Referrals from friends have a significant impact on typical buying behavior, an effect commonly known as social influence. Social influence is part of the reason so many marketers, as well as investors, get excited about the marketing potential within social-media networks. However exciting the potential, marketers and investors still need to understand the relationship between social influence and online sales. What effect, if any, do friends have on one another's buying decisions when engaging on a social network?
A study by Harvard Business School titled "Do Friends Influence Purchases in a Social Network?" addressed this question by using data gathered from a Korean social-media website called Cyworld. Cyworld contained a personal page format similar to Facebook and also allowed users to purchase virtual items to display on their pages through e-commerce. The study addressed the relationship between social influence and buying decisions between 208 randomly selected members in the study group. It also included the real impact on sales.
Do friends influence purchases on a social network?
The findings identified within the data set three main groups of users, which I refer to as the Normals, the Joneses, and the Hipsters. These findings are consistent with the middle-status conformity thesis, a sociological phenomenon that segments individuals into three groups on the basis of status within the community.
These individuals represented 48 percent of the study group. These users had the least amount of activity on Cyworld. They visited other users' pages the least (low outdegree) and also had the least amount of other users visiting their pages (low indegree). The Normals showed zero social influence on buying decisions, resulting in a zero impact on sales.
These individuals represented 40 percent of the study group. They showed a moderate level of activity on Cyworld in terms of time on site, indegree, and outdegree. The Joneses showed a "keeping up with the Joneses" behavior by purchasing items that friends within their network purchased in the past. The study found a positive social effect on the Joneses' buying behavior that resulted in a 5 percent increase in sales for this group.
These individuals represented 12 percent of the study group. These users showed the highest levels of activity on Cyworld in terms of time on site, indegree, and outdegree. The Hipsters showed a 14 percent negative effect on buying behavior as a result of social influence. The study postulates that, "To maintain distinctiveness, these users tend to reduce their purchases of items when they see their friends buying them."
Given the findings from this paper, here are a few suggestions for companies that wish to boost e-commerce sales using social influence:
1. The Normals are irrelevant. Omit these users from your ad spends.
2. The Joneses are the most important group, given that they show the highest conformity and make the most purchases. Find the Joneses that interact with your industry on social networks and target them with display advertisements.
3. The Hipsters are irrelevant in terms of direct sales but can be highly instrumental in driving buying activity from the Joneses.
Find the Hipsters in networks that don't know of your product. The Hipsters will have a higher probability of sharing your product if it's unknown to their network, thus preserving their unique distinction. Sharing product samples is also a great way of boosting the Hipsters' personal statuses, which drive positive downstream effects for your campaign as a result of their sharing your product with the community. Try to coordinate the Hipsters' mentions of your product with a display ad campaign targeting the Joneses. You will get more for your money, because referrals combined with ads show positive interdependence in online marketing.
Please let me know if you implemented these strategies. What were the results? I would love to do a follow-up post with more real-world examples!