TokBox Works to Break Video Chat Out of Its Siloes, Flank Google+, Facebook

8/4/11Follow @wroush

Imagine that “EJ,” the now-Internet-famous San Francisco resident whose home was systematically robbed and vandalized by Airbnb guests, had been able to interview her potential renters beforehand via video chat on Airbnb’s website. She might have picked up on danger signs and avoided the heartache of having her home utterly plundered and her identity stolen. Well, as it turns out, facilitating “richer communication between guests and hosts before booking” via video chat is one of the new safety features Airbnb is now considering, at least according to this guest post on TechCrunch by CEO Brian Chesky.

Until recently, though, it would have been hard for Airbnb or any other startup to add such a capability on a whim. Video chat was a standalone, siloed technology—think Skype or Chatroulette—and building two-way video or group chat capabilities into a typical business or social-networking site would have required expensive software licenses and months of integration work.

But video chat technology is gradually becoming more accessible, and a 30-employee startup in San Francisco called TokBox is leading the change. Once known for running its own branded video conferencing service, TokBox has spent the last couple of years reinventing itself as a provider of free software widgets and application programming interfaces (APIs) that make it easy for other Web companies to add group video chat features to their sites. The startup officially killed its own service this spring in order to concentrate fully on the new platform, which is called OpenTok. “It was a hard decision because there were lots of people who loved that service, but it was also easy because we believe the OpenTok opportunity is much larger,” CEO Ian Small told me in April.

That’s proving true so far—the company plans to announce soon that its platform has been installed on 13,370 websites. (Why such a weird number? Partly, Small says, because the startup was too busy to put out a press release when it passed the 10,000 milestone, and partly because it’s a geeky inside joke—13,370 is 10 times 1,337, and 1,337 spells “leet” in the alternative Internet alphabet known as leetspeak.)

The beauty of OpenTok’s technology is that it lets developers drop live, Flash-based video-chat windows into almost any kind of website, anywhere on the page that they like. A site called Pokerview, for example, uses OpenTok to let poker players see each other during games; the video windows appear in a circle around a virtual poker table. Another site, TubeIt, lets people view each others’ webcams while they collectively watch YouTube videos.

The OpenTok concept is thus an inversion of the usual video chat paradigm, where the video stream itself is front-and-center and users have nothing to focus on but each other. “Content fuels conversation,” says Small. “The ‘aha’ moment for us came when we noticed that people who watch YouTube movies together hang out for an hour or more. They’re half talking, half watching, half chilling. So we said, ‘Why are we spending time trying to figure out how to fit the Web inside our service, instead of taking our service and fitting it to the Web.’”

But a lot has happened in the video chat world since my first talk with Small almost four months ago. The two biggest news item, of course, was the introduction of Google+, which includes a group video chat feature called Hangouts. Then there was the deal between Facebook and Microsoft’s Skype division, allowing Facebook users to make Skype video calls from within the social networking site.

It’s rarely a pleasant experience for startups when Internet giants start treading in their markets—but Small told me this week that he long expected such moves from Facebook and Google, and that the new features are actually helping to increase awareness about video chat technology, as well as consumers’ comfort level with the whole experience. “Since Hangouts and the Facebook-Skype integration launched, in that five weeks or so we he have seen literally a 200 percent increase in inquiries about OpenTok,” Small says.

And it’s far better to have Facebook or Google doing OpenTok’s advance marketing work than Chatroulette, he adds, considering that site’s reputation as a hotspot for nudity. “A year ago, lots of people were having their first experience with two-way video as a result of Chatroulette, and it didn’t necessarily give people the best experience, so it was a two-edged sword,” he says. “I think Facebook and Google are going to double down on that awareness trend, but people are going to have good experiences, so it’s a one-sided sword, which I like a lot better.”

It has taken a while for TokBox to find a path to growth. The company was founded by former Cornell students Ron Hose and Serge Faguet in 2007 with backing from Sequoia Capital, and was known early on for providing a six-way video chat feature on the social networking site Meebo. By 2008 the company had raised $14 million in Series A and B funding from Sequoia and Bain Capital, and was taking the cheeky step of sending a taco truck to Yahoo to recruit newly laid-off employees. But uptake for TokBox’s chat system was slower than expected, and in mid-2009 Small, a former executive at database company MarkLogic and one of the original developers of the QuickTime video protocol at Apple, was brought in by Sequoia as a turnaround CEO.

The original idea for turning Tokbox from a standalone product into a platform that developers could tap came from customers, Small says. “We talked to a bunch of people who were interested in integrating video chat into their websites, but they said ‘We don’t want it in this one-size-fits-all way.’ The only model at the time was the Brady Bunch model—a box with six heads in it. We started hearing that it would be interesting if we could break up that metaphor. We did some experimenting, and when we saw what happened when you stripped the chrome off and started delivering video into the middle of a Web page, we knew there was something fundamental about it.”

By September 2010, Small had decided that the company would make a full pivot, turning off its branded video chat service and shifting all its resources to the OpenTok platform, which it unveiled in November. At the same time, it raised $12 million in Series C financing, with DAG Ventures now in the lead but with Sequoia and Bain still along for the ride.

TokBox is now betting that more and more website owners will see that adding face-to-face communication can make almost any other Web-based activity more social and more engaging, whether that means playing games, shopping, watching movies, or even coding software. “Some enormous percentage of communication is non-verbal—from 50 to 90 percent,” says Small. “To pick one example, think of two things that are very popular these days in startups—distributed engineering teams and pair programming. Those two things don’t go together very well, but what if you could take an online development environment and put a little video of the person you’re working with right next to the code base? Now all of a sudden you can see out of the corner of your eye that the other guy is frowning. Which lets you ask ‘Why don’t you like what I’m writing here?’ That’s what you lose without face-to-face.”

Under the hood of OpenTok is a sophisticated, cloud-based, telecommunications-grade system for opening video chat sessions and allowing an unlimited number of people to watch or “subscribe” to the streams that the main participants are “publishing.” This flexible publish-subscribe model lends itself to many sorts of online gatherings, depending on what rules third-party developers want to specify. “If three of us are all publishing and the business logic says ‘Subscribe to everyone publishing,’ we have a traditional business conference,” explains Small. “If we were all publishing to the same session but the logic says, ‘Connect to one other person randomly,’ you have Chatroulette. If there are 1,000 people subscribing and just two people publishing, you have a talk show.”

Only a company as big as Google could muster the resources to replicate a system like this internally, Small says. “If you ask me whether I’m surprised that they could pull it off, no I’m not,” he says. “They’ve had a video chat offering inside Google Talk for a while, and Hangouts is built on that same technology. Frankly, when we were looking at the market 12 months ago, we projected that both Facebook and Google would be in this market with solution as destination chat sites, because Google already had the technology, and Facebook and Skype have had a partnership for a long time.”

But neither development presents a big obstacle for TokBox, in Small’s estimation, since Google Hangouts is tightly integrated into Google+, and Skype isn’t opening up its platform to developers outside of Facebook. Both systems are examples of what Small calls “a destination approach” to the world. “What about the millions of other sites around the Web that might have a reason to want to have face-to-face communications between their users?” he asks (a point he also addressed in a recent post at the TokBlox blog).

If Small is right about that “millions,” a lot of people will be knocking at TokBox’s door over the next few months. Which means the startup will have to start thinking pretty soon about the last piece of its formula: a revenue model. Right now, websites can tap OpenTok—the widgets, the APIs, and the whole cloud-based back end that supports the actual video conversations—for the unbelievably low price of zero dollars. Small says the basic service will stay free, but that the company will eventually introduce premium features that entail a subscription fee, such as high-definition streaming, encryption, or video archiving.

The startup may also look for ways to introduce advertising into its video streams—but in a way that gives publishers a cut of the action. “Right now we are trying to open up a future, let people get their toes in the water, and use it to their hearts’ content,” says Small. “But we are, after all, delivering live video, so delivering advertising is not that hard, if we do it in a way that is respectful of the sites and lets them participate.” Airbnb—which is not yet a TokBox user, though the two companies share an investor in Sequoia—probably wouldn’t mind that tradeoff.

Wade Roush is a contributing editor at Xconomy. Follow @wroush

By posting a comment, you agree to our terms and conditions.