The TV Revolution Will Be in Slow-Mo, Says Flingo CEO
Henry Blodget, the CEO and editor-in-chief of Business Insider, argued in a column this week that the television industry could suffer a sudden revenue collapse as consumers shift their TV viewing to on-demand, Internet-mediated services like Netflix and iTunes. Blodget thinks network and cable executives are in denial about the danger—just as newspaper publishers were until about 2001, when their advertising revenues really did go off a cliff.
I’ve offered arguments similar to Blodget’s in the past, and to be honest I’d like to believe they’re true, because I think the TV business needs a good jolt. But lately, people whom I trust in the media business have been telling me otherwise. They say the real shift to all-digital, on-demand TV consumption is going to be a long time coming, and that there will be plenty of opportunities along the way for both the media giants and small startups to try out new business models.
One of those people is Ashwin Navin, who’s best known as the co-founder of BitTorrent Inc. and has spent the last four years building a new TV-tech company called Flingo. The San Francisco startup, which raised $8 million this February in a financing round led by August Capital, is betting that an array of new digital services will pop up alongside traditional TV, but that consumers will be very slow to let go of old-fashioned “linear” TV, meaning the programming being pushed to them live by their cable providers.
“We have 50 percent DVR penetration in this country, but 85 percent of TV watching is still live,” says Navin, who is Flingo’s CEO. “The behavior is built in. People aren’t voting with their feet.”
Flingo’s answer is to create technology to help bridge the Web and traditional TV. The company started out building software that video publishers could use to bundle up their content into “channels” for delivery via devices like the Roku Player or Internet-connected TVs. If you subscribed to the TMZ channel or the Funny or Die channel on your Roku box, for example, Flingo was the company powering it. But the company soon pivoted to what Navin now believes is a much bigger market: tools to help developers and publishers build apps that interact directly with linear programming.
For anybody who wants to build an app that interacts with TV, the first step is simply figuring out what the user is watching. So Flingo built a video “fingerprinting” engine that can suck in a short snippet of video, extract key features, compare this signature to a database of past signatures, and—within a second or two—determine what show the video is from. (It’s like the Shazam app for identifying song titles, except it produces an answer a lot faster, because Flingo has more information to work with.) Flingo’s algorithms are part computer vision, and part real-time search and pattern-matching; co-founder and chief technology officer David Harrison, who was also the top engineer at BitTorrent, is a hard-core computer vision PhD who did his postdoc in a computer-vision in the Berkeley Video and Image Processing Lab.
Navin describes Flingo’s service as “GPS for television,” in the sense that it provides contextual data that can be tapped to enhance other services. “Yelp and Google Maps and all these other great apps would be really difficult to use without GPS,” he points out. “Identifying video signals—independent of the delivery mechanism, whether it’s cable, satellite, or over-the-air—could be just as significant, and could help smart TV applications go from being kind of a novelty to being really valuable to the user. The only difference [between GPS and Flingo] is that the TV industry is much, much bigger.” The company is simultaneously working with TV manufacturers like Samsung, LG, Sony, and Vizio to integrate its software into their next-generation smart TVs, and with TV networks like Fox, A&E, and Showtime to come up with interesting ways to apply the information.
What I find intriguing about Flingo is that it’s looking for ways to succeed in the TV industry without trying to undermine the industry’s existing business models, the way some bigger players, like Netflix, Apple, or Google, seem intent on doing. This theme—about using Internet tools to enhance rather than reinvent the TV experience—came up over and over in my talks with Navin. (I visited Flingo for the first time in the fall of 2011, and went back for an update earlier this week.) What follows is an edited version of our conversations.
Wade Roush: How does a guy who was the co-founder of BitTorrent, the file sharing service that was in some ways the bane of the entertainment industry, become the head of a company that’s working so closely with the TV industry?
Ashwin Navin: BitTorrent was very successful with a population of technology early adopters. A hundred million or so people use it every month around the world. We couldn’t get the mass market to use BitTorrent, but we saw that the mass market was hungry for better content experiences. We started this company believing we could reach a much bigger population by building great software for inside the TV. We think Flingo represents a much bigger opportunity to affect many more people and open up new business models that are consistent with the way broadcasters and cable networks create TV today.
The central theme is that TV is the last frontier for connectivity and service. The Internet has changed our lives in every way, but TV continues to sit on this island and hasn’t really been improved or affected by the Internet at all.
WR: Many people have tried to draw a sharp distinction between TV and the Internet by saying TV is a lean-back, passive experience and the Internet is a lean-forward, interactive experience. How do the two fit together?
AN: We need to take the best of both. There is the ability to very explicitly say, this is what I want to watch, put it there. But there is another experience that we are very optimistic about, which is, I’ve got my TV on, and I’ve got my laptop or tablet going. The things I use the laptop or tablet for are the interactive things—e-mail, Facebook, Twitter. So I’m watching video casually and interacting socially. To bridge those together, it would be nice if the Web were explicitly aware of what is going onon the TV.
So we have two interests. One is to take the context of broadcast out to the Web, and the second is to bring the context of the Web back into television. I think the most immediate impact is going to be on the traditional TV experience. A lot of companies are trying to replace the broadcast delivery mechanism, but there are some major challenges and it’s going to take longer than some people expect.
WR: Yes, you were speaking about Henry Blodget’s post, where he predicted that TV advertising revenues are getting ready to go off a cliff because of digital delivery. You think he’s wrong about that—why?
AN: I think there are a lot of reasons. First of all, the Web made a very disruptive substitute for print, but it does not make an immediate substitute for video. It’s complementary, but not as valuable or relevant to people as prime-time TV is. The media companies on the TV and film side are looking at what happened to newspapers and music and saying that embracing digital is the right thing to do, but don’t make it a substitute for your core business.
Look at it this way: In an environment where you have 9 or 10 percent unemployment, and people should be feeling the pinch, you only have 3.5 million cord cutters. The people in San Francisco and New York who blog about the demise of TV are not necessarily representative. It’s the 99 percent that we need to be listening to.
WR: Behind what you’re saying, I think I hear the message that Flingo is a small startup that needs to work with big media companies, so you don’t want to talk in a way that antagonizes them.
AN: I like to say that there are two types of technology companies in the world: revolutionary and transformational. Skype was a revolution in the phone industry, because it replaced long-distance calling. Yelp is transformational. It’s not a replacement for restaurants—it enhances the experience. We think of ourselves as more transformational than revolutionary. We are not trying to replace TV or cable. We are trying to enhance the experience of watching TV.
WR: But surely, at some point on-demand, Internet-mediated video really is going to replace traditional TV for most people.
AN: Everyone in the business knows that, even now. That’s the reason they’re bidding up the value of content, to the point where Netflix is now running at a loss. They know it’s a matter of time, a race against the clock, before people’s behavior changes. But betting on rapid change in behavior patterns in this area has usually been a bad bet.
WR: Okay, describe Flingo’s business and how it fits with current TV-watching behavior.
AN: We started out building software that video publishers could use to create channels for the over-the-top services like Roku or the app stores or channel stores on connected TVs. That was our legacy, and it gave us relationships with manufacturers of TVs and chipsets and publishers of content. But what we realized is that if you add up all of the usage of smart-TV apps, it’s a very small fraction of what people do with their TVs. If you look at the numbers, 90 percent plus of the utilization of the screen is to show live cable, satellite, or over-the-air broadcasts. In other words, all of this TV “smartness” is not really relevant.
So for the last two and a half years, all we have been doing is figuring out how to integrate smart TV apps with traditional TV, and that’s the framework we call SyncApps. It allows apps to synchronize with a broadcast. It doesn’t matter whether you have time-shifted the broadcast or done some other manipulation. We can identify the content using computer vision and signal processing to make a confident determination of what that show is.
WR: How do you put that information to work?
AN: We’re presenting SyncApps first and foremost as a social TV experience. In our proof-of-concept you see the show title, a URL, and a short code in the corner of your TV screen. You visit the URL on your laptop or phone or tablet, and punch in the code, and you see that we have pulled all the discussion happening about that show on Twitter and Facebook. You can watch the conversation in passive mode, or could then push a check-in or comment back to Facebook.
It’s a second-screen experience. There’s no need to turn the TV itself into a computer when 60 or 70 percent of people already have a computer in their hands while they are watching TV.
WR: But there’s no Flingo app, right? The vision is that it would be makers of second-screen apps—whether they’re independent publishers or the broadcasters themselves—offering that synchronized experience. For example, a company like Dijit Media could use this to automatically synchronize their social TV app with what’s on TV.
AN: Exactly. There is an API that Dijit or GetGlue or Miso or any of these companion experiences could use to contextualize their applications. We want to be the GPS, not the aggregator. I think there are going to be multiple flavors of these companion apps out there. Some will appeal to sports fans, and others to movie watchers, and others to the networks themselves. CNN is going to want an app and probably wouldn’t be super excited about some other app running against their network. We want to stay at the technology layer and empower the ecosystem.
WR: How does Flingo earn revenue in this equation? Do you charge app makers or networks for a certain number of API calls?
AN: Right now the API is open and free, but I imagine we will have some parameters around that in the future. Right now the API returns the name of the show and the time stamp, so you know where you are in the show. But it can also expose the name of the advertiser during a commercial—and this is where the business model becomes a little more clear. Today, a TV ad buy is restricted to the primary screen, but imagine if you could synchronize that with an advertising experience across multiple screens. Then the value of an ad buy would go up. Even if you were time-shifting, the advertisers would still have the opportunity to be integrated with a show, through all these other access points that don’t disrupt the linear programming.
WR: You’ve mentioned being like the TV ecosystem’s equivalent of GPS. In the actual location world, that’s a role played by Skyhook Wireless, which has a giant database of WiFi access point locations, which they license to mobile companies so that handset users can discover where they are on a map. But now other companies like Google and Apple have come along and sampled WiFi network locations and built their own databases. What’s to stop someone from coming along from behind Flingo and creating their own database of video fingerprints?
AN: We believe the technology is hard, and we have filed for a lot of intellectual property protection. But what we have found is that pure technology companies are not as exciting as those that can build ecosystems. About half of our company is engineering, and the other half is doing nothing but ecosystem development, meaning working with chipset and TV manufacturers and TV networks to make sure they have the tools to influence the SyncApps experience. If there is a Twitter hashtag they need to be showing between minute 20 and minute 24, for example, they can do that easily with our Web-based content management system. Those are the kinds of things our partners will have, and that is not going to be something that will be easy for anyone to displace.
WR: What about the companies offering over-the-top Internet video experiences, like Google, Apple, Boxee, or Roku. If those players gain a lot more users, how does that affect Flingo?
AN: There are companies that think of broadcast TV as an app, and companies that think of broadcast TV as a platform. Google TV is a perfect example of those who think of it as an app. You’ll see thousands of apps on Google TV and broadcast is just one of them, and everything is search-driven. Other companies in that general mode of thinking are Netflix, Roku, and Apple. They are going to keep marching down that path, because they are trying to displace the Comcasts of the world with something else.
Then there is the broadcast-as-platform group. We fit into that camp. We are saying that out of the 1,000 things in the Google interface, there is only one that people actually use, and it’s the HDMI cable-in, meaning traditional broadcast TV. So let’s unclutter and improve the broadcast experience. Other companies in this general bucket are companies like Sling and TiVo that allow place-shifting or time-shifting. We’re saying, let’s make it more social and enhance broadcasting rather than revolutionizing it.
WR: Even though you’re trying not to sound too revolutionary, I’m guessing that some networks and studios are a lot slower to warm up to your technology than others.
AN: There are two types of media companies. There are the ones that get really excited about standards and new technologies and ways to influence early adopters. And there are others who can’t be bothered to look at the newfangled stuff and will wait until it is baked to jump in. For years, I was the guy who convinced Hollywood to embrace BitTorrent and connected TVs, so I have a pretty good handle on which networks fall into the first camp and which ones we can bring into Phase Two—which ones will say “Come back when there are 30 million of these things out there and it will influence my Nielsen ratings” and which ones will say “Holy shit, this could influence my Nielsen ratings and I need to be involved from Day One.” We like those companies and want them involved.