Most of us have heard the term “going viral” when talking about something that has quickly spread from one person to another. What most of us might not realize is that the terminology is rooted in a longer history of contagion modeling used to study mass outbreaks of contagious diseases called pandemics.
A new model by researchers at MIT looked at the spreading of a contagion in the early stages to determine which airports would have the greatest impact on spreading. Rather than create a random traffic pattern for travelers, this report differed in that it used cellphone data as a basis for individual travel patterns.
The conclusion was that simply being a busy airport didn’t automatically mean that there was a significant role in terms of spreading the contagion. Atlanta’s Hartsfield-Jackson Airport, number one in terms of number of flights, ended up being 8th in terms of contagion influence. Logan International Airport in Boston was 15th.
The top contagion-spreading airports were Kennedy Airport (#18 in world airport traffic), LAX (#6), Honolulu, San Francisco (#24), Newark, O’Hare in Chicago (#4) and Dulles in Washington.
One of the interesting results is that Honolulu, which only has about a third the traffic of Kennedy Airport in New York, could be nearly as influential in terms of contagion. The reason according to researchers is that it’s geographic location means that it is connecting to more, large and well-connected hubs giving it additional influence.
A potential lesson here for designing viral projects is to not think of just connecting to those who can generate the most traffic. Having 500 people all know about something is not the same as 500 people having equal influence in terms of how that message will be disseminated. On the face of it, this is probably not much different than typical common sense.
The insight we might take from the MIT study is that the key to effective viral projects is to look for people that have significant influence not only in their field, but have the ability to cross fields and retain the same level of credibility in other distinct groups. This means that the most significant influencers will not always be the ones with the most people in their network, but those with the ability to influence the largest number of distinct groups.