Monday 16 October 2017

What's wrong with this KM picture?

There is a common diagram in use in the KM world, which has at least 3 major flaws, so apply it with caution.

I am going to be contentious again in this post, and draw attention to the failings of a very common KM model. I do this because I think we can do better, and because this model draws us into ways of thought that can cause us to miss opportunities.

This is a very common sort of picture in the KM world - a one-way learning loop with "knowledge capture" or "knowledge collection" as step 1 or step 2.  Here is a version I drew just for this blog post, but you can find many other examples.

Simple and seductive as this model is, there are three main flaws which you need to be aware of if the model is not to lead you astray.

Firstly, knowledge doesn't have to be captured, to be managed. 

As Stephen Denning pointed out recently,  there is some knowledge that can be collected or captured (he calls it precision knowledge) and some that cannot (he calls it intuitive knowledge). He concludes that "It is only in the area of precision knowledge that a knowledge collection (captured knowledge) will offer a clear guide to action".

Can you manage knowledge if it is not captured? Sure you can - you manage it through conversation - through arranging the right conversations between the right people;

  • conversation within the team, for example through after action reviews
  • conversation from on team to another, for example through peer assist or knowledge handover
  • conversation within a community of practice, for example through knowledge exchange
  • conversation from one person to another, through mentoring, coaching, knowledge interviewing
Now, in each case, there may be a case for capturing the knowledge as well, if there are other people who also need to learn, but the point is that capture is only an option, and that the capture is not the point - the point is transfer, and that the transfer will be richer than the capture, and will include the intuitive as well as the precision knowledge. 

Secondly the model is a one-way push model, with no room for Pull. 

It starts with a piece of knowledge, which needs to be captured. There is no indication in the loop that there is a need or a demand for that knowledge - only an arrow from "knowledge" to "capture".

Knowledge management is not as simple as a one-way loop. There are two driving forces for knowledge, Push and Pull, and Pull is a stronger driver. Knowledge flows most easily when it answers a need, and knowledge flow is less like bottles moving along a conveyor belt than it is like electron holes moving through a transistor. The push loop above has no place for "knowledge seeking", for example, as this is a Pull activity and the arrow would go the other way.

Thirdly the model usually has no place for co-creation of knowledge.

It usually is seen as a transaction flow from knowledge creator to knowledge re-user, but we know that these flows are not simple, and that when they work well the knowledge is not transferred like a unit - like a bolt in a piece of machinery - but changes its nature with every interaction. As people discuss what they have learned in a lesson capture meeting, they co-create the knowledge through conversation. As knowledge is stored in a wiki (for example), it is co-created through interactions in the wiki, as new insights are added and misconceptions removed. As knowledge is shared in a community of practice discussion, it changes again, growing and becoming more robust with each interaction. The knowledge user adds new knowledge to what they already know.

The danger of the model

The danger of seeing KM as a push-based flow of knowledge starting with capture was once again brought home to me recently, when talking with a young knowledge manager. Her approach to KM was to start with capture, and she was desperately trying to convince people in her organisation to submit "knowledge documents" to the knowledge base. She was failing, as people did not see the point. They did not believe in KM, they did not think publishing would have any effect, and they did not see "whats in it for me." Instead I suggested that she started to work on developing a demand for knowledge, and on setting up "knowledge conversations" such as after action review.

No comments:

Blog Archive