Social Network Analysis: What to Map


SNA2
A lot of work has been done in the past couple of years in the field of Social Network Analysis (SNA) — the assessment of the real person-to-person links in and between organizations and groups, as distinct from the ones that you might expect to see from the ‘official’ organization charts. These analyses have already provided some interesting insights — identifying who in organizations are the ‘loners’, where the knowledge-sharing ‘bottlenecks’ and ‘roadblocks’ are, and the types of networks that people tend to form, which can be described (according to SNA Guru and my fellow KM specialist Valdis Krebs, from whose site the illustration above is taken) using attributes like degree centrality (the number of connections), betweenness centrality (the number of connections to the otherwise-disconnected), and closeness centrality (the shortness of the average path to others). This insight can then be applied to other network theories like Malcolm Gladwell’s Tipping Points: Which network nodes indicate the critical communication roles — mavens, connectors and salesmen? (Heather in the illustration above has what Valdis calls a high betweenness centrality, and is likely what Malcolm calls a connector, while Diane could be either a connector or a bottleneck, or both).

tipping_pointThe advantage of the data that is used to construct SNA maps like the one above is their objectivity: Generally, this data is highly quantitative, and is collected by polling e-mails or telephone calls or other evidence of the presence of links. Even the terms ‘strong’ and ‘weak’ ties are defined in ways that are objectively verifiable — they refer simply to the number of degrees of separation between the nodes. The popular expression ‘the strength of weak ties’ (the fact that most of the value in your network tends to come from people two or more degrees of separation away) is a bit of a misnomer: It might be more accurately (if unpoetically) called ‘the power of large numbers of more remote ties’. I have argued that because of the high degree of ‘signal loss’ both going out to and coming back from these more remote links, this power is heavily diluted.

The next step, then, is to explore this ‘signal loss’: What causes it, and why is it greater in some situations (e.g. when you try to get an e-mail through one filter or intermediary to a celebrity) than in others (e.g. in successful viral marketing, where information spreads like wildfire).

A new book entitled The Hidden Power of Social Networks by Rob Cross and Andrew Parker suggests that organizations would be wise to ‘surface’ and support these de facto networks, instead of ignoring them in favour of the formal hierarchical ones. Management can do this, the authors say, by:

  • Bridging strategically important disconnects between departments or organizations
  • Improving a networkís ability to sense and respond to opportunities
  • Aligning the organizational context to energize and support networks
  • Identifying overburdened employees and redistributing workloads
  • Eliminating information bottlenecks
  • Recognizing and supporting key ìconnectorsî
  • Pulling in peripheral people who represent untapped expertise

To do this effectively, they say your organization needs to look beyond the objective quantitative attributes of network links (the frequency of connections) and survey you and everyone else in the organization to get your assessment of each of the following 16 qualitative attributes for each person you’re linked to

  1. The perceived value of information received from each person you’re linked to
  2. The extent and sophistication of problem-solving and decision-making through collaboration with each person you’re linked to
  3. The extent and quality of ideas and innovations you get from each person you’re linked to
  4. The degree of awareness of your skills by each person you’re linked to
  5. The degree and quality of your access to each person you’re linked to
  6. The degree of engagement with you exhibited by each person you’re linked to
  7. Your perceived level of safety of communication with each person you’re linked to
  8. Your perceived quality of conversations with each person you’re linked to
  9. The extent to which you think your productivity is improved by each person you’re linked to
  10. Your perceived sense of the extent of power and influence of each person you’re linked to
  11. The degree to which you like each person you’re linked to
  12. Your perceived degree of closeness with each person you’re linked to
  13. The extent to which each person you’re linked to supports achievement of your career goals
  14. The extent to which each person you’re linked to supports achievement of your personal goals
  15. The extent to which each person you’re linked to energizes (or exhausts) you
  16. The extent to which you trust each person you’re linked to

What could you do with this information? Imagine you could recast the SNA map at the top of this article with a second map that showed the perceived quality of information transfer (each way) between you and others in your network, and a third map that showed the perceived degree of trust (each way) between you and others in your network? If the perceptions each way, or between the three maps, were markedly different, this could be startling and potentially very useful knowledge. Unlike the authors of The Hidden Power of Social Networks, however, I think in the hands of management it would be dangerous, disruptive, and perhaps even abusive. But suppose the perceptions of you by others (how ‘well’ you do in each of the 16 qualitative attributes above) was available exclusively and privately to you? This, I believe, could be astonishingly valuable as a self-assessment tool.

Recall that Peter Drucker, who is consistently rated as the most universally trusted and respected business adviser of this generation, has argued:

  1. that it is presumptuous of managers to tell knowledge workers what to do when today almost every employee knows how to do their job better than their boss does (in fact it is likely their boss has never done this job and is almost completely ignorant of how it should be done); and
  2. that the greatest management challenge of this century will be to achieve significant improvements in the effectiveness (productivity) of knowledge work

What better way to communicate how well each employee is doing at their job than by giving them a map that shows how well they are perceived to be supporting their networks? Suppose for example you got this map, and it showed that your peers rated you extremely highly, but your boss rated you low. Imagine what eye-opening conversations could result: Does this mean your boss is an idiot, or does it mean you’re doing all this work helping others do their jobs at the perceived cost of getting your own essential work done? And if management perceives it is the latter, are they right or are you just under-appreciated and suboptimally employed, and should you in fact have your job description changed so that helping those people who rate your help so highly becomes your job?

As long as management is kept out of this evaluation process, I think this would allow each of us to self-manage our jobs and careers, and hence to become dramatically more effective in our work. It could tell us that we’re doing the wrong job. It could tell us that we’re working for the wrong organization. It could tell us why we’re not getting ahead, why we’re not happy, and what we need to do to turn ourselves around, to fine tune our performance, and to leverage our perceived strengths in the organization.

It could employ The Wisdom of Crowds to solve Drucker’s ‘greatest management challenge of this century’. But ironically, and delightfully, it could solve it with no involvement from management at all.

This entry was posted in Using Weblogs and Technology, Working Smarter. Bookmark the permalink.

4 Responses to Social Network Analysis: What to Map

  1. Valdis says:

    Great summary and insights Dave! Thanks for the links. Here is a direct application of SNA and Gladwell’s Mavens and Connectors …and it is all about you bloggers! ;-)http://www.orgnet.com/NEOBlogosphere2005.pdfAlso, I have lot more comments on your post, would you be interested in doing a podcast with me on this? It is too much to type… especially in a small box like this!Valdis

  2. Dave Pollard says:

    Thanks, Valdis. Sure, I’m up for a podcast.

  3. Dan Bassill says:

    I have a database of more than 13,000 contacts who I have been connection for more than 12 years in an effort to create a community of people who’s collective efforts results in more youth born in poverty entering jobs and careers by age 25.I view network analysis tools as a way to help people understand who the people in this network are, how they are connected, and how we have been working with them. At the same time, it’s an opportunity to look for voids in the network, and for others to find a role.Thus, I’m looking for interns / volunteers who would apply their knowledge of SNA to help me map this network. I think it has huge potential. Thanks for posting this information.

  4. Noel Guinane says:

    There is no substitute for hands-on management from a management team. It’s also unlikely SNA information would be made available to any one outside that management team.Apart from mapping a gossip mill – who talks to who and who likes who – what is the purpose of SNA and how exactly will it achieve it?

Comments are closed.