any big-company experience,” says Camp. “During the negotiations, there were so many different points I hadn’t necessarily thought about, such as the finer points of post-integration dynamics.” For example, Camp says it became more and more difficult to get the management approval the company needed for new hires and engineering changes, especially after the 2008 recession led to company-wide hiring slowdowns and even layoffs. “There was a list of probably five to 10 things that we really needed to do, and I remember realizing that it would be very difficult to make all those changes and get back to the agile situation we were in before,” Camp says. “So at some point we suggested a spinout.”
It was a good moment for such a suggestion. John Donahoe, who succeeded Meg Whitman as eBay’s CEO in March 2008, seemed more focused on eBay’s core auction marketplace than Whitman or founder Pierre Omidyar had been, Camp observes. And discussions were already underway over the potential spinoff of Skype. (In late 2009 eBay sold 70 percent of its stake in Skype to a consortium of venture and private-equity investors.)
“We just got lucky with the timing,” says Camp. “We suggested it at the same time they were doing the other spinoff, and we were able to convince eBay executives this would be a good idea. We also got help from Ram [Shriram]—the first term sheet came from him. It was a pretty complex negotiation, but finally we got to a permutation that everybody was happy with.” In April 2009, eBay sold StumbleUpon to a group of investors that includes Shriram’s Sherpalo Ventures, Accel Partners, August Capital, and Camp and Smith themselves.
Which led StumbleUpon to Phase 4—independent again. Today the company’s 58 employees occupy offices just a block away from the startup’s original SoMa location. They’ve been busy building things like the mobile apps and a custom URL shortener (called su.pr), shoring up the service’s backend infrastructure, and improving the self-service advertising portal, which lets advertisers sign up for pay-per-visit traffic directly. (Up until 2006, the company obtained advertisers through other networks such as Commission Junction or ePilot.)
The Web-only browsing option developed at eBay turned out to be a very good bet: today one-third of stumbles happen that way. Another third come via the Firefox extension, and the last third are through add-ons for other browsers such as Chrome and Internet Explorer, plus the mobile apps.
Camp argues that StumbleUpon’s paid-inclusion model works better for advertisers than other forms of traffic generation, such as Google or Facebook ads, because the company is better at choosing sites users will like—and, just as important, at steering them away from sites they won’t like. “We collect thumbs-up and thumbs-down information for every site, and people rate ads as much as they do other content,” he says. “So we can easily figure out if an ad is good or bad, and if it’s bad we can deprioritize it.” Within the StumbleUpon ecosystem, moreover, there’s always the possibility of free exposure—what the brand advertising world calls “earned media”—if users actively share the paid sites with their followers or others.
Very few StumbleUpon users complain about the sponsored content, according to Camp. “If our algorithm works well, you will find it interesting,” he says. Some of the company’s top advertisers “are just publishers who want to drive traffic to their microsites or blog posts,” Camp says, citing Intuit’s Mint.com as an example.
Activity on StumbleUpon has doubled every year since the company started, and now exceeds 6 billion stumbles per year. “It’s getting crazy,” Camp says. “We are going to be at 100 billion in a couple of years.” Which means the company’s engineers are working hard implementing distributed databases and the other key technologies that will allow the company to scale up its backend to keep up with the growth.
A year from now, the company could have 20 million users, 40 million, or even more, Camp predicts. “If you look at Facebook, they hit the hockey-stick part of their curve in 2007 when a couple of things came together,” he says. “We don’t know when that is going to happen for us. Maybe 2011. So we’re trying to get ready.” In the end, it might just be serendipity that rules StumbleUpon’s fate.
