Euan is definitely back from his holiday with a provoking post in which the key paragraph states Much of corporate IT has been designed to replicate a hierarchical view of organisations which pigeon holes people and bears little relation to the real world they work in. Now this triggered a key issue for me on the application of complexity thinking to systems design (which applies to IT and other systems). The classic (in the sense of classically misguided) approach is to design the ideal system based on how senior management/the consultant/some poor sod who has been told to do the specification (delete as applicable) think the best of all possible worlds should be.
We saw (and still see) a lot of this in knowledge management so I will use that to illustrate the short comings of the ideal and the evolutionary opportunities of the real.
An idealistic approach goes something like this:
In a perfect world everyone would update the KM system with details of their latest projects and contribute to a CoP relating to the area of their expertise. Then when a new project starts all members of staff would check out the KM system for relevant material, use it and this safe money and time by not reinventing the wheel. So we will set up a system based on best practice (add industrial best practice if you are in government to ensure that the approach is not applicable) and then train people to use the system as it is in the overall interests of the organisation for them to so do. They will realise that a small amount of their time given now to maintain the system will save them (and the organisation) time in the future so that they will see it is in their interests to do this. Oh and we have some add hoc collaboration tools being used, some outside the firewall but these cannot be considered as a corporate activity, and the new system requires us to abandon these in favour of a common approach.
Now of course this fails to appreciate some of the basic rules of knowledge management in respect the conditions under which we share knowledge. It also fails to understand the epistemological messiness of human knowledge creation and use. Critically it fails to realise that the few minutes to maintain this system cannot be considered in isolation. These few minutes have to be added to the many minutes spent on email, the minutes spent on responded to climate surveys etc from HR etc etc.
Of course the ideal is never achieved. Utopianism is doomed as it leads to inertia. If we have the best possible world why would we change it? Utopian writing from the Garden of Eden through Thomas More and beyond all describe a static system. However when people fail to use the system, or it is not used properly the whole thing is either swept under the carpet and forgotten; a lot of KM systems presented at conferences by their enthusiastic creators are de-facto in that state fairly shortly after, or possibly during the presentation. In other cases a new argument is brought into play, namely the question of cultural alignment. The argument goes something like this:
The system has failed for cultural reasons so we will institute a cultural change programme to align personal values and objectives to the greater good of the organisation as a whole. We will invest in further automation and control mechanisms to ensure alignment of those goals with the needs of the system. Personal incentive plans etc will now include incentive payments (or disincentives) to ensure full participation.
In other words we want our ideal system and it’s idealised outcomes..
Now for some types of system – much of ERP in manufacturing processes, payment systems etc – I can see that this is the right approach. The extension of the approach to any highly human based system such as CRM or KM on the other hand is highly dubious and to my mind doomed to failure in whole or in part.
Fortunately there is an alternative, which is to switch from an ideal, to a naturalising approach. Here we treat the system as a whole (both the people bits, the organisation its environment and the technology) as part of a complex system. We know that in a complex system we cannot manage outcomes, only starting conditions. So that is what we do. Instead of designing the system based on ideal outcomes we start with a semi-robust architecture into which we introduce various tools. For collaboration for example we can use list serves (still the most universal tool for cross silo collaboration, blogs, wikis, a test version of that tool the IT department keep saying is wonderful. That really cool narrative software from the interesting start up company. We make these tools available to different groups of people. We train, we celebrate and promote success, we see what is imitated and what is not. Critically we spend time with the failures so see if they present a greater opportunity for learning. Failure is more useful learning mechanism than success which is why I worry about story based approaches which only produce, or tend to produce positive stories (its another example of ideal not the real but more of that on another occasion).
As patterns of use stabilise, then we start to introduce some stability. We start to take a Just in Time approach to KM. We realise that informal communities are a lower cost option to knowledge storage and use than formal systems. May be a common document store so that collaborative systems use and pass around references not large chunks of data for example.
Basically we allow the system to evolve from the interaction of its elements, not the idealism of a designer who does not share the context of the present let alone the future.
Cognitive Edge Ltd. & Cognitive Edge Pte. trading as The Cynefin Company and The Cynefin Centre.
© COPYRIGHT 2023
Around this time five years ago I had just landed at Heathrow on a flight ...