Netflix Uses Data to Drive Creativity, and It Terrifies Hollywood
“There’s a whole thing going on in Netflix right now and in Silicon Valley saying, ‘We’re going to use algorithms to make creative decisions.’ I say ‘posh.’ You can’t. It’s not like making a computer program that can work in the fixed and formal rules of chess.” —John Landgraf, CEO, FX Networks
Is big data a help or a hindrance to creativity? I had the chance to explore the topic with Michael D. Smith, a professor of information technology and marketing at Carnegie Mellon’s Heinz College and Tepper School of Business and the author of Streaming, Sharing, Stealing: Big Data and the Future of Entertainment, cowritten with Heinz College professor Rahul Telang.
Netflix has a lot of data about what we like to watch. It began collecting data years ago, when it was in the business of sending DVDs through the mail. It turned to that data when it began commissioning original movies and series. It was able to make smart decisions about what genres, directors, and actors we wanted to see, and how to fund each project based on expected viewership.
That approach has led to some critics, like FX’s Landgraf, accusing Netflix of programming by algorithm, of replacing intuition, experience, and Hollywood know-how with a computer routine. To critics, the Netflix formula produces hollow results and ignores the value of the creative process.
What Smith and Telang show in their work is that Netflix doesn’t use data for creative decisions; it uses data to match content with viewers. Netflix is excellent at getting out of creative peoples’ way, Smith says. Unlike studios, it lets its talent do its work without a lot of notes or advice. When projects are completed, it uses its data to match them with viewers who will mostly likely enjoy them.
Doing so not only drives views, it also helps with discovery. Netflix churns out a lot of original content, and subscribers could easily feel overwhelmed if they had to wade through it all. But they don’t—they only see the titles Netflix is pretty certain they’ll like.
It’s something the big broadcasters can’t do, Smith notes. Broadcasters can only show one program at a time, so they go with whatever they expect to get the biggest return in each slot. Netflix doesn’t have to turn every show into a blockbuster, so it’s free to create niche hits that appeal to specific groups. The power of Netflix’s data is in microtargeting.
With that in mind, Netflix’s four-picture deal with Adam Sandler starts to make more sense.
Hollywood creatives have been surprisingly vocal about their dislike of Netflix. Consider that Steven Spielberg joined the Netflix hate squad in March, saying Netflix output should be considered TV movies and not be eligible for Oscars.
Some of this comes from wanting to protect the traditional theater model or the glamour of a theater opening. Some people look at Netflix’s data and think it has an unfair advantage.
“Netflix can do things that would be very difficult for Hollywood to copy,” Smith says. “There’s a quote that we use in the book where [Netflix chief content officer] Ted Sarandos is sitting around a table of TV execs and he says, ‘The difference between my business model and everybody else here is that we can hit singles and doubles and the business is just fine.’ I think what he’s saying is the traditional business of Hollywood is, blockbusters are everything. You’ve got to have the blockbusters. It’s hard to sell a piece of niche content in any of the existing channels. What Ted Sarandos is saying is that we can sell niche content. If I can find an audience who likes this and is going to subscribe next month, it doesn’t have to be a big audience. I just have to be able to find them.”
In the 2018 Oscars, Netflix had two big winners, Mudbound, which was nominated for four awards, and Icarus, which won best documentary feature. Both were movies Netflix acquired rather than commissioned, but the company still used its data to find each an audience (after brief theatrical runs). Some Academy members were vocal about not wanting to vote for a Netflix movie.
While studios condemn Netflix’s business model, they’re doing their best to reproduce it. The mergers and acquisitions we’re seeing these days are all about major media companies trying to own customers’ data so they can serve targeted content the same way Netflix does. Models are changing, and studios see they’re at risk if they don’t own the data, so there’s a race to best serve the viewers.
“Netflix is trying to vertically integrate upstream into content creation faster than the content creators can vertically integrate downstream into distribution,” Smith says.
Big data is firmly a part of Hollywood today, but—for now, anyway—creative decisions are still left to creative people.
[This article appears in the June 2018 issue of Streaming Media magazine as "Netflix Uses Data to Drive Creativity, and It’s Terrifying Hollywood."]
Related Articles
In 1993, Omni Magazine showed its readers in-car video technology that let people watch as they drive. It was both decades ahead of its time and hopelessly misguided.
04 Nov 2019
Is the leading SVOD worried about the upcoming debut of Disney+? No, it says, noting that Fortnite is a bigger competitor than HBO.
18 Jan 2019
With its logo program, Netflix hopes to make the technical side of uploading finished movies a little less confusing for storytellers.
14 Sep 2018
The SVOD leader is diving deeper into interactive storytelling, this time partnering with Telltale Games for a series based on the popular video game.
14 Jun 2018
Strong demand outside the United States—especially in Asia Pacific—will propel record growth at Netflix, forecasts Digital TV Research,
21 May 2018
Sometimes an industry benchmark—such as Netflix's ISP Speed Index—only tells half the story about video delivery. Before picking a codec, dig a little deeper.
16 May 2018
A group of powerful studios wants to shut down Set TV, which sells a $20 per month subscription service that delivers 500 premium channels.
24 Apr 2018