are starting to ‘Like’ stuff, but it’s not the most common activity,” he says. “The fact that Facebook has 500 million users but we send websites almost as much traffic as they do indicates that people are sharing and discovering about 30 to 50 times as much on StumbleUpon as they do on Facebook.”
There is, however, an important social element to StumbleUpon: you can follow other users and see what sites they like, and they can follow you. In fact, if you scan your Twitter, Facebook, or Gmail contact lists and connect with those same people on StumbleUpon, it will greatly improve what engineers call the signal-to-noise ratio—in this case, the proportion of StumbleUpon recommendations that match your own interests. That way, says Camp, “You get the best of the social graph and the collaborative filtering graph. When you put those vectors together, you can’t get that kind of information density anywhere else.”
With users so committed that they’re stumbling hundreds of times per month, it’s no wonder advertisers are attracted to StumbleUpon. “If you’ve got really good stuff, you can use StumbleUpon to seed traffic,” says Camp. “Every viral advertiser you can think of its out there using StumbleUpon. We’re not allowed to name names, but pick a big site that calls itself viral, and there is an 80 percent chance that they are using StumbleUpon ads to do it.”
As a privately held company, StumbleUpon doesn’t release financial data, but Camp says that the company is breaking even at the 5-cents-per-visit level. That’s less than the pay-per-click costs that many advertisers are accustomed to shelling out for Google AdWords ads, so StumbleUpon may have room to increase its rates. “Once we up the price we are going to have a huge margin,” Camp says.
How StumbleUpon’s model gained so much momentum—and pulled Camp and his co-founder Geoff Smith out of their Calgary student digs and deposited them at the center of the Bay Area’s Web startup culture—is mainly a story of luck, skill, and persistence. When I met with Camp last month, he broke the tale down into four phases.
Phase 1 took place in Canada in the fall of 2001. Camp had just started studying for a master’s in software engineering at the University of Calgary, and was interested in collaborative systems and recommendations. Thanks to his government stipend, he had some free time, and he began working with Smith—the roommate of a childhood friend—on a collaborative filtering extension for the then-new Mozilla Firefox browser.
A prototype was ready by February 2002. “We were the prime example of getting into the Mozilla platform early,” Camp says. “I think we were the 138th extension for Firefox ever written, and we got a top rating very early. People liked it, and we followed the comments and fixed anything that was wrong. After a while we had 4.8 stars out of five, and we started to get thousands of installs a day, for free.”
This phase continued for nearly three years. With co-founders Eric Boyd and Justin LaFrance, Camp and Smith incorporated StumbleUpon, but the team didn’t have
