There are a handful of fascinating bits this New York magazine profile of Jonah Peretti, founder of BuzzFeed. We read, for example, about one of Peretti’s first encounters with going viral--an email thread with a Nike customer service rep that was forwarded far and wide, essentially making Peretti a celebrity labor activist overnight. But the article’s most interesting insights focus on something much more current: How, exactly, BuzzFeed plans to make money.
The site’s model is based around sponsored content, meaning that instead of selling advertisements to companies like Virgin Mobile and Pepsi, BuzzFeed sells them posts on its site. Sometimes, these are tangentially related to the brands at hand (JetBlue, for example, sponsored “The 50 Most Beautiful Shots Taken Out of Airplane Windows), sometimes they’re not.
But with these posts, BuzzFeed isn’t just selling advertisers its built-in readership. It’s also selling a sophisticated understanding of what makes things go viral on the internet. Part of that knowledge stems from BuzzFeed’s origins--not as its own editorial entity, but a sort of pre-viral aggregator of other web content.
Originally, BuzzFeed employed no writers or editors, just an algorithm to cull stories from around the Web that were showing stirrings of virality. In return for functioning as a sort of early-warning system, BuzzFeed persuaded partner sites to install programming code that allowed the company to monitor their traffic. (The network now encompasses some 200 sites that serve 355 million users.) This analytical capacity, which the company doesn’t talk about much, has given BuzzFeed an enormous trove of data about what information people are reading and how they are sharing it.
And how, exactly, do those mega-stories spread? The article introduces us to Duncan Watts, BuzzFeed’s official science adviser and a social scientist:
But his research suggests that the commonly understood, Gladwellian model of virality, with its linear progression through influencers and tipping points, doesn’t really reflect the way viral messages spread. Instead, he says, they tend to grow from seeds scattered in little clusters, popping up all at once like toadstools after a rainstorm. BuzzFeed has found its most popular posts don’t take off because Kim Kardashian shared them but because many people did in small groups—the median figure is just nine Facebook friends.
A real virus is constrained by biology. A disease begins at an origin point—its Patient Zero—and it will continue to spread so long as its reproduction rate stays above one, meaning that each person communicates it to at least one other on average. If its reproduction rate drops below one, the contagion will die out. But a marketer doesn’t have to be constrained by nature’s rules. Seeds can be planted in many places at once, and even if they reproduce only at a fractional rate, that can add up. Watt’s calculus looks like this: n = pN(1+ R+ R^2 + R^3 + … ) = pN/(1- R).
Essentially, the idea is that you don’t need George Takei to share your post for it to make a splash. A handful of disparate, less-influential individuals sharing it can have the same effect.
What’s even more interesting, though, is how Buzzfeed’s applying this understanding to sponsored posts. They pay to plant all those pre-viral seeds, buying their way into people’s social media feeds through Facebook’s sponsored post system and others. Essentially, BuzzFeed’s paying to advertise their advertisements.
Where BuzzFeed gets scientific, Peretti says, is after publication. “We understand how the system works,” he told me. The company practices “viral optimization,” Peretti’s term for promoting the messages that work and ruthlessly starving those that don’t. BuzzFeed has released some selective data about the fractional proportion of sharing it achieves—its so-called “lift”—and claims that for the median advertising post, 10 paid views yield around three shares. Peretti adds that the brands that have embraced the format most enthusiastically have better results. Virgin Mobile’s ratio of shares to paid views is better than one to one.
This isn’t the first time we’ve seen the idea of engineered virality put forward as a new frontier of advertising. ThinkModo, a relatively new agency in New York City, specializes in creating videos that will explode on the Internet, hopefully dragging a brand along with them (they, as it turns out, were the masterminds behind Bubba Watson’s hovercraft golf cart).
Still, it’s fascinating to see how much faith BuzzFeed’s putting into decoding that viral DNA. They may not be able to guarantee a hit post, but they have apparently figured out how spending a little cash can put a story on the right track.