PJ Lamberson, Visiting Assistant Professor of System Dynamics
CAMBRIDGE, Mass., March 16, 2010 — Few topics are of more interest to businesses, advocacy groups, and other organizations trying to win public favor these days than networks.
Facebook pages and Twitter feeds are ubiquitous. Marketers of products and ideas strategize on ways to expand their connections and harness the power of social networks.
But simply adding network connections may not be the best way to acquire adherents, according to MIT Sloan School Visiting Assistant Professor PJ Lamberson. Under certain conditions, more connections may actually hinder adoption of a product or technology or idea—even if it is something that people will enjoy and appreciate, Lamberson has found in his research.
"In some cases, even though on average people would be better off using the technology or product, adding more connections can decrease how much it spreads," says Lamberson. "It's kind of surprising. We would have thought that adding connections would always help something good spread."
While Lamberson's findings are of obvious interest to marketers, they also can be used by individuals or groups trying to influence public attitudes and behaviors, such as support for health care reform or participation in recycling.
Lamberson reached his conclusions by building a mathematical model that simulates how people behave when they consider whether to adopt a new technology, such as an iPad or TiVo. In his model, individuals' decisions are shaped by both their prior beliefs about the technology and the information they receive from friends, family, and others they know who have already tried the technology.
"Based on those inputs and how connected people are, I'm able to predict how many people will end up adopting a technology," he says.
Some of the outcomes predicted by the model are not surprising. If a technology is not very good, then adding connections to a network decreases adoption rates. And once a technology has passed an initial threshold and proven effective, then more connections give rise to still greater adoption until most people are using it.
But problems can arise if people have high expectations for the effectiveness of a technology. Even though the technology may be beneficial, if adopter’s experiences are not sufficiently positive, then adding more connections can decrease adoption, Lamberson found.
In some cases, a technology can become stuck, and an external stimulus may be needed to reach a tipping point where people will start adopting it again, according to Lamberson. In the case of a product, a deep discount or free sample might accomplish this. For a behavior like home energy conservation, a temporary government rebate for home insulation could set in motion continued adoption of conservation measures.
Lamberson's findings are presented in the forthcoming paper, "Social Learning in Social Networks," to be published in the B.E. Journal of Theoretical Economics.
A specialist in systems theory, Lamberson designed his model to account for complex interactions within networks. His approach is a departure from standard network simulations, which assume that products spread through networks in the same way as infectious diseases.
Lamberson's model does not assume that people are perfectly rational and self-interested, which are basic assumptions in classical economics. "People don't use these perfect decision rules," he says. "They use approximations, which makes decision-making easier for them."