So, what's not to like about Knowledge Management? It sounds very Web 2.0, Enterprise 3.0 or something like that. It's probably what distinguishes the sheep from the goats in the automated business service management business in a million PowerPoint presentations.

And yet, the whole idea disturbs me, with its connotations of managing nice clean packages of knowledge, away from contamination by nasty, troublesome "wetware" artefacts. I feel much the same way about "Human Resources Management"; what is wrong with managing people?

Of course, "knowledge" (which I define as information that you actually use to do business with) is what we all depend on for our success; and I presume that we all manage it in some way already. But "Knowledge Management" (with capital letters) has connotations of sucking the contents of peoples brains into some kind of repository, so that its original custodians can retireor get made redundant. Of course, there are some issues with this sort of knowledge management process not unrelated to the fact that many of its "victims" are reluctant to share everything they know, possibly in the hopes of coming back as a part-time consultant on twice their previous remuneration. And what is captured is seldom used (which, in my book, stops it being real knowledge), because the repository interface is usually rather poor and needs a lot of learning, which not only discourages use but also discourages any effective knowledge maintenance, so the expensively collected "knowledge" first becomes out of date and then thoroughly distrusted.

There must be a better way, especially as most professionals like sharing knowledge and being professional generally; and my experience is that sharing knowledge gets you thought of as an asset and, in fact, protects your career when things get tough.

The first stop on this better way is to change the way you think of the join between "management" and "knowledge". Think of it not as the managing of little packets of knowledge apart from the people who discover or develop it but as the the use of knowledge (as opposed to gut feelings, guesswork or even untested theory) as an enabler for good management.

The end point on this better way is probably effective knowledge transfer. This involves creating or discovering knowledgean activity that needs clever people who build knowledge out of books, training courses, talking to peers and their practical experience; and who then transfer it to wherever it can be used. This is a two-way process because as knowledge is used it is refined and extended, and this updated knowledge must be added back into the pool for reuse.

Knowledge transfer must be institutionalised, which means that it becomes a routine part of "the way things are done around here". This, in turn, implies organisational maturity, so that the stages in any process are reviewed with the aim of discovering what can be learned from them and added back into the process (knowledge is an enabler for process improvement); and good people management, so that all involved see transferring knowledge as an important part of their career progression. It also implies tool support (perhaps we should all review Vanevar Bush's "Memex" concept from some 65 years ago), but the tools must be easy to use and supportive of people and knowledge transfer processes, not seen as a "silver bullet" replacement for them.

One area of tool support that interests me is some sort of visual feedback on knowledge quality and knowledge currency; I don't know of many "knowledge management" tools that do this yet, but if you consider knowing what the "end-to-end user experience" is as one aspect of the knowledge around a system (and I do), Compuware's monitoring and systems management tools monitor their own usage, so they add to the body of knowledge they collect with information about areas of the organisation that aren't making effective use of the tools available and can track process improvement. Always remembering, of course, that the information collected about user experience is only knowledge if it is used effectively, presumably to improve user experience and, thus, user loyalty.

"Knowledge", considered properly, has a very wide scope and nowhere is the idea of "people and process before (not instead of) tools" more important than in the field of effective knowledge transfer.