Haas Newsroom


EmailThis

Marketing and Social Networking: when measuring influence, quality connections top quantity

 

April 22, 2009

 

 

Haas School of Business Media Relations:

 

Pamela Tom                        
(510) 642-2734                          
ptom@haas.berkeley.edu     

          

Ute Frey                        
(510) 642-0342                          
frey@haas.berkeley.edu 

 

Taylor has 483 friends on Facebook. Cameron has 832 connections on LinkedIn. And Paige boasts more than 1000 followers on her Twitter page.  Are these three online devotees more popular than the average social networker?  Possibly so, but marketing professor Zsolt Katona, Haas School of Business, University of California, Berkeley, and a team of researchers found that as the number of one’s online contacts increases, the average influential power of that networking individual decreases.


From a marketer’s perspective, better understanding of consumer influencers is invaluable in developing successful marketing campaigns and effective viral marketing strategies. Katona recognized how social networks have transformed word-of-mouth communication between consumers and sought to identify how the networking structure drives consumers’ decisions. He says many viral marketers target people with the highest number of connections, but this study challenges that approach.


This study suggests businesses should focus their marketing efforts on individuals with fewer online social connections, “These findings have important implications for firms that want to better understand how customers adopt their product or service … Our model may serve as a methodology to identify customers who play an important role during the diffusion of the new product or service,” says Katona.


Furthermore, Katona says the findings suggest that marketing products and services on social network sites without careful examination of the network structure may not be as effective as advertising on search engines. 
The paper, “Network Effects and Personal Influences: Diffusion of an Online Social Network, “does not study network growth or evolution, but focuses on the diffusion process of an existing network. Katona and co-authors Peter Pal Zubcsek, Ph.D. candidate at France’s INSEAD, and Miklos Sarvary, INSEAD marketing professor, also found online social networking produces a “clustering” effect that drives influence.


Being in a cluster means network friends actually know each other as opposed to having many friends who do not know each other. If the person's friends know each other the local network is denser; a so-called cluster is formed. “This finding is important as it shows that beyond sheer network size, strong communities are more relevant for word-of-mouth influence,” says Katona. Consequently, if one were trying to maintain his influence in the network, he would also have to maintain the same density of his relationships when the number of contacts grows.


Researchers studied a major social network in Europe from its debut in 2003 until 2006. The intent was to study the site during a period when it was neither advertised nor benefited from much media exposure. Consequently, membership growth was due entirely to word-of-mouth prompting.  At the time of the study, the site had no competitors.


The network hosted three million clients, each with an average of 50 friends and a maximum of 100. The study looked at the first 138,964 registered users: 53 percent female; 47 percent male; the mean age, 21 years-old.  The site’s service provides tools for people to send messages to friends, share pictures, and maintain a profile page. The network supports only mutually accepted friendships.


The study defines influence by how quickly a person gets another person to sign up as a new friend. Researchers analyzed how many members joined the same online social network over time. After 36 weeks, the social networking site had grown by seven times more members and 12 times more links than in the initial examination period.


The results indicate on average, the person with more friends has less influence. Katona says although the study could not the determine the intensity of communication between members, it is clear that members have limited time and therefore as the number of friends increase, one has less time for each friend.


At the same time, female friends proved more influential than their male counterparts. In comparing different age groups, older people were more prone to influence while younger individuals showed higher influence on the older group than the latter’s peers.  Although these demographic results may be specific to the certain site, it is interesting how groups that traditionally have lower social power become influencers in this context. Katona says while the study’s demographic information was limited to age, gender and city of residence, it would be interesting to examine how other variables influence social connections, such as the school a person attended.


This paper is currently under review by the Journal of Marketing Research.

 


###