Harvard Professor Finds That Innovative Ideas Spread Like The Flu; Here's How To Catch Them

Network science reveals just how innovation—much like the dreaded flu bug sweeping the nation—is contagious. Here's where you need to position yourself to catch one (and maybe avoid the other).

You can hear it in the way we speak: Songs are infectious, trends are contagious, videos go viral. We use disease to describe data. Information acts like illness. As it turns out, so does innovation.

While Google helps us understand the way the flu moves and a Facebook app can ferret out who might make you sick, what's missing from the sniffly conversation is that disease and ideas both travel along social networks—the real-life kind. In other words, there's a reason that the only thing that swept through your office faster than that stomach bug was the gossip about who gave it to who.

Enter network science, an emergent discipline drawing from sociology, medicine, and statistics. Harvard Medical School professor Nicholas Christakis is one of the most prominent figures in the young field—he's made foundational investigations into how products, sneezes, and behaviors are spread by networked contagion.

"Things don't just diffuse in human populations at random. They actually diffuse through networks," Christakis says, noting that we live our lives in networks—the same as hunter-gatherers. Today our networks are the web of relations you have between friends, coworkers, siblings, and relatives. Where you are in that web lends special properties.

The key is exposure. The more central you are to a network, the more connections you have to others, the more you are exposed to whatever's going around the network—whether it's Carly Rae Jepsen or H1N1. So if you want to hear the next great pop song, find out about the next great innovation, or catch the next great flu—the center is the spot for you. As Christakis notes, the people with the densest, most central connection form a "sensor group," a kind of canary in the mineshaft for what's going to be cool next—or give you a cold.

"Individuals located centrally within a network will be at both an increased risk for the acquisition of a pathogen," Christakis says tells Fast Company, "and an increased risk for the acquisition of novel information."

Of course, being aware of ideas doesn't necessarily mean anything—it's what you do with it. "While your exposure to new ideas or to new information is an attribute of where you are in the network, your tendency to adopt a new thing is an attribute of you," Christakis says. "It's a similar case with germs: being at the center increases your risk of exposure, but whether or not you get infected depends on the resilience of your immune system."

Why managers need to know their networks

"If you're trying to get people to work better," Christakis says, "it's not enough to think about individuals, you've got to think about how the group of individuals is connected or organized."

Harvard has licensed Christakis's ideas out commercially, including to Activate Networks, a consultancy that has built on his ideas and that he cofounded with medical statistician Larry Miller. Miller described to Fast Company how Activate evaluates companies and their customers through the prism of networks—with important conclusions for managers and employees alike.

Larry Miller

Miller and his team recently worked with a large engineering company, the kind with several thousand engineers at a given facility, the kind whose job was to innovate. The consultancy was able to map the networks—most of the teams were six to seven people—and superimpose the company's metrics of success, like patents or commercial entries, onto that survey.

"What we found was striking," Miller says. More than a single person's individual work, the quality of relationships that a team had was a crucial—especially having someone on the team highly connected to the rest of the network. The most successful teams were the most interwoven, because as Miller says, "If you're doing great work back in a closet somewhere, it doesn't matter."

In this case, too, information—here the process of innovation—acts as contagion. Successful teams were able to spread their ideas through the organization, gathering commentary, criticism, and broader support. Movement of information from person to person and across departments advanced the innovation, Miller says, which couldn't have happened if it stayed on an island.

The benefits of a broad network aren't limited to a team: They serve individuals as well. Activate Networks' research has found that high-performing people tend to have strong connections, both laterally across departments as well as up and down the hierarchical chain. When you have dense, rounded connections, both you and your company benefit, as ideas then spread throughout an organization beyond their home silo. Those ideas receive feedback or get shot down, and you'll look good for being the source of the ideas. And the pattern is circuitous—if you have relationships across the organization, you're more likely to receive ideas and best practices from the collective.

Maybe you could call that having an infectious personality.

Drake Baer covers leadership for Fast Company. Although he's a little scared of the flu, he'd like to be a part of your network. (Really.)

[Image: Flickr user Steven Depolo]

Add New Comment


  • This seems to be following Burt's paper on Structural Holes and Good Ideas - that information is clustered in groups and in between is a huge chasm of no-information land. If you're the person who manages to bridge over that chasm, then you're more likely to succeed.

  • Shafiq Khan

    I do not agree with the professor because innovations means adoption of a new concept & defining some old concept as incorrect but this does not spread like flu but faces opposition & resistance.Take the example of following
    Main-stream physicists are misleading the ignorant people as the very paradigm of physics they know is already shown to be fundamentally incorrect & baseless because the very space-time concept has been mathematically, theoretically & experimentally shown as fundamentally incorrect & baseless in the article " Experimental & Theoretical Evidences of Fallacy of Space-time Concept and Actual State of Existence of the Physical Universe" published by a peer-reviewed Indian Journal of Science & Technology available on www.indjst.org March 2012 Issue (first article). Thus Special Theory of Relativity, General Theory of Relativity, E=mc^2 & Big Bang Theory being correct is out of question. Now those who believe that these theories have an iota of truth should first attend to the open challenge which is at http://www.worldsci.org/php/in... and http://www.gsjournal.net/Scien... I would like to inform that Open challenge to Einstein's Theories of Relativity is now the 'The Science News' on http://www.facebook.com/TheSci... (January 23) and also on University of Dhaka at https://www.facebook.com/pages... Now main-stream physicists have no alternative but to face this open challenge or leave their jobs, if they have any conscience left with them.

  • Abdallah Al-Hakim

    Increasing your interactions and networks has always made sense to me. Every person represents a possible node to a new network and exploring different networks is very stimulating. This process has been enhanced via the social web. I have not tried the 'Activate networks' but other sites such as Engagio (http://engag.io) are building social interaction graphs based on social gestures. In Engagio's case it is focusing on the commenting activity which is independent of other social gestures such as retweets or Likes. This opens the door for some potentially interesting discoveries to be made. 

  • Steve Ardire

    Christakis work is more traditional SNA/ONA stuff that players like Activate Networks licensed for commercial apps which is good for some problems to address.

    However today there's more and more dynamic information like activity streams where relationships continuously change, which impacts the direction of information flow so the key value add is capturing the context of conversations and making predictions on the ever-changing dynamics happening within a network.

    This means factoring in the fluid nature of topics based on changing interests and how that translates into identifying the right key influencers at any given time.