Leading Organizational Learning

awry is in failing to see the tremendous differences between
(1) what is required to take a function that has involved human
judgment and reducing it to a problem of information processing
(for example, Amazon’s book recommendations) and (2) what is
involved in supporting a function in which human judgment con-
tinues to play a central role (such as determining how to resolve a
complex insurance claim).
The first problem, decision and process automation, requires
that a model be able to drive usable outputs in most or all cases
(potentially escalating a process to a human decision maker if cer-
tain conditions hold). The second problem is not so much about
“managing knowledge” as it is about providing a context and a set
of tools that enable human actors to maximize their effectiveness.
Too often, knowledge management initiatives, as distinct from
decision and process automation, fall into the pattern of encourag-
ing participants in a business process to turn their judgment and
experience into documents (“codification”) and then creating a
technology that enables others to obtain these documents (“access
knowledge”) to inform decisions and actions. The stories told to
obtain sponsorship for these initiatives seem plausible enough, rest-
ing as they do on the intuition that no one has available the expe-
rience base or the expertise that the full range of employees possess
collectively. We can all think about situations in which someone
took action without the benefit of what a colleague knew, resulting
in a costly missed opportunity. The promise that such mistakes will
be avoided in the future pulls powerfully on heartstrings and purse
strings both.

Challenges of the Codification-Access Model

Unfortunately, this codification-access model rarely works. The
knowledge management community has focused on the problems
associated with incenting codification and generating awareness of
the valuable knowledge to be found in their systems. These prob-
lems are real but less fundamental than three others:

28 LEADINGORGANIZATIONALLEARNING