The End of Knowledge Management is Already Here

4 minute read
Stephen Fishman avatar

Despite what people think, the end state of knowledge management is already here. All future things are uncertain and that is not going to change no matter how much information, or how many artifacts, you have at hand. Entropy dictates that the problems of uncertainty multiply with the increase of information or artifacts.

The End is Here

In the past week we've seen analysis-based coverage on where document management went wrong, by Joe Shepley, and an enterprise-search proof of life claim by Miles Kehoe. Both make valid points, but are more evocative of a greater truth for both capabilities.

Shepley's instinct is right in direction, but we diverge slightly at the final synthesized perspective. I don't believe document management went wrong by itself -- I indict the entire raison d'etre of knowledge management. The end of knowledge management (and document management with it) is already here.

Since their inception, people have conceived of knowledge and document management as ends to achieve where everything important (in either document or factoid form) is known at the immediate moment needed and placed in the hands of the seeker. The idea is flawed at its root and will leave practitioners in these fields forever flummoxed.

Much like the problems these intertwined disciplines are trying to address, an unending set of dynamics are at play which make the aims of each intractable. The most easily understood of these is this: enterprise search lacks the basic conditions to perform like Internet search.

The Holy Grail is a Myth

Enterprise search, which is supposed to bring mass accessibility to the content stored and organized by the organizational learning systems, cannot ever perform like Internet search (at least not using any of the current models for indexing and relevance). Both knowledge management and document management are doomed to fall short of expectations when you realize that enterprise search tools will never perform like Google's internet search.

Ask anyone who has a history managing an intranet why their internal search functionality is not as good as Google and you'll see them wince. The playing field is just not fair (if you doubt me, ask yourself why Google appliances don't dominate the internal search marketplace to the point of monopoly). Google's algorithms depend on contextual assumptions that only apply in the Internet -- enormous scale, voluntary cross linking and referencing, and an entire industry devoted to indexing, scraping, tagging and SEO. Since none of these are present inside the walled gardens of internal content stores, the best technology for finding relevant content just doesn't apply here (at least not as people envision it).

Learning Opportunities

Aside from the prime technical/contextual constraints, there are basic limitations of humans living in an indeterminate universe. Heraclitus, a pre-Socratic philosopher from Ancient Greece, said "no man steps in the same river twice, for it is not the same river and he is not the same man." The same axiom can be applied to enterprises. Employees everywhere call out for a single version of the truth without realizing there is no such thing -- truth is contextual. The constraints of dogma would be suffocating if they ever succeeded in finding it.

From Tragicomedy to a New Beginning

My personal favorite dynamic that stifles knowledge management's efforts is that despite the increasing calls for it, precious few would ever use it. Look at the way most people in enterprises arc in their approaches over their career. In the first half, most people are predisposed to rebuild the wheels built by others. This isn't due to a lack of information, but by a lack of understanding that they stand on the shoulders of those who came before them.

The hard truth is this: even if you could build it, most of them would not come. Because sophisticated organizational understanding (i.e., culture) is manufactured via time, collective experience and repetition which cannot be delivered singularly via the online channel (and if it could, wouldn't that mean that the AI singularity had arrived?).

So what is a well intentioned organizational efficiency crusader to do in the face of such obstacles?

Reset your goalposts. Rather than seeking certainty via masses of information and artifacts, look to deliver a refined perspective and insight. Taking a page out of the DevOps and manufacturing playbooks, start by applying “hard to game” metrics to your efforts like “mean time to insight” (MTTI). I’m not sure how to measure this one, but it doesn't surprise me that several people have tried before and documented their efforts somewhere.

About the author

Stephen Fishman

Stephen holds a M.S. in Management from The Georgia Institute of Technology and a B.S. in Electrical Engineering also from Georgia Tech. Stephen has worked as a practitioner and leader across business, design and technology domains for enterprises and brands like PepsiCo, AutoTrader, Cummins, Chick-fil-A, the American Cancer Society, the CDC, Macy's, GM, Home Depot, Lowe's and others.