Too modest to puff his own horn, Thomas Stewart takes on (the failures of) knowledge management.

Meltsner, Kenneth
Tue, 15 Jan 2002 17:57:50 -0500

What keeps me up at night are managers that think there are silver bullets.  Magic fixes for software reuse, finding information, organizing information, getting people to talk to one another, etc.

[And now we're being told it's usually foolish to worry about re-use, what with the growth of extreme/agile programming techniques.]

It's more than a bit odd to invoke Dave Winer (or David Weinberger, who isn't interchangeable despite some similarities) in this forum, but the conversation does come first.  If you have a chaotic, poorly organized folder full of useful and useless information from a mailing list, at least you have the information somewhere in that folder.  If it's really valuable, you can send an intern on an search-and-destroy tech writing mission. If it's not that valuable, it's not likely an automatic categorization tool will find any gold nuggets.  Even if no one mines the email ore, at least you had some good exchanges and solved some real problems.

One Booz Allen Hamilton consultant I worked with had a nice pyramid view of Knowledge Management.  At the top of the pyramid was "KM for experts."  This focussed on getting the experts connected to one another, getting them to talk, etc.  In the middle was "KM for competent people."  Here, you wanted support tools, outlines of methodologies (but not mandatory processes), contact lists, etc.  Finally, at the bottom of the pyramid was "KM for the masses."  Fill-in-the-blanks processes, codified manuals, expert systems, "performance support," etc.  

The key, she told me, was knowing what your target audience was.  If your consulting firm thought of itself as the elite problem-solvers, the gurus and wizards of their disciplines, you needed to focus on KM for experts.  Competent consultants needed the mid-range, and companies that were built on hordes of freshly minted BS-holding kids needed KM tools for the masses.

This breaks down, of course, as an elitist view of the world.  The copier manufacturer with the case-based support system for the call takers thought they were at the "masses" level of KM.  Turns out they needed to cultivate their guru and his disciples instead.  It might not be an 80-20 proportion, but I'd bet that a lot of help desks find that most of their resolutions (hard and easy alike) come from a relatively small part of their staff.

[Erik V-something wrote up a similar case of a case-based help desk system, and how, despite the fact that the users loved it, it wasn't effective at fixing customer problems.  I'll try to dig up the reference.]

Ken Meltsner