I was always bugged with the buzzword "Knowledge Management." Not because it is a buzzword... but because it appears to NOT be a buzzword. A buzzword should either be really concrete, or vague enough to lead to questions -- like "Enterprise 2.0." Instead, "Knowledge Management" is somewhere in the middle, and sounds like annoying advice:

• The key to getting rich is making more money!
• The key to winning races is going faster!
• The key to a smarter enterprise is managing your knowledge!

As such, I feel that the very phrase "Knowledge Management" might have led people to ask the wrong questions, and implement the wrong solutions... I think Chuck Klein down here in Albuquerque said it best:

We don't need "Knowledge Management." We need knowledge capture and context management.

That puts it pretty well... the goal is to capture as much knowledge as you can, and store it safely and securely. At the same time, you need to constantly gather more and more context, so you know what information to get to which people, and when. Information without context is worse than useless: its merely clutter that wastes everyone's time.

Too many projects lose focus on the context management problem... some of the easy questions revolve around things like metadata and keywords, but that is rapidly becoming insufficient. As the amount of information you manage gets larger and larger, you need to ask a lot of hard questions before you can maximize the value of your system:

• Who is the intended audience for this item?
• Where does this item fit in my taxonomy?
• If people like this document, what related items would they like?
• How would people find this item, if your search engine did not exist?
• Under what conditions should we archive this item to reduce clutter?
• Under what conditions should we destroy this item to save storage space, and mitigate risk?
• Who is the current user?
• Where is this user, and how are they accessing the system?
• What is the user's search, download, and feedback history?
• What is the most effortless way to gather feedback from this user?
• Based on this user's past behavior, what information are they likely to want next?

Some good Knowledge Management folks already ask these kinds of questions... but I feel that not enough clients understand what kinds of questions to ask. If we used more specific terminology -- like context management -- it gets people thinking about the problem in a very concrete way... and I feel would lead to better implementations.