Dead Presidents, them dead presidents...

Hmm, shouldn't the application of knowledge be measurable using the traditional methods of time and/or money? Regarding the allocation of money, I'm sure many would rather stick to the status quo use of golf-course deals and hiring consultants -- both often done sans theories behind the proposed initiatives. ;-)

The notions of going after knowledge work productivity via indirect measures seems appealing. Mostly, because it requires you to develop a theory about how the application of knowledge creates some new value. On the downside, I suspect most organizations aren't keen to rely on theories when it comes to allocating money to solving problems.

Seriously though, it might only take a few people to bring in k-logs through the backdoor, and begin using them informally. If the first k-logged tidbit which hits paydirt when used has a pointer back to the k-log which caused the initial spark -- and management is made keenly aware of it -- it might be enough to trigger wider-spread adoption, at least in the unit(s) involved. Think of the Tom Watson IBM legend you mentioned earlier, but with more people expected to do the thinking, reflecting (and k-logging)...

I think the issue is more about a company's cultural attitude toward risk and the willingness to allow some folks to swing for the fence occasionally. For instance, 3M once expected many employees to devote about 15% of their time to R&D efforts -- I'd think the engineers who were to do this could certainly be categorized as knowledge workers.

Looking back in a few years, we may see many companies that adapted more rapidly to change had expectations about employees being paid to 'surf and k-log'.

Greg poses an interesting grassroots entry strategy for getting KM justified by way of k-logs. If you establish a beachhead for k-logging you create a contemporaneous account of how and where ideas surface in the organization. If you then have a successful innovation, you've now got an audit trail.

If the old adage was "success has many fathers, but failure is an orphan" you now have a DNA test to establish paternity. Which suggests that one of the threats to k-logging success could be free-riders who've managed to achieve power by taking credit for others' ideas and work. Something to ponder.

9:33:39 PM •  • comment  

Krishna Narayan, one of my partners when I was at DiamondCluster, came in to offer some real world perspective on KM. A lot of his initial focus was on what I might label "straightforward" KM opportunities. For example, he talked about efforts in one chemical company to discover patterns in customer buying habits that might suggest either future sales opportunities or, in other context, that the customer was quietly acquiring the makings for something illegal or dangerous. On the one hand you could label this as an example of datamining. On the other, you could list this as a KM application.

Krishna also offered several examples that centered on the knowledge capture/retention issues raised by tunrover in organizations. Two examples came from governmental settings. One was a concern within the State Department about how to capture the insights and lessons learned from key embassy staff around the world reaching retirement. The other dealt with a similar issue around knowledge about key Federal information systems held by Federal IT managers approaching retirement. In both instances, contracts were being let to KM vendors to develop and use systems to somehow capture and disseminate this knowledge. The central question, however, that didn't appear to get asked or answered in these projects was what was the value in trying to capture knowledge about aging and obsolete systems or about long-dormant political issues. The value might be there, but there wasn't much evidence that anyone was trying to identify it.

Two thoughts in particular stuck with me.

One was the notion the KM may be analogous to AI when it comes to definitions. Once you figure out how to do it with technology, somehow it doesn't seem like KM or AI anymore.If that helps us move away from endless discussions of "what's the right definition of KM?" that would be all to the good. A more productive question is "what makes knowledge work different?" In particular, we ought to be thinking about where our assumptions about work in general lead us down unproductive paths when applied to knowledge work.

The second thought that I've been noodling with since Tuesday is how optimizing and satisficing apply in the KM realm. Herb Simon coined the term in his research on economic decision making. His insight (nicely summarized in a column by Ronald Marks) was that, contrary to conventional economic theorizing, few decision makers ever seek optimal decisions. The world works just fine if you find satisfactory ones. Most of the discussions on KM that I've encountered bog down in a quest for optimal answers. Knowing that there are instances where what we know in one part of the organization fails to be applied elsewhere, we obsess about designing a KM system that will somehow ensure that all knowledge gets captured and routed to the right place. We'll get far more progress out of searching for answers that satisfice, that move us a little farther along the road.

2:08:49 PM •  • comment