At a company where I worked many years ago, circulating correspondence was an everyday practice. It was also one of the simplest and best knowledge management techniques I've ever seen.

Whenever you wrote a letter -- and we wrote a lot of letters -- you made two copies: one to file, one to circulate. Every week (or every so often) you took the circulating set, culled any that included confidential dope or made you look more stupid than usual, stuck on a buck slip, and put them into your outbox. By the time the folder returned, it was generally time to refill it and send it out again. Everybody participated, including the chairman and the president.

Has knowledge management been oversold? What's it good for? Discuss.

"A whale ship was my Yale College and my Harvard," said Herman Melville's Ishmael; when it came to learning my job, circulating correspondence was mine. Reading my superiors' letters opened a window into how they conducted business with the world outside; I aped things more experienced colleagues did, and saw how they handled tricky situations; I copied useful addresses into my Rolodex (another antique). I learned who knew what, and that made me better at asking for advice.

Circulating correspondence was obligatory, easy, and genuinely useful. As such, it stands in stark contrast to much of what today passes for knowledge management -- an activity that has assumed immense importance in the corporate world. Knowledge management undergirds a growing edifice of ideas, techniques, and technologies. International Data Corp., a research group that focuses on technology, estimates that poorly managed knowledge costs the Fortune 500 about $12 billion a year. Reasons for the lost money, by IDC's reckoning: "substandard performance, intellectual rework, and a lack of available knowledge management resources." You don't need numbers like IDC's to know that the need is there: Just consider how much time you waste searching for information that ought to be at your fingertips.

The response to the need for knowledge management has been astounding. In Intellectual Capital in 1997, I wrote, "If the subject of intellectual capital ever spawns a business fad, it will be under the guise of 'knowledge management,' because there's money to be made selling software, systems, and consulting services with the touted goal of allowing every person in an organization to be able to lay his hands on the collected know-how, experience, and wisdom of all his colleagues." I was more right than I dreamed. Knowledge management has become "KM," and there are national and international KM conferences, local KM forums, the Journal of Knowledge Management, and Knowledge Management magazine. By IDC's estimate, knowledge management software and services will be a $6 billion industry in 2002.

As time marches on, and Moore's Law with it, this technology gets swifter, stronger, and subtler. Why, then, is there a nagging sense that all of it misses the point? Or that much of the time it yields no more insight than a file of circulated letters? A couple of stories will get us toward the answers.