November 2013Marketing

Social Influence — It's Weaker Than You May Think


For the past few years, marketing has been abuzz with so-called “Influentials” — who they are, how their social impact can make or break a new product or service, and, most importantly, how to get messages into Influentials’ brains and mouths. But Wharton marketing professor Christophe Van den Bulte says social influence is weaker than many people think, and he has the research to show it.

“For instance, there is now extensive data on physicians and their acceptance of new drugs. A popular theory is that some doctors are Influentials because they’re the ones others go to for advice and refer their patients to. There is no argument that some physicians are indeed opinion leaders. But the theory goes further to claim that Doctor A prescribes a new drug because she was influenced by Doctor B. That second claim is not always true. You could instead attribute the behavior to the fact that they have similar patients, similar training, or that they were both influenced by a salesperson at about the same time. What looks like social influence or word of mouth is often really similar needs, taste, or situation.”

In one study, Van den Bulte analyzed a famous data set many believed showed peer influence in the timing of physicians' first prescription of a drug. “However, once I controlled for advertising,” he explains, “the influence of peers disappeared.” In another study, he and Wharton colleague Raghuram Iyengar did find evidence of peer influence even after controlling for marketing effects, but only for a quarter of the physicians studied. So, he notes, “If only a small group is influenced by their peers, social marketing alone cannot be the solution to all of your problems. You cannot rely that heavily on peer influence.”

Van den Bulte, who teaches in Wharton’s Competitive Marketing Strategy, says that what often looks like influence can instead be explained by homophily. The term, which literally means “love of the same,” was coined by sociologists in the 1950s to describe people's tendency to interact more with people who are similar to them (think taste, needs, and budgets) — giving credence to the adage “birds of a feather flock together.”

“Now, that idea of homophily has huge potential for marketers — it means that you can leverage social networks even when you do not know who the Influentials are and even when there is no social influence at all,” Van den Bulte points out. One example is customer referral programs, where firms reward current customers to bring new ones.  

“We looked at referral programs at a retail bank and a search engine optimization company. We found that those acquired through referral programs are better customers — they stay longer and are more profitable [the margins for referred customers at the bank were 25 percent higher, which combined with the lower churn resulted in an ROI over six years of 40 percent]. And the nice thing about these kinds of programs is that they are simple and low tech. You don’t need new data, and you don’t need to go online. They have been around for a century and they work.”

But what explains the success of referral programs? “It’s most likely another example of similarity in taste and needs. Member-get-member programs work because current customers have friends and family who are similar to them. So, if you are good at delivering what your customers want (and you must be, otherwise they would not be your customers), the odds are that you will also deliver what their peers want.

“So, what is really going on in these customer referral programs is that you essentially outsource the targeting decision to your existing customers. This is much smarter marketing than simply leveraging social networks to create buzz and build credibility through word of mouth.

“For marketers, leveraging the ‘birds of a feather flock together’ phenomenon is a new and extremely valuable insight. It’s not just a nice story in theory. We have the numbers to back it up, and so do people at AT&T and Microsoft Research. I expect it will gain more traction in the next few years,” Van den Bulte notes.

“I also hope that sound theory, solid analytics, and evidence that low-tech network marketing programs can be quite effective will help marketers think about what connections really matter,” he adds. “Right now there is too much focus on online. My concern is that these networks are shallow; they represent weak ties. But marketers tend to look at them because that is the data they have access to. It is not necessarily the right data. What you want to leverage are strong ties, and you don’t need to find Influentials or go online to do that.”