Whether you spread an idea, a product, or a service, you always have a choice. You can broadcast or you can connect. Broadcasting involves massive mailings or buying media time and packaging your message so that it can be transmitted simultaneously to all nodes in the network. Connecting involves starting a dialogue with certain individuals in the network that you are trying to influence. This chapter describes the latter: how to identify network hubs, how to connect, and how to get them to help you create change. The following case study describes how the concept was used in the medical field.

In the mid-1980s, Canadian physicians were abuzz about the use of cesarean sections. It was clear that the surgical procedure was often used when it wasn't needed, causing unnecessary risks and expenses. The Society of Obstetricians and Gynecologists of Canada issued guidelines that unambiguously recommended a lower rate of cesarean births. The guidelines stated that the fact that a woman had had the procedure before was not a reason to reach for the scalpel again. Most practitioners—94 percent of them to be exact—said that they knew of these guidelines, and 56 percent reported that they had discussed them with a colleague. But when researchers checked the monthly discharge data from hospitals to see whether there was a reduction in the number of cesarean sections as a result of the new guidelines, the answer was no.

It's not that the physicians disagreed with the guidelines—about 85 percent of practitioners said they agreed with them. In fact, many physicians claimed that they had changed their practices. "I think they truly were under the impression that they had changed their behavior," says Jonathan Lomas, a researcher who studied the case. "But in fact they hadn't."

Lomas and his colleagues were intrigued and decided to search for the most effective way to influence these physicians. They tried several methods and eventually found that the best way to change people's behavior in this case was to work with local opinion leaders.

The first step was to identify the network hubs. To do this, the researchers sent out questionnaires to physicians in each hospital asking them to nominate colleagues in the area who matched the "opinion leader" profile. When the opinion leaders were identified, they were invited to participate in a workshop that explained the rationale behind the guidelines. In the weeks that followed the workshop, the researchers asked the physicians to do a little "marketing." They asked each physician to mail a personal letter to his or her colleagues with some material from the workshop, to follow up with another letter, to host a meeting with an expert on the topic, and, in general, to talk with colleagues about the issue. The results were impressive: Seventy four percent of the physicians in the hospitals with trained network hubs gave women the option of choosing regular labor, compared to 51 percent and 56 percent in groups where different strategies were tested.

Also in this chapter: • Where to Find Network Hubs? • Give Them Something to Talk About • The Bias Toward Mega-Hubs • Why Reach Hubs Early?