- Haas Home
- Haas Newsroom
- Cover Story
- Featured Stories
- In Brief
- Innovation Wizards
- Power of Ideas
- Connected at Haas
- Alumni News
- Personal View
- About CalBusiness
- Past Issues
Quality Tops Quantity in Social Networking
Prof. Zsolt Katona shows that influence declines as online contacts increase
Taylor has 483 Facebook friends. Cameron has 832 LinkedIn connections. And Paige boasts 1,000 Twitter followers. Are these three online devotees more popular than the average social networker? Not necessarily, according to Assistant Marketing Professor Zsolt Katona.
Katona and a team of researchers recently found that as the number of one’s online contacts increases, the average influential power of that individual decreases. Their results challenge the approach taken by many marketers who target people with the most online connections.
In their paper, “Network Effects and Personal Influences: Diffusion of an Online Social Network,” Katona and co-authors Miklos Sarvary, a marketing professor at INSEAD in France, and Peter Pal Zubcsek, an INSEAD PhD candidate, study influence in a major social network in Europe from its debut in 2003 until 2006.
Katona and his co-researchers analyzed how many members joined the same online social network over time, looking at the first 138,964 users. The intent was to study the site before it was advertised or received much media exposure. Consequently, membership grew entirely by word-of-mouth, with members telling friends about the site.
The study defined influence by how quickly a person got someone else to sign up as a new friend.
On average, the study found that a person with more friends has less influence. Katona says although the study could not determine the intensity of communication between members, members have limited time and therefore, as the number of friends increases, one has less time for each friend.
Of course, being able to identify the “influential” customers is a marketer’s dream. Katona found that influence is driven by a “clustering” effect in online social networks. He defined a cluster as a network of friends who actually know each other as opposed to a group who may have friends in common but do not know each other. The more friends know each other, the denser the cluster.
Katona found that the strength of relationships between a member's friends – the density of that member's cluster – correlates with the influential power of that member. In other words, influence is all about the relationship among one’s friends – not just the number of friends an individual has.
“This finding is important as it shows that beyond sheer network size, strong communities are more relevant for word-of-mouth influence,” says Katona, who noted a few implications for firms. “If a firm wants to introduce a new product or service, the best way to market it is to target influential customers. These customers might influence other’s preferences and tastes, and thus, by learning about them, the firm can design a better product.”