Maybe Hierarchies Are All About the Power

Now, my latest conundrum: In Here Comes Everybody, Clay Shirky makes a very compelling argument that says that hierarchies in management were necessary for communication. Because, he says, we now have electronic communication, these hierarchies can be much shallower -- we can have more producers and fewer managers.

One of the major case studies in Gary Hamel's The Future of Management is W. L. Gore & Associates, which was founded by the frustrated DuPont engineer W. L. Gore because he saw so much lost innovation at DuPont through archaic management processes. He built his new company around the goal of fostering innovation, and when Hamel wrote about it in 2007 it was doing $2.1 billion in sales with 8,000 employees.

Because his goal was "maximizing innovation" rather than "maximizing profits," the company was organized very differently, in ways that often make MBA/bean-counter types shudder:

• No "core" business; people can put lots of energy into finding the next big thing, rather than being forced to milk the last big thing. Just because a division makes medical equipment doesn't mean it can't make guitar strings (this actually happened at Gore, which now has a big share of the guitar-string market).
• No management layers and no organization chart: Gore wanted to get rid of the "facade of authoritarian hierarchy."
• No bosses, ranks, or titles; plenty of leaders. Leaders are chosen by their groups, "authority" only comes because people want to follow you. Leaders don't tell people what to do; instead they provide mentoring and support.
• "Lattice" management: "... a dense network of interpersonal connections where information can flow in all directions, unfiltered by an intermediary. In a lattice, you serve your peers, rather than a boss, and you don't have to work 'through channels' to collaborate with your colleagues."
• The CEO is voted in by the employees based on her being "someone they want to follow."
• Associates find the teams they want to work with, and who want to work with them.
• 10% time for "dabbling" (which has produced most of Gore's products). This long-predates Google's 20% time.
• Ideas are allowed to gestate until they either bear fruit or show themselves to be impractical.
• If you have an idea, you can recruit people to help work on it and create a team, by convincing them it's worth working on.
• An associate can say "no" to any request, but commitments are considered serious.
• Compensation is based on a comprehensive peer review; generally at least 20 peers are involved. No other ranking, degrees, experience, etc., has any impact on compensation; only the peer review.
• Every associate is a shareholder (the company is privately held), and there is profit-sharing.
• Plants are kept small so people know each other; grouped together regionally so that cross-pollination can occur. Financially this is far less efficient but the innovation benefits vastly overbalance any short-term costs.
• Projects regularly review themselves to make sure there are real opportunities, that the company can win in the marketplace, and that the product has enough value to be profitable.
• Gore has been included in every one of Fortune's "100 best companies to work for."
• The company has never had an annual loss.

This is as near as anything I've seen to "designing the company around the employees," and people apparently love working there, as much or more as people used to love working at Hewlett-Packard before that company got ruined by MBAs.

Thousands of executives have visited Gore over the years, but there's not much talk about the company, and very little emulation. Notice that lattice management takes all the power away; this is about as flat as it gets. I believe this demonstrates my argument that no one in power wants to change management when it means giving up power ... and those who believe it's worth giving up the power don't believe that anyone else is willing to do it.

Here's the corker: W. L. Gore & Associates was started in 1958, over 50 years ago! This was long before anyone thought we'd have the ubiquitous electronic communication necessary to make Shirky's thesis (electronic communication reduces or eliminates the need for hierarchy) possible. Despite this, they've been amazingly successful.

This experience is remarkably reminiscent of when I decided that the web should allow me to reinvent conferences. I began imagining all sorts of ways to redesign conferences around self-organization, using computing networks to collect all the information (speaker bios and abstracts, in particular) that we had collected by hand; allowing the attendees to vote for the talks they wanted to see rather than using an advisory board to choose them, etc. Basically, I was re-creating the same conference we had been putting on by hand, just using computers to do it. In hindsight, it was just like early "computer-based education," when people automated flash cards on the computer: you had the same experience as before, just on the computer. Then along came open spaces and turned everything upside down, not only eliminating the need for computers but eliminating eyes-forward speeches at the same time -- changing everything about the conference. W. L Gore did this for management while cheap computers and the internet were still science-fiction fantasies.

Is there something about the availability of computers and the internet that enable certain ideas like open spaces to evolve, even if those ideas don't rely on either technology for execution? Now that we have computers and the internet, is it possible that the ideas expressed in the organization of Gore can now propagate? I maintain that these ideas are too revolutionary to introduce to established industrial-hierarchy companies; no one in those organizations is going to voluntarily give up their power after fighting so hard for it. But new organizations have the choice -- and precedent, as evidenced by Gore -- to create flat, lattice-managed companies.