Knowledge management is a phrase capable of driving the strongest-minded IT director to despair, and it can have a similar effect on all members of an organisation subjected to it.
The issue has been beset by the problem of definition. The literature of academic management theory does not help - there are at least 25 definitions floating around the journals, warns Ashley Braganza, senior lecturer in information systems at Cranfield School of Management.
But the last thing a company should do is get bogged down in a debate about "what is knowledge?"
Instead, urges Braganza, what is needed is a more practical approach which focuses on the needs of companies making a living in a tough world. From a pragmatic point of view, what, he asks, are the characteristics of knowledge as found in organisations? Only when we know that, can we start asking whether it can or should be managed or why, and how.
Knowledge exists in two forms - tacit and explicit. Tacit knowledge is the knowledge and behaviour that people bring to bear when doing their jobs, while explicit knowledge is the tangible and codifiable information available to them.
"Good knowledge management is how to bring these together," says Braganza. Knowledge, he points out, also exists in different places, at multiple levels with different degrees of accessibility or comprehensibility.
A further, and very real complication is that knowledge about the same thing can be different for different groups of people.
"What does a pallet of baked beans in a supermarket mean?" asks Braganza. "To sales it means commission; to marketing, revenue; to finance, an invoice; to the warehouse, shelf space; and to logistics, a size of truck.
"The same item can have so many different implications or meanings that it becomes an amorphous mass of everything and nothing. Is it any wonder that knowledge management initiatives drift off into a morass of ambiguity?"
Braganza says the worst mistake that an organisation can make when it comes to knowledge management is to ask staff, "What are your knowledge requirements?'"
"No one goes into work saying, 'Hmm, what are my knowledge requirements for today?'" says Braganza. "They say, 'What have I got on today?' That is the explicit knowledge they want."
They also want tacit knowledge, by way of the behaviours they bring to bear, depending on the job they need to do.
For example, someone involved in new product development will need to know how to listen, take on ideas, make suggestions, brainstorm - knowledge that is actually counterproductive to, say, someone working in the warehouse, where creativity is undesirable.
"There is no point them saying, 'Now, where shall I put the baked beans today?'" points out Braganza.
What is key, says Braganza, is to let those who use knowledge define what they want by way of knowledge. It has to be demand-pulled, not supply-pushed by IT deciding what others need to know to do their jobs better. And for the pull to be most effective, he argues, it has to take place at the level of the business process.
An organisation needs to ask, in the context of knowledge management, what its business strategy and goals are, and who its stakeholders are. But it is the levels below that - the business processes that need to be done, and the activities that fulfil those processes - where knowledge management should be targeted for maximum effectiveness.
"You have to go in at the business process level," he says. "That is the unit of analysis."
But, Braganza warns, a business process is cross-functional, so it must take account of different meanings attributed to items, such as pallets of baked beans, by different functions.
An organisation keen on knowledge management should therefore identify all the business processes within it, and all the activities that take place within one process.
Then, in light of understanding what the company's business goals and strategy are, it can start to understand what knowledge is needed to transform those processes either to meet changed goals, or the same goals more effectively.
"You have got to ask the people doing the day jobs," says Braganza. "People in the organisation need to get a handle on it. Knowledge management cannot be 'done to' an organisation."
Nor do you necessarily have to apply knowledge management to all four corners of an organisation. You can target the areas it will benefit most. "How do you know where it will pay off most? By understanding the business strategy and who the stakeholders are you can prioritise your efforts," says Braganza.
Drawing up a grid that sets out the priorities should take weeks, not months. "You have got to know what you have got and where you need to apply knowledge management most. If that takes months to do, you have lost the plot," he says.
As for why to do it, Braganza says, "Knowledge management is about getting the right information to the right people at the right time in the right format and in the right order that has meaning to them - those individuals will then exhibit the right behaviour.
"That means they must act on that knowledge."
Since the software that organisations need to handle knowledge management is out there already, and vast numbers of IT systems are already delivering information to vast numbers of staff, the key issue with knowledge management, says Braganza, is to ensure that systems are delivering the knowledge that their users actually want and need.
Ashley Braganza was speaking at the Cranfield School of Management IT Directors' Forum in September