The idea that it’s possible to reach massive Internet audiences has long tempted media makers of all stripes. While amateurs depend on serendipity (and cute cats), marketers are pursuing more sophisticated strategies.

Last Thursday, Wired Business editor Ryan Tate published a blog post describing ad agency Portal A’s strategy for creating viral marketing hits. Their team hopes their latest video, a humorous piece promoting a video chat service, can capture the attention of online audiences. According to Tate, “Portal A, more than any creative shop, has cracked the code of viral video.”

For anyone studying the influence of entertainment, the idea that there’s a secret to engineering viral content is alluring. According to this notion, once we discover precisely what makes a viral video succesful we’ll have the tools to maximize the reach of our message, to add an “X factor” as Portal A co-founder Nate Houghteling puts it.

As with many alluring claims, deeper investigation complicates things. Our recent talk with computational social scientist Sharad Goel and study of Bully’s Change.org petition have added empirical and theoretical nuance to commonly held views about the viral spread of information online.

The fact that this video hasn’t behaved like a virus shouldn’t surprise us. Most information shared online doesn’t.

The first step is to clarify what we mean when we say “viral.” During his talk, Goel suggested that many people confuse the word “viral” with “popular.” The more complex but precise analogy posits that the spread of information online occurs in the same pattern as the spread of infectious diseases. The viral model implies that information moves away from its original source across many steps as more and more users adopt a given idea or product.

To his credit, Tate invokes the biological roots of the metaphor when he wonders if Portal A’s video “‘will spread like a weed’ across social networks.” However, YouTube data shows that the video, while popular, has failed to generate the same level of buzz as past sensations. With almost 180,000 views in the first week since its posting, the video has certainly made an impact, although at a smaller scale than the biggest diffusion events that capture millions of viewers.

The fact that this video hasn’t behaved like a virus shouldn’t surprise us. Most information shared online doesn’t. Goel and his colleaegues found that the vast majority of information shared on Twitter doesn’t spread at all, and information that is shared rarely behaves like an infectious disease. Rather, most large-scale online events remain close to their source. Creating content that yields these types of hits may produce some successful videos, but it won’t change the underlying structure of a network in which most content fails to gain traction.

Our free lunch may never arrive, but Tate includes some details in his article that might provide a better account for the video’s success. Instead of sitting back and watching the video spread like a disease, there’s evidence that Portal A and their clients understand, at least implicitly, that a more proactive approach based on traditional media exposure can play a big role in generating online views. Tate notes that the video has “won coverage” in a variety of influential online publications such as “TechCrunch, All Things D, Ad Age, Mashable, Ad Week, and Business Insider.” Agency Portal A was the subject of a big profile in an even more traditional medium, the print edition of the San Francisco Chronicle.

In the meantime, we’ll keep our cameras focused on cats and hope for the best.

We suspect that these facts aren’t lost on the producers who create these videos for a living. Kai Hasson, another Portal A co-founder, has a realistic view and frankly admits that “you can’t really predict” which videos will go viral. While accurate prediction remains difficult, studying the structure and dynamics of online social networks can add a dose of healthy skepticism to claims about viral content.

Rather than “winning” coverage because of any inherent “viral” properties, empirical work like our own Bully study suggests that it’s mainstream media coverage that propels popular content. Goel observes that the structure of these large information cascades follows a shallow “broadcast model” rather than long term peer-to-peer spread. A likely explanation for Portal-A’s success is coverage in popular media outlets or endorsement by people who can harness online attention at a large scale, like Wired’s Tate.

In the meantime, we’ll keep our cameras focused on cats and hope for the best.

Image: Portal A, Co.CREATE