Using Networks to Find Knowledge

Last week Ralph Ohr left me with a challenge to think about how to use experts to get the best outcomes on making decisions under conditions of uncertainty. We constantly miss disruptive changes in the operating environment and I suppose if I really knew the answer, I wouldn’t be posting it on a blog.

Sometimes predictions are genuinely impossible because of true uncertainty. The future is the future and nothing in the past can help us predict some events. Rather than making predicitons, operational flexibility is probably the best response to this type of uncertaintly.

On the other hand, sometimes the emerging disruptions are right under our noses and the problem is getting over myopia. Experts can suffer from myopia as well as the rest of us so perhaps the issue is finding the right expert with the right interpretation of what is happening.

This means that an organization’s networks can be crucial in determining the successful search for the right person with the required knowledge. Tim and I have been doing some work for a large organization that is trialling new ways of doing business. It seems that one of the issues that is faced by this organization is building the expertise networks acrosss the business to find the right person to give their opinion and expertise at critical decision making points. However, in large organizations, this is difficult. A colleague from a large multinational mining company sums up the issue in the following diagram.

As you can see, effectively using the knowledge of the business means trying to get better connections to reduce the size of the “I don’t know who to ask” space.

So how can we do this? One possibility is that we direct our questions to people in the organization that we know are very highly connected. However, one simulaiton study of search in a real organizational network has found that this might result in more steps needed to find the right person. In this simulation, a slightly more efficient search could be conducted by going to the manager who is responsible for the subject area that is being investigated or by starting the search in the right department.

The simulation study needs more investigation and we have a PhD student looking at the problems of search in large organizations. However, if formal lines of enquiry can be shown to be associated with efficient search networks then this suggests that organizations may be able to connect pockets of knowledge by identifiying subject matter experts and linking them to critical projects.

Please note: I reserve the right to delete comments that are offensive or off-topic.

8 thoughts on “Using Networks to Find Knowledge

  1. Hi John,
    when it comes to knowledge, networks might be a tricky concept. How are links conceptually built and maintained? What is the relationship between experts, their knowledge, and the perceived value of their knowledge?
    In my opinion, the answer is “it’s all in the head”, and each member of an organization has a different view of the relationships he entertains with others, even if the “objective” network remains the same. Focusing on the *who* might alter the possibilities to reach the right person, since linking someone to a definite piece of knowledge might be made differently by different people. That’s probably why the results of your study say that going to the manager is more efficient.
    The “what” part of the question seems to me as important as well; different ways toask a question mightr trigger different reactions and different linking behaviors inside the network.

  2. Hi Thierry. There’s some really interesting research on the first topic. One US study shows that people tend to overestmiate the connectivity of their social network by a big margin. This probably means that the network maps that people carry in their heads are a bit unreliable. You’ve got a good point about content and context within networks. Being able to define the question is just as important as being able to locate the right answer.

  3. Hi John,
    Interesting points you make and nice read. I have had experiences with this area myself and have formed my own thoughts on this subject.

    One thing I have noticed is that people in general equate authority with knowledge – which is a false assumption to start out with. In many firms, some of the most knowledgeable are not the managers or the VPs but someone at the bottom of the totem pole. I think the key to finding expertise in an organization is like solving a puzzle – one has to find the nodes that link the most structural holes in the knowledge network.


  4. Thanks for picking up my comment for this post, John. I really apreciate it!

    I think you describe a valid approach in order to locate appropriate knowledge experts in organizations.

    When it comes to recognizing disruptive changes, I’d like to add the following:
    As you have pointed out, all people tend to suffer from myopia, more or less. Experts are likely to suffer a little bit more as their views and way of thinking in a particular field are much more entrenched. That makes them very valuable for improvements and developments in their respective field of activity, but it might pose a raodblock to take different views and question fundamental assumptions.
    The complementary part here could be diversity and cross-functionality. I think, for disruptive innovation it becomes increasingly important to integrate and combine specialized knowledge. This is related to the claim towards a T-shaped professional for innovation. As human’s capacity is restricted, either the vertical (expert) or horizontal bar (integrator) can be pronounced. Maybe we need to think about how to properly educate those “integrators” in order to handle radical innovation decently. For sure, they need to be strong networkers.

    What do you think?


  5. Hi Ned:

    Intuitively, I think you are correct and some of our research on knowledge communities supports the idea of brokers as key nodes in the network. I thought I’d put the results of the simualtion study up because it suggests another view. Maybe the nodes aren’t as good at connecting as we think they should be. There might be reasons for this. One study suggests that these nodes become filterers rather than connectors because it makes that person more powerful to control information. Rewarding the brokers for the rught behavious seems to be important.

  6. Hi Ralph
    I think you are onto something here. The network issue isn’t the only dimension to the problem. It’s also about how we can get experts to think differently and engage with other knowledge domains. Some techniques such as devil’s advocate and dialectic enquiry might help. I’m also liking the idea of a ‘premortem’ to provoke discussion around strategy. It seems to work well in getting people to ask the right questions.

Comments are closed.