Video Advertising for the Long Tail: Behind Blinkx’s Acquisition of Burst Media
How did Burst Media, a Burlington, MA, company that helped to invent the idea of the online advertising server at the dawn of the dot-com era, end up as part of a San Francisco-based video search company called Blinkx? And how will owning Burst’s network of niche publishing sites change the way Blinkx does business? Those are two of the big questions that came to mind when I was first reporting on the acquisition last month, but I wasn’t able to sit down Blinkx CEO Suranga Chandratillake to ask them until last week.
Blinkx paid $30 million in stock and cash for Burst, which, like Blinkx itself, was a public company trading on the London Stock Exchange’s AIM market. The deal officially closed yesterday. Blinkx says it will continue to operate the Massachusetts company under the Burst name; Burst co-founders Jarvis Coffin and David Stein, formerly the CEO and chief technology officer, respectively, will depart the company after a brief transition period.
Blinkx, a spinoff of database company Autonomy that used to be in the desktop search business, now makes money by selling lucrative pre-roll video ads, which play before the Web videos that people find using Blink’s search tools. But right now all of that discovery happens on Blinkx’s own site. Chandratillake says that Blinkx has always had its eye on capturing a larger chunk of the overall cash that advertisers spend on video ads. So the basic idea behind buying Burst, he says, was to bring the company a network of “long tail” websites that may have relatively small audiences on their own but which, in aggregate, are large enough to attract advertisers who would otherwise put all their money toward TV commercials.
“The Web is increasingly becoming ‘deportalized,’” says Chandratillake (which is pronounced Chawn-dra-TILL-uh-kuh). “At least half of YouTube’s views come from videos embedded in sites outside of YouTube.com. What an advertiser cares about is getting an engaged person watching content on a particular topic, and being able to place their message in front of that user. Increasingly, those users are spending their time out in the long tail.”
Assembling advertising platforms consisting of multiple long-tail or vertical sites—think fashion or mommy blogs, for example, which tend to have small but fiercely loyal audiences—is a strategy being pursued by more than a few media companies right now. Bay Area players in this market include Federated Media, Glam Media, and Say Media. But Chandratillake says Blinkx probably wouldn’t have gone down this path this year unless the circumstances had been exactly right—meaning, unless a company like Burst had been available for purchase at a bargain price.
Burst’s story goes back to the mid-1990s, when the Web itself was still new. In January 1996, David Stein developed one of the very first ad server programs that allowed Web publishers to hand control of display advertising such as banner ads over to remote advertising networks. “They were movers and shakers in the early days, and had built up a really solid advertising network,” says Chandratillake. “They chose about 10 years ago to avoid the big sites and compile a collection of smaller sites—the argument being that Valueclick and Interclick and the other ad networks tended to compete for access to larger sites, and that advertisers buying [on Burst] would have an audience that isn’t overlapping with the other networks.”
But the display ad market became commoditized nonetheless, and declining display advertising rates throughout the 2000s left Burst with very tight margins, Chandratillake says. “Even though they had a healthy gross margin, they just couldn’t cover their underlying costs, which meant they just kept missing profitability, usually by just 5 to 10 percent of their revenue,” he says. Which meant the company was getting beaten up by shareholders, especially through the recession. “Being a non-profitable business was a very difficult place to be on the AIM in 2008, 2009,” says Chandratillake.
“When we first came across [Burst] they were pretty unloved in the stock market,” he continues. “The share price was low, they were out of cash, and they weren’t able to raise any more money. But if you dug into what the business really was, it was pretty compelling.”
In particular, Chandratillake was impressed by the scale of the network Burst had built. Its 50 sites reach 130 million unique users and collectively make up the 36th largest media property on the Internet, according to comScore. “The number one biggest complaint video advertisers have about the Web is the lack of scale, and the fact that the audiences are still very small compared to TV,” says Chandratillake. “So we said, maybe we can take something like Burst, which has this fantastic, passionate audience that’s very engaged and doesn’t overlap with other websites, and if we can convert even a portion of those users to start watching videos and seeing video ads, then suddenly we get to massively increase the scale of the video advertising we can sell.”
So, does that mean Blinkx is morphing from a video search company into a vertical advertising platform like Glam Media or Say Media? Not exactly, Chandratillake says. “The acquisition does change parts of what Blinkx is, but it also leaves in place many parts of what Blinkx is,” he says.
For one thing, Blinkx’s core business of selling advertising on its own site is still growing rapidly—the company has been profitable since the second half of 2009, has a growing cash hoard (it paid for Burst mainly by issuing new stock), and expects to see a doubling of its ad revenue in 2011 to $66 million, up from about $34 million last year, he says. So consumers will still be able to visit Blinkx.com and search the company’s index of 35 million hours of video from 800 content partners—the largest video catalog on the Web. But now Blinkx will also be able to offer publishers in Burst’s network the ability to serve videos related to their existing content, and to share in the higher advertising rates that video ads bring.
In fact, when it comes to serving contextually relevant videos, Chandratillake says Blinkx will have a leg up over competitors like Glam Media and Say Media. “They all have a video component but it’s not central to what they’ve done,” he says. “It’s all we’ve ever done at Blinkx. Whatever article you might be reading, I can almost guarantee you we will have a video on the same topic. No one else can do that.”
Of course, everything Chandratillake told me was theoretical up until just yesterday, when the acquisition was legally completed. Now Blinkx will need to approach the individual sites in Burst’s network—sites like Cooks.com, Grandparents.com, Momdot, MomsWhoThink, Politico, Runescape, and Sidereel—and look for opportunities to incorporate more video alongside the conventional display ads Burst was already serving on their pages.
“We know how big Burst is, but we don’t know to what extent we can convert it into a video network,” Chandratillake acknowledges. “Some of the sites that Burst represents are absolutely going to stay text sites. Some will change. They have a whole bunch of sites around video game content, and a lot of those will take video extremely well. They happen to have a bunch of education sites, where the crossover is less obvious.”
One of the key challenges as Blinks integrates Burst, Chandratillake says, will be “figuring out how to introduce video into these sites in a way that maintains the publishers’ control and consistency, while also delivering the quality, appearance, and presentation, which is very important from the advertisers’ point of view. But the great thing about the increasing deportalization and widgetization of the Web is that this sort of thing is easier than it ever was.”
In other words, there are lots of non-intrusive ways to offer video content to site visitors; a video might appear in embedded players, for example, or it might pop up as an overlay. “A lot of these smaller sites don’t have the scale to produce their own video, and if this means they can get video for free and get larger checks because of it, they will like it,” says Chandratillake. Burst has a strong publisher services team that will be able to shepherd the individual publishers into the new video era carefully, Chandratillake says: “That’s part of why we bought them.”
But even if Burst hadn’t come along at the right price, Blinkx would have had to do something eventually to build an audience worth video advertisers’ attention, Chandratillake says. “This allows us to flow in the direction we think the market is going anyway,” he says. “If you look at the way advertisers buy TV, there is event-style advertising like buying ads on American Idol or the Super Bowl, but a lot of people are taking the channel viewpoint, selling to kids on Nickelodeon or to cooks on the Food Network. This will allow us to create an advertising product that looks like a collection of cable networks, just on the Internet—the difference being that you will get a level of targeting and analytics that you don’t get on TV.”