Social data interpretation: The human factor

Research claim that A full 90% of all the data in the world has been generated over the last two years. This tsunami of digital data have brought along incredible insights but also many many headaches. One of the most prominent challenges relates to the inferences we draw from these data. Kate Crawford argues in “The Hidden Biases in Big Data” on Harvard Business Review that “We give numbers their voice, draw inferences from them, and define their meaning through our interpretations. Hidden biases in both the collection and analysis stages present considerable risks, and are as important to the big-data equation as the numbers themselves.” O’Reilly Radar’s Mike Loukides stresses, in “Data Skepticism“, that “even when you have unlimited data, you have to be very careful about the conclusions you draw from that data. It is in conflict with the all-too-common idea that, if you have lots and lots of data, correlation is as good as causation.”

This particularly strikes a nerve when it comes to Social data (and Social Network Analysis). You must have, at least once, come across titles such as “Social Networks are making us anti-social” or “Facebook causes divorce” or “Twitter moods help predict stock markets” etc. This might reflect a mere misinterpretation of the original studies (causation sells way better than correlation as the comic promptly illustrates) or a defect in the analysis and interpretation processes of said studies. We have a tendency to jump into such conclusions because our minds react better to narratives and “because” is a good ideas’ connector.

When examining the virality of content in social networks, a tweet for instance, the observed contagion phenomena is often explained through ‘Peer influence’. It could be the case, but it’s good to stop and think ‘Maybe the answer isn’t that simple’, maybe there are alternative explanations. Sinan Arial, an associate professor at the MIT Sloan School of Management, delivers a compelling talk about social contagion and highlights just how Homophily (the tendency of similar people to bond together) is a viable explanation for some diffusion phenomena often attributed to peer-influence.

When stressing how ‘Big Data’ (Social Data) will revolutionize business or how ‘visualization will save big data’, vendors fail to stress these interpretation issues (which is understandable). Human intervention is an omnipresent part of the conception and analysis process. And unless our analysts (or data scientists if you like) are open minded enough to consider alternative explanations, or we come to find more appropriate models, we might just be digging ourselves into a much bigger hole.


Links & Notes

Here are few links that captured my attention this week.

  • If a network is broken, break it more: New research from Northwestern University suggests that, instead of replacing the damaged lines in a network, we could restore the whole network by strategically disconnecting even more lines
  • Social Network Analysis: making invisible work visible: The paradox is that organizations continue to allocate a significant proportion of their IT budgets on communications infrastructure and ‘social software’ and virtually nothing on systems and tools that can analyze how effective this investment is.
  • Unstoppable march of big data: The driving force of big data is not technology, but the economics of data storage. Cell phones, lifts, doors and weather stations always used to throw out tremendous amounts of data every few seconds; we just never used to record it. Now we can.


  • Interesting app I spotted this week called Coffitivity that supposedly offers just enough noise to get work done
  • A great compilation of resources for obtaining, handling and visualizing data
  • A video, inspired from Sherry Turkle’s book Alone Together, “the innovation of loneliness” is worth pondering as you unplug this weekend!

“Network theory” vs “Theory of networks”

It should also be noted that SNA theorizing encompasses two (analytically) distinct domains, which we refer to as “network theory” proper and “theory of net-works.” Network theory refers to the mechanisms and processes that interact with network structures to yield certain outcomes for individuals and groups. In the terminology of Brass (2002), network theory is about the consequences of network variables, such as having many ties or being centrally located. In contrast, theory of networks refers to the processes that determine why networks have the structures they do—the antecedents of network properties, in Brass’s terms. This includes models of who forms what kind of tie with whom, who becomes central, and what characteristics (e.g., centralization or small-worldness) the network as a whole will have.

Source: On Network Theory – Borgatti & Halgin

The Challenge of Network Analysis

“To represent an Emprical phenomenon as a network is a theoretical act. It commits one to the assumptions about what is interacting, the nature of that interaction and the time scale on which that interaction takes place. Such assumptions are not “free” and indeed they can be wrong”

~ Carter T Butts – Revisiting the Foundations of network analysis

How Networks Can Revolutionise the World

The world is one big network problem. Some even affirm that we can only survive the future through connections. One thing is sure, Network theory is becoming more and more profoundly important to understand the world. 

In a compelling RSA talk, the economist and author Paul Ormerod makes the case of how “we have to see the world through the lense of networks in order to be able to design and think about better policies to achieving our ends whether in politics or the corporate sphere”. Through examples of network systems that range from popular culture to modern economy, Ormerod presents how observing network properties such as resilience or fragility can help explain some aspects of today’s world, including the economic crisis.   



Our greatest defect: crossing the chasm.

“Cros­sing the chasm bet­ween “Idea” and “Exe­cu­tion”. That is where the bodies pile up” ~Hugh macleod

Ok, maybe it’s not “The” greatest defect but it sure is one of the most significant and handicapping ones. I see this everyday and I’m sure you do too. Hundreds if not thousands of ideas that thrived throughout the ideation process but never survived the dreaded period of execution. Dozens of people who seem to have a sizzling motivation, a good vision, a cutting edge idea but never succeeded to concretize it in the real world. I fall for the same trap as well. More often than I like to. When I tried to understand why, I found out that our defect was in our little understanding of the two networks in play: the idea network and the execution network.

Network of ideas: Reaching outside the bubble

In the words of Thomas Alva Edison “Success is 10 percent inspiration and 90 percent perspiration.”  So in order to really succeed in making ideas happen you need to be inspired first. That’s what networks of ideas are here for: Causing the sparks that could lead to the ultimate breakthrough innovation. However, most of the great inventions in History emerged from a cross-pollination of ideas. Which means that only a diversified network could render innovative ideas. And since diversity comes from bridging the white spaces between disciplines, fields, cultures etc., this translates into the following rule: Leveraging weak ties and avoiding flocking with similar people is your passport to better ideas.

Mistake number 1: Cocooning in your bubble and waiting for the spark to come your way

Network of execution: It’s about collaboration      

There was a nice equation in belsky’s book, that I think summarizes the whole issue promptly. Making ideas happen = Ideas (we covered that) + Organization + Communal forces + Leadership capability. 

Organization is a rather intrinsic aspect. A skill that everyone needs to acquire in order to get things done. I really think it belongs to our circle of control, which means that if we are motivated enough to attain in, we certainly will.

Communal forces and leadership capabilities are the trickiest. In the network of execution, we solemnly work alone. Collaborating with others is often the stepping stone to cross the chasm between vision and reality. Collaboration leverages communal forces and needs great leadership to occur.  It often requires stronger ties that those of the network of ideas, and a common vision with all entities involved. If 10% is for inspiration, 90% of the effort needs to be devoted to building a strong community with a common goal. Needless to say, this is a daunting task. Most communities are fragmented chunks and subgroups that come together because of the energy a new exciting idea brings about. If not under the right leadership, the energy fades and the network doesn’t stand the test of times.

Mistake number 2: Deluding one’s self that the energy of the idea network will fuel the execution network forever

The network of ideas and the network of execution are very different and sometimes contradictory. While one is built on weak ties, the other needs strong foundations to thrive. While one can be transient the other requires medium to long term vision. While one’s texture is built on barely connected subsets, the other needs constant stitching to make a solid construct. Making ideas happen equals getting the best out of each network and leveraging both at the right time with the right people.

A touch of the not invented here syndrome?

Eugene Eric Kim did a great job mapping the different relationships among collaboration-related skills.  


What I discovered very early on was that there was an awful lot of great knowledge about how to collaborate effectively. The problem was that this knowledge was largely locked in silos. Ironically, the people who best understood collaboration were not collaborating with each other.


I was disappointed, but not surprised, that “collaboration” as a skill was mostly lumped with technology skills. Folks in the Enterprise 2.0 space, for example, have almost no overlap with organizational development professionals. It’s a troubling trend. Although people are fond of saying, “It’s not about technology, it’s about people,” there’s not much practice validating that mantra.


On the flip side, it’s disappointing that organizational development professionals have stayed removed from some of the amazing trends in the technology sector.

Social Business / Enterprise 2.0 are all about tearing down knowledge silos. Ironically, practitioners of these fields seem to be stuck in their own bubble. The future  lies in effectively spanning the white spaces.

Understanding Brokerage in Organizations


Source: A. Mrvar: Network Analysis using Pajek

I stumbled upon this article by two Harvard Business School researchers who were looking at Employee-suggestion systems from a different angle. While process improvement isn’t my field of expertise, this passage really caught my eyes:

Tucker also explains this finding in terms of “boundary spanning.” Nurses are at the far end of an internal supply chain. Even if they discover a gap between what the supply chain is providing and what the patient needs, they usually don’t have the authority or knowledge to go back to those supply departments and fix the problem; a higher-level person needs to be involved.
“This finding tells us that process improvement in hospitals will require people to work across departmental boundaries, where the problems happen, rather than within a particular department,” Tucker says.

Boundaries are often the equivalent of information flow and collaboration breakdowns in organizations. They arise for different reasons. The most common boundaries, as identified by Kate Ehrlich are:

  • Functional: Breakdowns between divisions (e.g., marketing and finance)
  • Geographic: Breakdowns between geographically separated locations (e.g., US and European offices, East Coast and West Coast offices)
  • Tenure: Breakdowns between long time employees and new employees
  • Organizational: Breakdowns because of M&A scenarios, or among leadership networks

So, if boundaries are obstacles to information flow for most networks, spanning or bridging them is a goal organizations seek to attain (as it’s the case of the hospitals in the HBS’s study). Employees who span these “Structural holes” and tie together otherwise disconnected people and information/knowledge entities, are referred to as brokers.  In order to understand the role of these employees, I’ll refer to Mrvar’s classification of the different types of brokers:

  • Coordinators are those who mediate between the members of the same group
  • When twomembers of a group use amediator fromoutside, this mediator is called an itinerant broker
  • representative is someone who regulates the flow of information and goods from his own group
  • Gatekeepers regulate the flow of information and goods to his own group
  • And finally liaisons are those that mediate between two groups while not belonging to either of them

Due to their unique position, brokers gain enough Social Capital to make them as much indispensable for the network as they are dangerous. Their bridging role translates into Control of the flow from one part of the network to another. They can thus be great change agents. Yet, the negative spin suggests that brokers can play a “Tertius” Strategy where they induce competition or conflict between neighbors who are not linked directly. This could render information retention problems, more conflicts and hence structural holes (which were supposed to be spanned in the first place).

For this reason, it is very important to identify and recognize your spanners early on. This is challenging because, as these employees sit in the white spaces between network pockets, they are not highly visible  and are frequently not in a position of formal authority. Cross and Thomas state that leaders can only recognize 30% of their key brokers which shows how much potential is lost because we are not leveraging network players as we should.         

Are you aware of the brokers around you? What are you doing to turn their power to the advantage of the organization?

It takes a network!


“It’s a profound thought….How every person is a new door, opening up into other worlds.” ~Six Degrees of Separation


Networks are everywhere. Whether we acknowledge it or not our networks shape us to some extent. Making the best out of our networks is an enriching experience that can help us thrive as individuals, communities and societies. 

I pondered on this as I was preparing a talk for TEDxENSEM that tackled the theme: Dare to be different. It seemed to me that to be different, one has to leverage that web of networks he’s often unconsciously embedded in. Acknowledging this fabric of interwoven ties helps us unleash their power. In the era of interconnectedness that we are witnessing today, being a great network player boils down to being able to strategically weave various networks. And that’s what we should strive for. 
The video of the talk is yet to be available, until then, here are the slides of the presentation.

The promise of Social Network Analysis


Photo courtesy of quinn.anya

Here is a fact: Organizations are a set of interwoven networks, embedded in bigger networks. They thrive or die according to their networks’ health. And while most organizations are aware of that, few ever act with a network-aware mind.

A social network approach is primarily concerned with the interconnections between [actors], rather than being focused on their attributes or behaviors. The patterning of such connections – the configuration of positions and relationships – constitutes the structure of a social network, from which the social behavior of individual members can be analyzed and interpreted. This structural arrangement has important implications for the [actors] involved as well as for the overall social network, insofar as it enhances or constrains their access and control abilities.

~Emergent Leadership in Virtual Collaboration Settings: A Social Network Analysis Approach. J. Sutanto, C. Tan, B. Battistini et al

Thinking organizations as networks relies on different lenses:

– A micro lens zooms on the employee and his ego-centric network
– A macro lens xrays the interactions between different subgroups of the organization (business units, project teams…)
– A holistic lens studies the organization taking into account its context (socio-economic context, partners, …)

Each lens requires different network measures and concepts. And each lens answers a different set of questions. Example: The HR department needs to know how the new recruits are doing after 6 months of hiring them. A viable approach would be to conduct an ego-centric network analysis on the recruits. The main objective is to identify the ties among the new recruits and other employees. If the recruits are still peripheral it’s time to take action to help them integrate. Launching an internal mentorship program for instance can help new recruits meet key collaborators that could help them advance their work and nurture a sense of belonging.

Thinking organizations as networks doesn’t necessarily come with extraordinarily out-of-the-box answers but it surely sheds the lights on problems from a different angle. The emergent body of research and application of Social Network Analysis has provided some important insights on how thinking with a network perspective can be associated with organizational benefits (better collaboration, enhanced innovation etc.). However, there always seems to be quite a chasm between academia and corporate business and many techniques developed by the research community still haven’t made it in the real-world yet. An interesting classification I came across the other day aims to cross this gap to some extent as it tries to map SNA techniques to business processes. The framework is based on the APQC Process Classification Framework and lists the various uses of social network analysis depending on the business process at hand (Operating or Management and support process).

Source: Social Network Analysis and Mining for Business Applications. F. Bonchi, C. Castillo, A. Gionis, and A Jaimes, Yahoo! Research Barcelona 

I have come to think of this framework as a good list of the promises Social Network Analysis makes. While it is true that many techniques stated above are still in their infancy and face numerous technical and cultural challenges, it is only a good thing to keep an eye on their progress. You may never know when the opportunity of applying them presents itself.
We will go into the details of these techniques and the challenges they face on our upcoming blog posts. Until then have a look at your business processes and see if any of these techniques would fit. We would love to hear your feedback!