Dangerous decentralization

In a previous post, I observed that in the political realm, even so-called grass roots organizations are highly centralized. While such organizations claim to be the voice of the people, in reality the role of the people is to take orders from Beltway-based consultants.

Obviously businesses face similar issues. On the one hand, businesses want to provide a cohesive, common picture - hence, centralization. On the other hand, businesses want to harness the power of their employees to come up with creative solutions - hence, de-centralization.

Jim Clemmer tells of an organization that tried to do both:

We once halted an executive retreat and everybody went home after the group of seven division presidents and corporate staff vice presidents couldn't agree on whether their values were centralization or decentralization. Trying to do both at once was ripping the organization apart. The CEO never could decide which direction he wanted to commit to. He was eventually fired as frustrations and infighting rose while organization performance fell.

Clemmer himself has a clear preference regarding centralization vs. decentralization:

I am an extreme (some might argue dangerous) decentralist. Since I began my management career, I've given people high degrees of autonomy. I've run even small organizations to the point of such inefficient decentralization that people are running their own show. It works.

Clemmer's argument is that decentralization provides positives, and centralization negatives, that aren't necessarily captured by the metrics that many centralized organizations use.

What doesn't show up is the alienation, helplessness, and lack of connections to customers or organizational purpose that centralized bureaucracy often brings. The energy-sapping and passion-destroying effects of efficiencies may save hundreds of thousands of dollars. But traditional accounting systems can't show the hundreds of millions of dollars lost because of lackluster innovation, mediocre customer service, uninspired internal partners, and unformed external partnerships.

Notice that Clemmer is not only looking at things that could conceivably be measured - customer service quality, for example - but also "opportunity cost" types of measures - the strategic alliance that was not made, the product not invented.

Because some of things can't be measured, perhaps the best way to measure the value of decentralized is to find two similar firms - one centralized, one decentralized - and see which one performs better. Unfortunately, Clemmer's article fails to name any specific companies that follow a decentralized model. However, Clemmer's website includes a list of clients who have attended his seminars. While some of these clients are personnel/human resources groups and the like, it's instructive to look at the technology companies that are listed. Now this doesn't mean that the organizations themselves follow Clemmer's advice, but at least someone in the organization is paying lip service to being interested in them.

Advanced Technology Systems
Algorithmics
Aspen Technology
Bell Canada
Canadian Information Processing Society
Canadian Radio Common Carriers Asociation
Co-Operator's Data Services Ltd.
Communitech Technology Association
Entourage Technology Services
IBM
Ironside Technologies
Internet World
J.D. Edwards
Lumonics
Maritime Tel & Tel
Matsushita Electric (Panasonic, Technics, Quasar)
SAP
Sasktel
Software Support Professionals Association
Stentor Canadian Network Management
Sun Micro Systems
Symantec Corporation
Targray Technology
Xerox

Now I'm not familiar with many of these companies, but it's amusing to note that two of them - J.D. Edwards and Sun Micro Systems - are now part of Oracle. And Xerox is probably most famous for letting its technological advances be used by other companies (Apple and Microsoft in particular).

Does anyone know of an example of a firm that is decentralized, and successful?