He had spent part of his childhood in Mexico City, where his father, Tim, worked for UPI before settling in Silicon Valley. Tim, who taught Paul to code when he was eight, eventually founded his own company, which sold downloadable business software—at least until the dotcom crash of 2000. Paul headed to Mexico, where he found work as a $25-an-hour developer, all the while feeling painfully removed from the ferment and excitement of the great digital disruption unfolding at home. “I needed to be close to the change of history,” he recalled. “It’s like there’s an earthquake happening and the land is splitting and there’s this gaping hole. It’s that obvious.”

He returned to New York, where his skills led to work at a real-estate finance firm. But the work did not excite him. Berry enrolled at NYU’s ITP, where he’d taken a class as an undergrad, studying with Clay Shirky, the media theorist. To the school’s chagrin, he kept his job and maintained a full class load.

Two weeks before teachers were to submit grades, Berry unveiled Teachers On the Run, a site where ITP students could rate and comment on their professors. It became an in-house sensation, especially after one woman posted an anonymous comment criticizing a professor for staring at her breasts. “Everybody was just in a frenzy,” he recalled. Here, after all, was an unnamed student leveling a troubling accusation against a professor in a public forum. Was this slanderous? Was it a grudge? Was it wrong to post? Berry did not think so, especially after more women added their comments, supporting the charge. The school’s administration wanted it shut down, though Shirky stuck up for him. Berry had witnessed, as Peretti had with his Nike campaign, the power an idea could muster if it found its audience.

Teachers on the Run was a crude experiment, especially compared to Berry’s project a year later for Peretti’s class: Dog Island. Unlike Teachers on the Run, Dog Island was a hoax, a make-believe resort where dog owners could send their pets, for a respite, or forever. Berry chose a rough look for his first draft. But the sense of the class was that he had not succeeded in making Dog Island feel like canine paradise; it needed to be slicker. In Peretti’s class, Berry learned there was a process to creating content with viral capability: iteration. “In Jonah’s approach to viral, there is a structure,” Berry says. “You get close to chaos in how you develop. But there’s a structure of feedback from key people.”

That structure, he was beginning to learn, meant developing an idea, presenting it to an audience, and, depending on their reactions, tweaking, adjusting, even overhauling. “It’s insanely rare that on the first try you have it right.” Dog Island did find its audience. There were dog lovers who came to hate it, assuming that sending small pets to Dog Island spelled their doom. Berry did not dissuade them, suggesting that, alas, from time to time big dogs did kill little dogs, because that was the natural order of things.

“Dog Island was the most fun, because I’m a traffic junkie,” he recalls. “There is a thrill when someone tells you about something you’ve done without knowing you’ve done it.” Something so compelling, alluring, amusing, and so beyond the need for explanation, he adds, “they have to share it.”

Berry had discovered a way to make a living by becoming a creative player in the great disruption. In the culture of “scrappiness,” failure was part of the joy of the work, as he and Peretti found at HuffPost: “Let’s have an idea on Monday. Instead of having a lot of meetings about that idea, let’s just fucking do that idea. By Wednesday, we will have realized the flaw in the idea and we’ll have iterated it, so by Friday, it’s totally different, and it’s either executed or almost done.”

And measured. Everything had to be measured. “Traffic,” he says, “was the measure of success. It showed if we could be a real business.”

But what, exactly, was that business? By 2007, Huffington Post had taken $10 million in investment. This figure was considered modest by venture-capital standards, but it did suggest, none too subtly, the nature of the enterprise: Raise money, raise the profile, raise more money, and then, when the moment and price were right, look for an exit. That could not happen without revenue, and the revenue would not come without traffic. And traffic, much to his delight, was Paul Berry’s to chase.

Michael Shapiro is a contributing editor to CJR and teaches at Columbia's Graduate School of Journalism. His most recent book is Bottom of the Ninth: Branch Rickey, Casey Stengel, and the Daring Scheme to Save Baseball From Itself.