Friday, January 15, 2010

Intelligent Knowledge Management

@snowded has started a series of blogposts about knowledge sharing across silos (part one, part two). He starts by identifying a number of different types of knowledge that (some people think) it might be valuable to share.

  • Data - avoiding duplication between data stores - for example consolidating all health records in a single database
  • Information - ensuing all relevant information is available to support critical decisions and actions - see for example my post on Joined-up Healthcare
  • Transactions - a sense that the citizen or customer has to navigate tortuous pathways through the bureaucracy
  • Practice - good practice developed in one area is not appreciated or understood elsewhere
  • Ideas - innovative combination and variation of issues, ideas, solutions and problems from within different silos

Obviously a mixed bunch of stuff here, and it seems very confusing to lump all of this under the heading of "knowledge management". As Dave points out, some of these look more like information flow. However, suppliers of knowledge management products and services (consultants and tool vendors) typically want to find as many different applications for their stuff as they can. Knowledge is therefore presented as a way of achieving a bunch of different things.

  • more efficient information systems and working procedures
  • improved decisions
  • more effective and convenient customer service
  • faster, more efficient and more consistent innovation
  • greater cooperation and collaboration

Underlying the knowledge management agenda are a number of (usually) unexamined assumptions
  • most of the knowledge we need already exists in people's heads - we just have to get it out
  • knowledge is additive - the more knowledge we can "capture" or "extract" the better
  • explicit knowledge is better than implicit or tacit knowledge - codification is a "good thing"
Thus knowledge management becomes a combination of documentation/codification and librarianship, plus whatever persuasion or coercion may be necessary to encourage people to participate in this game. Added to this is a kind of open-ended information retrieval, as exemplified by IBM's Lotus Knows campaign. (See my comments here.)

Undoubtedly there are some situations where this kind of thing can deliver direct value to the organization - for example in pharmaceuticals, where any delays in the assembly of regulatory information for a new drug can cost millions of dollars in lost revenue. But is this really knowledge management, or just a sophisticated form of information management?

An alternative approach to knowledge management is to leave it in people's heads, but then to provide knowledge maps that tell everyone whom to ask. The trouble with this approach is that the genuine experts often keep their heads down, for fear of being swamped with enquiries from around the globe, while attention-seekers use this as an opportunity to promote themselves. Thus questions of motivation and excellence must be addressed, and knowledge management becomes a branch of Human Resources.

Meanwhile, I've been reading a couple of older critiques of the 'knowledge management' industry.

Although these papers predate the recent explosion of web-based knowledge management tools (including Enterprise 2.0, however that is defined), the issues raised in these two papers largely remain unaddressed. Bergman points out that knowledge management is often regarded as a technological initiative. He mentions a number of very expensive white elephants, which failed largely because they failed to address the real business objectives or the real cultural and working practices of the organization. Wilson raises some more fundamental issues, suggesting that the knowledge management agenda is muddled and self-serving, and that Polyani's original concept of tacit knowledge has been grossly misinterpreted in order to promote an utopian fantasy of the manageability of the human mind.

So what (if anything) can we rescue from this mess?

Well we might start by shifting from "knowledge as a resource" to "knowledge as a process", as Gil Ariely argues in his paper Knowledge Management as a Methodology towards Intellectual Capital (Sept 2003, pdf). See also Interview with George Siemens (Dec 2009). Unfortunately, this phrase was long ago coopted by solution providers - for example in their paper Knowledge Asset Management, (June 2001), Ron Young and Gregoris N. Mentzas describe what they claim to be "one of the world’s first truly holistic and integrated knowledge management solutions". Holistic, huh?

Perhaps knowledge management shouldn't be about grabbing and hoarding stuff but about deploying it intelligently. This points us towards concepts like Evidence-Based Management, where knowledge is collected, analysed and deployed within a management loop. Consultancies promoting this and similar concepts include Pfeffer's and Sutton's Evidence-Based Management, as well as Management Epistemology. and the Knowledge Management Consortium - see for example the KMCI paper Corporate Epistemology (November 2003, pdf).

This kind of approach shifts the emphasis from knowledge sharing to knowledge embedding - grounding the work in the best available and critically evaluated knowledge, as well as actively seeking well-grounded knowledge to support organizational learning. Obviously collaboration is important here, but there is a distinction between collaborating-in-the-work (for example shared responsibility for decisions) and collaborating-in-the-knowledge (for example, shared responsibility for collecting and interpreting intelligence, connecting the dots). This distinction builds upon the RAEW technique we have long used for analysing organizations. A key question here is the relationship between decision and intelligence - how closely the expertise and authority should be coupled/aligned to the work itself.

What characterizes the intelligent organization is not just having more knowledge than its competitors, or even doing better at sharing the knowledge it has, but how effectively it invests its entire knowledge capital to increase its own viability and survival. This is so not a technology question, or even a best-practice question, but a strategic question. But you already knew that, didn't you?

see also When does Communication count as Knowledge Sharing?

No comments: