Andrew McAfee in his interview with Dave Weinberger, acknowledged that knowledge management 1.0 did work on occasion, but was more often a failure. He said it was a failure, in part, because the system predefined the structure of knowledge and asked people to fill in the holes. But this predefined structure often did not meet an individual’s needs and, anyway, why would a busy person want to take the time to fill in the holes in someone else’s structure? All of these are good points but what about the times when knowledge management 1.0 worked?

Talking about the times that knowledge management worked sometimes reminds me of Tom Stoppard’s play, Rosencrantz and Guildenstern are Dead. Stoppard wrote about these two minor characters in Hamlet and covered what they were doing when they were not on Shakespeare’s stage. Now many people beat up on knowledge management and think it is a fool’s errand like the tasks of Rosencrantz and Guildenstern. The times it works are largely in the shadows. The times that I saw knowledge management work best were the times that it did not try to be an enterprise software solution on the main stage that provided answers to everything using a predetermined structure.

It worked best when it made available the advice that a supervisor would give to an individual worker at the task level around a specific process. This online advice was frequently needed because the company had laid off most of their supervisors or there were too many new hires for supervisors to accommodate, or tasks were now more geographically distributed. It worked for very specific work processes where following a specific process was good thing.

For example, my first exposure to knowledge management was in support of an underwriting process for property casualty insurance. The firm was losing buckets of money, in part, because there was no underwriting process in place. Risks were just priced which set up huge losses. So some of the top underwriters documented how they thought and put it in an accessible system that was aligned with the steps in the underwriting process. But this system was not shoved down the throats of the underwriters nor was it used to automate underwriting. It just gave everyone the benefit of the best thinking on each type of risk so they could make good decisions. It also gave them access to experts if they needed more. Prior to this the insurance company’s IT system was used to document stuff for the home office and was avoided, if possible. Now the knowledge management was designed by respected peers to help get the immediate job done and was very popular. I saw this again in many call centers and other process driven work. In these cases some structure was a good thing and those who used the knowledge management system performed at higher levels.

But so many tasks are not so clearly process driven and here is where knowledge management was, and remains, much less successful. It is also where Enterprise 2.0 has great potential for success. I happen to move from a large corporation to a personal enterprise of one just as I was introduced to blogs and then to what became Enterprise 2.0. Ray Lane’s concept of a personal enterprise was an easy fit. At the same time, I saw how these tools could greatly benefit my former colleagues in a large consulting firm. I also saw that, even more than knowledge management, it requires a fundamental shift in how work is done and a sense of openness most companies are not ready for.

The first large consulting firm that successfully adopts Enterprise 2.0 will have a huge competitive advantage, in part because, as Andrew McAfee noted, it is so hard for any company to make the transformation (see James Robertson’s post). It will be especially hard for the big consulting firms where transparency is seen with suspicion by many managers. And it will be especially powerful for these same consulting firms where sharing knowledge is the major product and the main competitive advantage. But it will take more than the desire for a “silver bullet” as Jim McGee wrote. It is not a software solution.