Thursday, November 01, 2007

HBR article demonstrates that leaders need to manage complexity

"We need to document our processes!"

I heard this again and again at various companies I worked at over the years. And that's a fine goal, to document processes. But the thinking--that if processes are documented then we will be able to perform high-quality work and be successful--is flat-out wrong in many circumstances.

Why? Because many (and many of the most important) business problems can't be reduced to a repeatable process. This view is described in an article in the November Harvard Business Review, "A Leader's Framework for Decision-Making," by Dave Snowden and Mary Boone (link - $$). (Prior references to Dave Snowden's work can be found here: 1, 2, 3.)

In it, Snowden and Boone describe the Cynefin framework, a model that helps put business situations into a context that guides how they should be addressed. The framework has four primary segments:

Simple - repeatable processes that can be described by best practices (e.g., how to determine whether a mortgage applicant is qualified)

Complicated - "the domain of experts," according to Snowden and Boone; where complete data is available, and issues can be solved with analysis (e.g., finding underground oil deposits)

Complex - where multiple variables interact unpredictably - "the realm of 'unknown unknowns,' ...the domain to which much of contemporary business has shifted."

Chaotic - where no manageable patterns exist, "the realm of unknowables" --e.g., September 11, 2001. In this case, the best response is to do something and assess what happens.

So, back to documenting processes. Simple processes and their best practice should be documented and followed. Complicated processes, too, can benefit from discipline, though there is value in dissent and dialogue. Documenting complex processes doesn't do much of value--repeatability is impossible and in fact counterproductive to attempt.

Here are some business processes that would fall into the complex domain:
  • new product development (how people learn about and use products can have a significant effect on how the product evolves)
  • entering a new market or geography
  • making an organizational change
  • a B2B sales pursuit
So how to manage these if they can't be boiled down to a cookbook? Boone and Snowden recommend involving more people in decisionmaking (sounds a bit democratic); setting some rules or guidelines to channel behavior (i.e., in a sales pursuit, we will never respond to a tender that we didn't know was coming); encouraging dissent; creating an environment where good things can emerge, and nurturing those things.

In my experience, managers are still trying to shoehorn all their business problems into the simple or complicated domains. The more quickly they accept the complexity of many critical areas, and manage them appropriately, the sooner we'll stop wasting human resources and start achieving better business results.

And that'll be something worth documenting.

(graphic: the Cynefin framework from Cognitive Edge via Wikipedia)

, , , , , , , ,

1 comments:

Michelle Malay Carter said...

John,

I agree with what you are saying here. Stratified systems theory, aka Jaques' Requisite Organization, takes this idea of complexity and codifies it. It divides work into distinct layers of complexity. Then, it provides a complimentary piece of information regarding human capability to solve problems at the various layers. Match human capability with its corresponding level of role complexity and watch the pieces fall into place within organizations.

To get the flavor of how this works in the real world, you can read this case study: http://www.peoplefit.com/downloads/Specialty_Chemicals_Success-pf.pdf

Regards,

Michelle