Sept. 7, 2011
Study suggests targeting talkative social media users for more effective marketing
More friends may not be better for trying to generate marketing buzz using social media.
A new study by researchers at the University of Iowa's Tippie College of Business finds that social media users with small numbers of friends might be more influential with their online network of acquaintances than those with larger numbers of friends.
As a result, businesses might be better off picking users with small online networks to sell their products or services, or to create marketing buzz for their business. The study found that someone who has a smaller number of online acquaintances but communicates frequently with them is a better target than someone who has more online friends but doesn't communicate as often.
"People who email each other more frequently are closer, more likely to influence each other, and have influence over others," said Gary Russell, a professor of marketing in the Tippie College of Business. "We found that a person who has a large network but doesn't buy anything isn't important from a marketing perspective. A person at the fringe of a network but who buys a lot is more influential than someone at the center but who doesn't buy a lot."
The use of social media as a sales and marketing tool is still in its infancy, but analysts expect sales to grow quickly. For instance, a report by consulting firm Booz & Co. earlier this year predicted that social commerce sales in the United States will jump from $1 billion this year to $14 billion by 2015.
But that won't happen until research speeds the development of a social media business model. Russell and Tippie doctoral student Sang Uk Jung are trying to develop part of that model. In a recent study for Jung's doctoral thesis, they analyzed transactions between players in an online multi-player game called Kilride. The game is a "huge phenomenon" in Asia, Jung said, especially in Korea and China, where online gaming is much more popular than in the United States. Jung, who used to work for the company that owns Kilride before going to graduate school, said some Internet cafes in Korea have nothing but the game because it's so popular.
Online games have never been used before for research purposes, but Jung said Kilride makes for a great tool because of its number of users and because it so thoroughly replicates everyday activities—people can make friends, create social groups, buy items with real and virtual money, just as they would in the real world. Since all of those activities are recorded, Jung said it provides a rich resource database to study transactions between people.
Russell and Jung looked at one month's worth of buying behavior to see which people had the most influence among their gaming network. They noticed that the most important factor seemed to be frequency of communication with other members of the network, not the size of the network.
Using the information gathered during that one month window, Russell and Jung predicted future buying behavior. When they looked at the purchasing records from later months, their predictions were largely true.
The researchers said the study suggests people online tend to self-segregate, just as they do in the real world.
"People who are nearer to me are more likely to be like me, at least in their transactions," Jung said.
Jung said he's planning another research project that will expand the study beyond a one month window to more thoroughly test their results.
Tippie marketing professor Qin Zhang is also overseeing Jung's thesis. Jung presented his findings at the 2011 INFORMS Marketing Science Conference. His paper is entitled "Identifying High Value Customers in a Network: Individual Characteristics vs. Social Influence."
STORY SOURCE: University of Iowa News Service, 300 Plaza Centre One, Iowa City, Iowa 52242-2500
MEDIA CONTACTS: Gary Russell, professor of marketing, 319-335-0993, email@example.com; Sang Uk Jung, doctoral student, firstname.lastname@example.org; Tom Snee, 319-384-0010 (office), 319-541-8434 (cell), email@example.com