That ideas can go viral is now a given in corporate marketing and the culture – it’s even part of the plot of the 2011 Pulitzer Prize-winning novel “A Visit from the Goon Squad.” But new research suggests the term “viral” marketing does not describe accurately what happens in the market.
“It is not happening at all,” says Sharad Goel, senior researcher at Microsoft Research in New York.
I saw his research into virality mentioned by one of Goel’s colleagues at the Marketing Science Institute Date conference in December. It turns out the paper, “The Structure of Online Diffusion Networks,” by Goel, Duncan J. Watts and Daniel G. Goldstein (all of Yahoo! Research at the time, now at Microsoft Research) was first presented last January, and is publicly available Intrigued by its counterintuitive finding, I got in touch with Goel.
Goel and his colleagues studied seven different online scenarios to see how they spread:
2 Zync, a Yahoo! Instant Messenger video-sharing application;
3 Friend Sense, a Facebook app introduced in 2009;
4 “The Secretary Game,” the online version of a classic hiring test devised by psychologists;
5 Yahoo! Kindness, a charitable website launched in 2010;
6 News stories sent via Twitter in November 2011;
7 Youtube links diffused through Twitter in November 2011.
Goel says they wanted to see whether these spread virally, “like the common cold, some sort of biological contagion. One person gets infected and their friend gets infected and a friend of their friend gets infected.” It’s called multistep diffusion in technical terms. And Goel’s data shows it’s a fantasy.“What we see is something qualitatively different. Most of the time it adopts and dies out within one generation,” Goel says.
Try 99 percent of the time, in fact. The best any of these seven online scenarios did was to get passed along more than once in six percent of cases. That happened both for Yahoo Voice and Friend Sense. Goel isn’t using models, by the way; this is real world data.
For marketers, Goel’s research means it’s time to abandon the idea that viral marketing via social media will lead to tenfold organic growth.
How about 20 percent growth?
But Goel has some good news for marketers: while things don’t go viral like the flu, they can get a 20 percent return – for every 10 adoptees of a conventional marketing effort, another two people will adopt something organically.
“Your business plan better not depend on 100x growth through word of mouth,” Goel says. “But 20 percent is nothing to scoff at. And you can put that into your business plan.”
Goel says other research shows that optimizing for content diffusion using things like adding a ‘share’ link can bump returns up to 30 percent or even 40 percent.
Marketing Gangnam Style
So, what about Gangnam-Style events, where something does become incredibly popular seemingly out of nowhere? Goel says the initial paper only looked at four outliers, three of which became popular through a conventional broadcast model, and the fourth was closer to a broadcast model than a viral one. (An aside: the broadcast model does map to a kind of virus — water-borne viruses like polio, where one person contaminates the water supply and everybody else gets infected from that one source. But it’s not the person-to-person model that is the conventional shorthand for viral marketing.)
Goel says “We know things become really popular, like Gangnam Style. On the other hand we know that almost nothing becomes popular and all of these things seem to die out within one generation. So what is going on with these really popular events?”
The answer is under peer review. Goel has submitted a follow-up paper examining what drives popular things online, and it is currently under peer review. He says he expects to post the paper in March.For anyone looking to spread an idea, product or piece of content it should be compulsory reading.
In the meantime, you can watch Goel talk about his work on Youtube (Excerpt from Sloan,MIT,fair use policy)