Yesterday I was asked – “What are the essential components of a knowledge management system?” Now this is no easy question because it is laden with assumptions, most often that we can actually build a knowledge management system from the ground up! The assumption being that a knowledge management system is a hard system, which it isn’t. Here’s my answer.

A knowledge management system is a soft system of systems with open boundaries. It is made up of components from the mnemonic PISHI (1), which stands for people, infrastructure, software, hardware, and information. This classification represents real things in the real world, and each of these things is linked together by processes so that an output is realised to meet an objective – in this case knowledge. In the modern business world, all components are essential.

People are absolutely central to a knowledge management system. People have tacit, implicit, and explicit knowledge, noting this distinction is artificial and a problem in itself. They also possess unsynthesised information, and interact in some way with the one or more of the other components to produce knowledge. For example, people manipulate information using software loaded onto computers (hardware) that are provided in office space (infrastructure) to produce knowledge artefacts like books and journals. Equally, people can exchange information via a video camera (hardware), which is located in meeting room (infrastructure), to exchange knowledge.

So what’s the problem? Well there are several. First, knowledge is not a static stock to be managed like an inventory item. It is dynamic and personal and has context. This means that it can, and does, degrade to being data or information over time.

Second, people synthesise information, context, experience, and understanding to produce knowledge; and that synthesis largely occurs in the human brain. The problem for a knowledge management system is how to get the knowledge out of somebody’s head, and then codify it so that it can metamorphose into useful public knowledge. Often it will not be possible so other means of sharing the knowledge must be found, like peer-to-peer meetings.

Third, unless there is an enterprise-wide lexicon, which enables users from different groups to share and reuse the knowledge most codification efforts will be in vain. This is particularly problematic for large multi-silo organisations, with many trades and professions each with their own lexicon. Many government departments suffer from this malady.

Fourth, because organisations cannot create knowledge per se they can only magnify the efforts of individuals if the individuals are willing to share. This can be only be achieved in a supportive environment that encourages organisational learning. Organisational learning occurs when a business puts in place mechanisms that encourage employees to contribute to business goals. These mechanisms include physical and virtual places to learn and exchange data and information, as well as hard and soft media, and a no-blame, supportive and encouraging culture. If these elements are present then the conditions exist for the externalisation of knowledge, and perhaps its eventual codification - but, the difficulty remains in getting people to change their work habits.

Last, a knowledge management system is inherently a soft open system. This means that boundaries are permeable and difficult to position. What may be useful to one person in one part of an organisation may be useless to someone else in another department. Any knowledge management initiative must therefore establish clear achievable goals that deliver benefits to the organisation, or a sub-set of the organisation, and take into account user and stakeholder requirements. The key is it must be useful and solve a problem, or problems.

In summary a successful knowledge management system is founded on a clear understanding of:what the organisation considers to be organisational knowledge;what the organisation knowledge goals are; where knowledge resides in an organisation, and its form; what knowledge components must be intensely managed; and finally the absolutely central role of people in any system.

Regards Graham

(1) The construct of PISHI has been derived from an incomplete systems engineering PhD thesis by Patrick Byrne. Patrick includes nature as a system component; however I do not believe nature per se is a component of a knowledge management system. Patrick’s company HolisTech™ holds the copyright on the concept; however he has kindly let me use the idea.