convey the essence of the film script using only other films or TV shows as references. In other words, one might say that film script A is a horror version of Singing in the Rain, or that film script B is like Twilight meets Happy Gilmore told like it was Memento. What I was told at the time was that films are expensive ventures, so as a rule, producers want to know that a film is sufficiently like an already successful film to itself be successful, while also being just different enough to seem fresh and original to viewers. This is why so many films fit neatly into genres, why there are so many sequels, and why it seems like we are currently seeing a lot of remakes.

To me, what Netflix has done is simply take this to a new level by using user data to (potentially) more accurately distill the past successes into characteristics that can be translated into ingredients for new productions. Just as there have always been original films made by people interested in taking risks, breaking with conventions, and who are just plain tired of seeing the same derivative stuff get made over and over, there will continue to be people who will make new materials.

What’s more, the commercial model of Netflix—a subscription rather than pay-per-view model—makes it possible for Netflix to distribute its risk on producing more original content in a way that a conventional film or TV production company cannot. While having original programming like House of Cards, Arrested Development (reprise), and Orange is the New Black are certainly draws for Netflix, their economic model does net (yet) rely entirely on the individual success of these programs in the way that the success of Hunger Games relies on ticket sales for the actual film. Probably 12 or more years ago, one of my college professors, Alex Juhasz, once mentioned in class that she thought premium cable channels like Showtime and HBO were going to be in the avant gard of Tv production because of their ability to disperse risk in the production of original content while keeping subscription as their main revenue stream. I think she was right (insert the long list of critically and popularly acclaimed HBO and Showtime original series) and I think that extends to Netflix.

It is much more likely that people will subscribe to Netflix because they want to see one of these new original series than it is that they will discontinue their subscription because they don’t like it.

Feedback

1 person reported using this


You need to login to submit feedback or edit your feedback of this post!