Simpler Organizations Are Harder To Design

Even the most deft leaders using the latest Web 2.0 tools can't transform organizational hierarchies overnight.

Venkatesh Rao, Contributor

May 11, 2011

4 Min Read

My favorite example of the complexity of organization design is from Alfred Thayer Mahan's classic, "The Influence Of Sea Power Upon History."

Until about the 12th century, naval warfare was galley warfare. Galleys were powered by primitive sails and oarsmen, and naval warfare used line-abreast formations. Lines of ships would face off and then charge ahead, trying to ram each other. If a galley rammed an enemy galley, a hand-to-hand melee would follow.

Between the 13th and 17th centuries, advanced sailing techniques allowed much larger ships to sail efficiently against the wind. These larger ships were equipped with cannons (more than a hundred on multiple decks at the peak of sail warfare), and they dispensed with oarsmen.

As a result of these innovations, naval strategy shifted from line-abreast formations to line-ahead formations. Ships would line up stern-to-bow, allowing their cannons to be pointed at the enemy line.

A basic change, right? The organizational structure didn't even change its fundamental shape. It simply went from one kind of straight line to another.

So how long do you think this shift from line-abreast warfare to line-ahead warfare took? By Mahan's estimate, about 30 years after all the enabling technologies were in place.

If that shocks you, it means you don't understand the scope of organization design.

The Scope Of The Problem

If a simple change from one line formation to another based on just three point innovations took 30 years, how long do you think a fuzzy shift from "hierarchical" to "network," driven by dozens of simultaneous innovations (the entire 2.0 technology set), is going to take?

The reason things take so long is that the devil is in the hidden details. Here's just a sampler of the details that matter in thinking about hierarchies:

-- Span of control: the well studied empirical idea that an individual can manage between three to eight direct reports, depending on the level of autonomy.

-- Decision-making speed: an aligned hierarchy where the lower levels aren't rebelling can understand and act on ground-level information exponentially faster than an arbitrary network.

-- Abstraction: Authority hierarchies naturally induce abstraction hierarchies: Higher levels get information at different scope/resolution levels and develop different kinds of situation awareness.

Generalizing, organization design is a problem with a vast scope. It involves information structure (who knows what, when), status dynamics (which arise from our genes), information flow patterns, financial versus social incentives, the speed at which the domain is changing, and many other factors.

The point is, all this has to be studied, experimented with, and thoughtfully considered in coming up with a design and a process to realize that design.

Here, too, the naive make simplistic assumptions that "emergent" designs are always better than "top down, imposed" designs. The biggest irony is that hierarchies are in fact networks themselves, in mathematical terms. It takes quite a bit of work to come up with realistic examples of networks that don't behave like hierarchies. A famous example is Arrow's Paradox. Kenneth Arrow, who studied the paradox in his PhD thesis, went on to win a Nobel Prize for his work on social choice theory.

That's how hard organization design is. Nobel Prize hard.

Simpler Isn't Easier

Let me conclude by countering a common reaction to this sort of analysis: "All that complexity is unnecessary. With the right tools, you don't need any of it and it's all simple. We can dismantle the whole useless structure."

Are 2.0 organization designs fundamentally simpler? Absolutely.

Do 2.0 models rely on "pull" more than "push," as John Hagel has said? Yes.

Are 2.0 models less waterfall and more agile? Right again.

Should they be grown via orchestration instead of being designed dictatorially? In most cases, yes.

Does that mean they're easier to create?

Absolutely not.

So how do you actually achieve all these beautifully simple states? Stay tuned -- we'll explore that in a future column.

Venkatesh Rao is a writer and independent researcher at and the author of Tempo. He can be contacted via LinkedIn.

Businesses have myriad technology options for pulling together people and ideas. But getting it right still isn't easy. Also in the new all-digital issue of InformationWeek SMB: A UC champion's survival guide. Download it now. (Free registration required.)

About the Author(s)

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights