Understanding the ways the social graph changes and influences decision making is something marketers and scientists are frantically trying to figure out. According to a new study out of MIT, people are more likely to start new behaviors when they're recommended by small clusters of people they know well. This may not sound intensely groundbreaking, but it runs counter to the decades-long assumptions social sciencists have been making. 

Apparently, social scientists believe social network behaviors are most affected by many distant connections, or "long ties." But now it seems that dense clusters of connections makes all the difference in influencing consumers. 

According to Damon Centola, an assistant professor at the MIT Sloan School of Management, who led the new study:

"For about 35 years, wisdom in the social sciences has been that the more long ties there are in a network, the faster a thing will spread. It’s starting to see that this is not always the case.”

Centola ran a series of experiments in an internet-based health community he developed over a two period. Users could rate and share information with friends. The 1,528 participants all had anonymous profiles and differing health interests. They were split into two groups. One connected through long ties and one organized around dense communities, where Centola matched participants with other users who sahred the same interests, called "health buddies." Six separate trials were run to discern which group was more likely to register for an online health forum website.

As it turns out, 54% people in clustered networks registered for the health forum, compared with 38% in the networks oriented around longer ties.

Moreover, creating a buddy system was very fruitful. Participants with more health buddies were also more likely to become regular participants in the health forum. Only 15% of forum participants with one friend in the forum returned to it, but more than 30% of subjects with two friends returned to it, and over 40% with three friends in the forum made repeat visits. 

According to Centola:

"The networks that make information spread more quickly actually make behavior spread more slowly."

While the study focused on health behavior, it's not hard to see how it coudl translate to other areas. Interestingly, Centola seems to define close connections as people who have similar interests, rather than actual real life connections. That's good news for marketers, who are interested in targeting specific demographics. If they can foster communications between people who share similar hobbies and purchasing behaviors, they may have more success increasing sales and other behaviors.