Sunday, June 16, 2013

Book Review -- Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives




It is always a pleasure when our expectations are unexpectedly exceeded.  When my wife and I were married, we were late getting to the inn where we were spending our honeymoon.  As the restaurant was due to close, I called ahead to ask if they could leave us some sandwiches.  We arrived to find they kept the restaurant open, had a table waiting by the fireplace, with a live classical guitarist.  They could have served sautéed shoe leather, and it would have tasted great.

Although not quite on par with my honeymoon dinner, I had a similar experience reading Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives by Christakis and Fowler (click here for link).  I expected to learn about the various forms of social media that have become such an important part of our communication landscape.  What I actually acquired was a new lens through which to view the world, and a desire to learn more about how this enhanced worldview can be used to drive positive change in communities and organizations.

I took three broad themes away from this well written and referenced work:

  1. Social networks have properties, which we tend to not understand.
  2. Social networks have complexity, which we tend to underestimate. 
  3. Deeper understanding of social networks has the potential to enhance spread of knowledge and behaviors (good and bad).


Upon completion of the book, the authors clearly succeeded in driving home the first two themes – I have a better understanding of network properties and structure, and a greater appreciation of the variability and complexity of network structures.  However, it does not deliver quite as strongly on the third theme.  How can one use this knowledge to enhance communication and engagement of a community?  It is almost as if I had learned for the first time of the existence of DNA, had an intuitive understanding that it should be clinically important, but had no idea how to translate this knowledge into meaningful activity.

Regarding the first point – networks have properties – the authors introduce the uninitiated to an entirely new network descriptive language.  A network, that is, a collection of people plus the specific set and patterns of connections between them, must have something to spread, or a contagion.  The contagion can be an infectious disease, information, happiness, depression, obesity or wealth.  Members of a network vary in terms of their location (central versus peripheral) and connectivity (high or low transitivity).  The nature of the network impacts the pattern and efficiency of spread of the contagion.

The second point – networks have complexity – is equally well covered.  Imagine you are starting a new organization and you outline your communication plan to your board or investors.  As the leader and chief influencer, when you need to engage your organization you plan to talk to one person.  That person will be assigned to talk to one other person, who will talk to one other person, etc., etc.  This is essentially a “bucket brigade” network structure, where the bucket of water, or in this case information, is transmitted from one person to the next sequentially.  Clearly, your board is unlikely to be impressed by this plan.  Although simple, it is full of obvious weaknesses.  It is vulnerable to disruption, since if one person does not carry out his assigned task, the flow stops.  It is also likely the message will change slightly as it is carried forward, and is unlikely to survive in its intended form to the end of the chain.


This is a ridiculous example, which no one would seriously propose.  However, what we generally put in its place is a corporate organizational chart.  As chief influencer, now instead of talking to just one person, you talk to your five direct reports.  Each of them talk to their direct reports, who talk to their reports, etc., etc.  This has the advantage of increased efficiency compared to the bucket brigade, but the improvement is still marginal.  The message is still subject to at least partial transmission failure if one person in the chain fails to complete the task, and the message is still subject to change.  As nonsensical as the bucket brigade model seems, the slightly improved branching tree model is probably the most common organizational structure for businesses, hospitals and physician groups.

In fact, the real social networks, which underlie the “official” organizational chart, tend to be much more complex, and continually evolve in reaction to the environment.  There tend to be hubs of hyper-connected influencers whose behavior has a disproportionate impact on contagion flow.  Instead of laboring over the production of the perfect “org chart”, with the expectation that the ideal design will drive great communication and engagement, perhaps we should concentrate more on identifying those nodes of influence within a group and working hard to engage them.

Understanding the architecture and laws governing networks has the potential for great good.  If we can learn to understand and leverage network dynamics, perhaps weight loss, exercise, anti-smoking and anti-drug campaigns could focus on far fewer individuals with equal or greater results to broader population interventions.  Perhaps the historically slow spread of medical discovery to routine and widespread adoption into medical practice could be accelerated.   Connected is an engaging and thought provoking work, and well worth the time invested for anyone interested in more effective group dynamics in any sort of group -- doctors, nurses, students, patients, employees, etc.