Interview with Ted Sarandos

Chief Content Officer, Netflix

In June 2012, Netflix Chief Content Officer and Vice President of Content Ted Sarandos sat down with MIP. During the interview, Sarandos discussed competition, Netflix's orignal content strategy, and the relationship between audience taste preferences and Netflix's proprietary metrics and algorithms.

Ted Sarandos has served as Netflix's Chief Content Officer and Vice President of Content since 2000. Prior to joining Netflix, Mr. Sarandos was Vice President of Product and Merchandising for Video City.

Original Content

June 2012

Who are your key competitors?

It’s a little bit of everybody, honestly. We compete for consumers’ attention and time. Comcast wants to make us obsolete by improving TV Everywhere. That’s fine. They just shouldn’t be able to do it for free. We pay a very large fee for those rights, and if they’re willing to pay for them too, then we’re just straight competitors. That’s OK. I’ll still bet on us over them.

Obviously, we also compete with HBO for content and subscribers. They’re probably our closest competitor because their product is so similar to our own. People say our investment in original content makes us more and more like HBO. I think it’s the other way around. HBO is becoming more and more like us by making their content available on-demand and on mobile platforms. Our current challenge is to make better originals quicker than they can perfect what we already do so well.

Actually, there’s a flaw in that logic. Bundling constrains the market for premium television. You can’t have HBO if you don’t have $125 worth of cable. Netflix is direct to consumer. For eight bucks a month, you can have the content you want when and where you want it. I would argue that makes us far superior. We’re not behind a big, expensive cable wall. How many more people would have HBO today if bundling wasn’t part of the equation? Of course, there’s the argument that bundling actually works in HBO’s favor, but I really don’t believe that’s true.

I don’t know what to think of Amazon as a competitor. We’re real competition in the U.K. where they own LoveFilm, and so we compete for content and subscribers. But they have this funky product here where they are adding streaming to freight, which risks contradicting their core business. I wonder if they’re just trying something new? I wonder if they’re dabbling? I wonder if they’re thinking about the loss of revenue from physical media over time? Maybe they are thinking about it as a way to feed content into the Kindle. They are hard to read in that way. They do a lot of things. But they also are really smart about e-commerce and Web design. If they are going to push this product, it has got to make more money than something else in that space. We have set the bar very high for future competitors in terms of content costs. I don’t think you can get in on the cheap anymore.

Why venture into original programming?

There are a few reasons. If services like TV Everywhere and HBO GO gain traction, then they will start to attack us on the things that we believe we still do better than anybody else. Subscription. Personalization. Encoding. Multi-platform delivery. We need to differentiate ourselves on all fronts.

Our data and algorithms help us perfect personalization. Likewise, we manage that data, including credit cards, more safely than anyone else. We deliver content on more devices than anyone else. We give access to full seasons. TV Everywhere only provides the last five episodes. Hulu is completely random and differs from show to show.

Ultimately, we wanted to produce original content because it’s time we have more control over the shows that matter most to our costumers. We’ve really come to appreciate the value serialized shows provide. So many people watch them and love them. Our data supports the trend, and that’s why you see such an explicit investment in television on Netflix. We’ve been able to grow the audience for serialized content by recognizing the behavior and securing more and more highly serialized, well-produced, one-hour dramas.

Yet you discover pretty quickly that networks don’t make very many of these shows anymore because they’re expensive and they’re perceived as difficult to monetize. HBO, Showtime, and Starz are making them, but they’re also the people who least want to sell to us in the season-after model because we are direct competitors. So, at a certain point I said, “Are we going to remain dependent on everybody else making good shows or are we going to try to develop some of them ourselves?”

It would be much easier for us if HBO, Showtime, and Starz would sell us previous seasons of their shows because they’re proven and they’re good at it and we would pay for them.  But the truth is that they don’t want to open that door. So it’s time to figure out if we can become good at it ourselves. 

Also, I think it’s the direction the entire entertainment industry is heading—networks and cable channels will evolve into something like Web channels, just like radio networks evolved into TV networks, and TV networks evolved into cable channels. Look at the widgets on a Samsung Smart TV. You see Netflix. You see Hulu Plus. You see MLB. It gives you a sense of things to come. Currently, the problem is that network brands don’t really mean anything. If they want to survive, broadcasters need to figure out how to make their brands meaningful. Cable is better at this. Comedy Central, for example, will be very powerful in this new world.

When a creative comes in here and pitches a program idea, how is it different, or is it different, from them going to a network or a cable channel?

It’s different today than it will be a couple years from now too. Right now, what I’m trying not to do is build a big development infrastructure. The existing departments in networks and cable channels are typically risk management. If a show doesn’t work out and they invested millions of dollars of development, it’s not because they didn’t invest millions of dollars in development. It’s not a good reason to invest that kind of money, but people do it. It’s the same way they make pilots and test them. They spend 8 to 10 million dollars on a pilot they test to 16 people and decide not to make it because of how it tested. What I said I would do early on is vet the projects better. Let’s shift the development burden to the producer. If they’ll invest a little more in the project and bring it to us better developed, a couple of scripts, talent attached, a bible, then we can make a bigger commitment to them, meaning, I won’t give them anything short of a full-season commitment. The way we got David Fincher to jump in with us on House of Cardswas we gave him a two-season commitment. Nobody else would do that, and they all thought we were nuts when we did. The truth of it is I feel much better spending what we did knowing that I’m going to end up with 26 hours of content that at worst is going to be mediocre, and I doubt by the way, that David Fincher would create a mediocre product and put his name on it. So that was the bet. That’s as far as I wanted to bet creatively. It had the stars attached. It had scripts written. It had a showrunner. It had a bible. It had executive producers with great track records. We could have done the same thing as networks and it literally would have been 8 to 14 million dollars to shoot that pilot. This Newsroom show that HBO is doing now had the most expensive pilot in history. I wasn’t going to take the risk of spending all that money and ending up with nothing to watch. How we’re different is if they are willing to develop a little bit more, we’re willing to make a much bigger commitment.


Audience Tastes

June 2012

How has your relationship with content providers evolved? Some love you. Others see you as a threat.

I don’t seek to be loved. I just seek to be respected. The reason why a network or a studio loves you is because you make them money. If I don’t make them money, then I don’t expect them to love me. But I do make them money, and more importantly, I make them money in unintuitive ways.Mad Men is the perfect example: I not only gave AMC a very high license fee for that show, but the ability to binge on Seasons 1 through 4 also helped launch the biggest premiere [Season 5] in the show’s history.

It’s the same for Sons of Anarchy. People will stream the seasons we have before jumping to the network to watch the latest season premiere. I realize not everyone jumps. Some viewers will just wait for the next season to premiere on Netflix. But, on a net basis, more people will migrate to FX, and they wouldn’t have done so without watching the series first on Netflix. So, FX grows their audience and we derive value from the license fee. And now the network is able to produce more seasons of Sons of Anarchy. We offer a really great economic sweetener: a buyer for highly serialized content, which is very expensive to produce and very hard to monetize.

But what happens to the value of your content once a viewer makes the jump back to the network?

We’ve had Mad Men for a couple years on Netflix. Last night, what was the most watched episode? Episode 1, Season 1. There are new people coming to this show everyday. You may have bailed out and watched Season 5 on AMC, but you’re also going to discover some other show that you have never seen before. We have thousands of them. The personalized website will help you find something that you’re going to love. What I really want you to do is find a show in which you can just get lost, a show that makes you want to watch just one more episode even though you know you have to get up early tomorrow morning.

We are uniquely able to build our business model around that sort of behavior. If we pick the shows right and we invest heavily in the right kind of content, we’ll make the viewers’ dreams come true. We connect people to media in a way filmed entertainment has lost to video games and the web. We are restoring a sense of connection between consumers and content. I think audiences have lost that emotional investment in content because television can no longer provide them access in the way they want it, or in a way that matches current lifestyles. Restoring that sense of connection is the biggest shift in the economy of entertainment.

Marketing has too much influence in the current entertainment economy. Marketing is the biggest item in this town on anyone’s profit and loss statement. Fill the seats in the theater on opening night. Make sure everyone gathers at the same time on the same night in front of the television. And let’s just hope everyone likes it so numbers don’t drop 80 percent the next night or the next week.

For me, I’m doing the exact opposite. I want everyone who watches something to love it. And I’m willing to let the content take a lot longer to resonate with audiences because there is long-term value in doing so: you can’t get as much content that really matters to you from anyone else for just eight bucks per month. There’s so much to watch that you don’t have to watch at once as long as you’re always watching something that you’ll love. During the early days of the Internet when everybody else was spending big money on Super Bowl ads, we were investing instead in technology, on taste-based algorithms, to make sure every single user had a personalized, highly effective matching tool to use when they visited our site.

For us, that’s why breadth matters. We are trying to match tastes, and tastes are really specific—even in your own household. So, imagine trying to do it across the country. We have to have a lot of titles to produce the results our customers want.

How have your metrics evolved with the launch of your streaming service? 

Here is what the data from our DVD business tells us: we know what we shipped to you and we know when you returned it. I have no idea if you watched it. I have no idea if you watched it 20 times. 

With streaming, we have insight into every second of the viewing experience. I know what you have tried and what you have turned off. I know at what point you turned it off. It’s very sophisticated. If there’s a glitch in the soundtrack or something wrong in the code, the data is so refined that it can detect mass quantities of people stopping at the same point and signal a red flag within hours of the content going live. That’s a much more efficient quality assurance process. We don’t have to wait for someone to complain. We don’t have to go back to the file and watch every second of it to find and correct the problem. 

How do you use this data when negotiating licensing deals? Do you share any numbers with content providers?

We share some high-level viewing data: how many viewers and how frequently do subscribers view the content. We don’t really use the data to tell us what we should and shouldn’t have on the site. We use it to indicate how much I should or shouldn’t pay. In other words, if I can get an enormous amount of viewing, I’ll pay an enormous amount of money.

We invest in a lot of content for really small audiences, too, because it’s still valuable for our subscribers who are really engaged fans of a particular program and, therefore, it’s still a valuable investment for us. We’re fortunate because we have unlimited inventory space. It allows us to value content in more ways than just mass numbers. For a lot of other buyers, however, their threshold is very high for what makes it on the air because they only have so much space they can allocate to programming—there’s a finite amount of hours in their schedules. Unfortunately in the traditional, linear television world, new series usually succeed or fail because of marketing. Or they fail because the measurement periods are way too small. It has little to do with the actual quality of the content.

So really, we can bring some equilibrium to a business that otherwise doesn’t have it or doesn’t even want it. Traditionally, content is worth what the buyer says it is worth, but our data draws from viewer behavior to bring a bit more science into the calculation.

Can you tell us a little more about the algorithms you use?

Our algorithms are incredibly precise and draw from multiple data points. The star rating is a dependable indicator. You watched this show, and you rated it five stars. If you rated the show but didn’t stream it because you saw it in a theater, on a DVD, or you just rated it because you like a particular actor it features, the algorithm weighs the rating slightly less. Similarly, if you only partially watch something and rate it, you degrade the ratings value, too. We account for all of those behaviors. Star ratings also are one of our internal metrics. How closely can we predict the rating you give a film or television show? We can predict within half of a star or quarter of a star.

We also have implemented a lot of predictive mechanisms:  people who love x, y, and z also love these films but hate those films. We’ve made really incredible strides to predict what people watch right after they finish something else. Basically, it’s a statistical push based on what other people have watched and really enjoyed immediately after viewing the same program you just finished. There are some really wonky results here, too, like right after somebody watches Thelma and Louise, they are much more likely to watch a Geena Davis movie than a Susan Sarandon movie.  

Algorithms drive our entire website—there isn’t an inch of uncalculated, editorial space. 

How are you promoting your original content if algorithms run the site?

We treat it the same way you might promote a lesser-known movie. First, we identify attributes of the production that we think best matches the taste preferences of a large population of customers. We call it a “cold start.” Then, we immediately replace our attributes with attributes culled from first couple hundred people who rate the program. So, our marketing improves.

You will get a more prominent presentation for House of Cards if you’re in a group of people that the data indicates as likely viewers. But if you never watch anything but Dukes of Hazzard, you probably won’t see an ad for House of Cards. [Laughter] We love that you love Dukes of Hazzard. It’s not a personal judgment. [Laughter]


Read the complete transcript of our conversation with Ted Sarandos in Distribution Revolution: Conversations about the Digital Future of Film and Television.

More From Critical Conversations

VFX artist and activist Dave Rand recounts his experience working at Rhythm & Hues when the firm filed for Chapter 11 bankruptcy. 

Steve Kaplan

Steve Kaplan discusses the challenges he confronts trying to convince visual effects artists of the value of organized labor.

Thomas Gewecke talks about the value of digital ownership for consumers and distributors.

Mitch Singer talks about taking advantage of "disruption" in developing tools for digital entertainment distribution.

Dick Wolf talks to MIP about launching Chicago Fire in a changing media environment.