Skip to main content

TCA: Netflix Executive Refuses to Reveal OTT’s Secret Sauce

Related: Complete Coverage of TCA Summer Press Tour

Beverly Hills, Calif. — Finally, a Netflix panelist at the TCA summer press tour at the Beverly Hilton said the magic words:

“I do know how many people watch this stuff.”

That’s because, as the OTT’s VP of product innovation, it’s Todd Yellin’s job to help figure out what Netflix’s 65 million subscribers watch, when they watch and how long they watch. Just don’t expect him to tell you — or even the streaming service’s A-list showrunners, how many viewers Orange Is the New Black has.

“You’d have to put me on a stretch rack and it would get really ugly because we don’t want people to get distracted with the numbers,” Yellin said. “We are very proprietary about the secret sauce about how we weigh different things in the algorithms because there’s been a lot of sweat and a lot of labor to get there.”

Two-thirds of Netflix’s viewing hours are from TV series, Yellin said. And the viewing habits of subscribers, Yellin said, are like their fingerprints.

“Rating counts a little bit in our algorithms but it’s not as important as history,” Yellin said. “Not just what I watched yesterday but how much did I watch of what I watched yesterday? What I watched last month? A year ago? How long did I wait?”

For example, after House of Cards launched in 2012, Netflix executives analyzed what shows viewers watched after it. The British version of House of Cards and The West Wing were obvious. It’s Always Sunny In Philadelphia was a surprise, Yellin said.

“We totally use who watches something,” he said. “We’re just using it to make the service better. Our content team is helping to create more diverse slate of originals so we have something for everyone.”

But the data, Yellin maintains, doesn’t directly influence the OTT’s creative output.

“It’s an ugly world where you could plug all these variables into a computer and then come out with the script that’s printing before our eyes,” he said. That’s not happening any time soon. That’s not going to happen in my lifetime. The data we use helps inform the process to make great programming but it’s one facet. A lot of is old-fashioned creativity and good taste and smart choices.”