Monthly Archives: June 2011

Agile knowledge management

Some brief thoughts on Agile KM…

Libraries are Creepy Knowledge in a dynamic context is constantly in flux. It is subject to continuous as people gain new knowledge from observation, experience (and often mistakes) and discard or modify obsolete views. Attempts to consistently harness this rapidly-changing knowledge therefore need to be agile in nature. To this extent, using the word “management” is a red herring in these contexts. A top-down approach is highly unlikely to anticipate the scope, lifespan, or use of the knowledge generated, and will lead to white elephant knowledge management systems: as a result of classic “build it and they will come” thinking.

As Johanna Rothman points out, knowledge generally exists in people’s heads, and the transfer needs to be done through f2f communication on teams, and in collaboration with other people. More generally, knowledge is created by people and only delivers value when successfully used by others, whether this is tacit or implicit knowledge. Fundamentally, knowledge is social in nature, and an Agile approach has something to offer.

The most efficient transfer of tacit knowledge is when person X, who learnt something, explains it to person Y, who immediately has use for it. Anything that reduces the richness of the communication media between those two endpoints, be it a repository, a remote communications link, or even the passing of time, will diminish the value of that knowledge. As Alastair Cockburn has identified, adding a whiteboard will likely optimise the transfer.

Valuable implicit knowledge is more likely to be embodied in the values and practices that the team adopts. Moreover, in an agile environment they will adapt over time to the changing context. True insight, wisdom, or generalist knowledge will endure as a longer-life meme, surviving social combat with alternative memes, and ideally changing its host DNA in the process.

As a necessary proxy for f2f communication, KM systems must not only be as social as possible, but implement effective memory, to ensure the fidelity and durability of the knowledge they store. A whiteboard, checklist, or wiki are examples of lightweight collaborative knowledge sharing systems. No matter how advanced, KM systems need to provide easy authoring and retrieval, often in collaboration with others, and the success of the system can only be measured by how well the someone’s knowledge actually generates value for someone else. If the KM is not doing this, it’s not effective. Taking an agile approach, through a people-oriented, and continuous-improvement process, will generate better results than a big-design-up-front KM system.