Reducing Complexity to Manage Better

Emma Stewart writes today in a Harvard Business Publishing blog about the futility of Scenario Planning exercises, saying that

.. if you want to avoid fixes that fail and move towards sustainability in today’s complex operating environment, I invite you to resist the instinct to resort to scenario exercises, or only do so in conjunction with systems thinking approaches.

I can see her point, but the elaborate systems thinking approach she is referring to does exist already – generally embedded deep into the fabric of any successful organization. The organizational processes and systems are designed and maintained to weather the ‘normal’ levels of complexity. Scenario Planning, on the other hand, tries to peer into the future in order to develop some decision making criteria as and when needed.

Complexity in business and its operating environment is a given constraint in management. Management challenges exist along two completely opposite paradigms.

At one end we need to delve deep into existing complexity – define it and automate the processes and decisions there – so that we are ‘shielded’ from the known complexity or the static view. The system or the model works just fine within the frame of its original assumptions. Taken to an extreme, all possible complexity can be modeled given enough time and data.

On the other end is the unknown complexity introduced as soon as we recognize that we live in a real world of dynamic, moving components that can interact with each other in unpredictable ways to create new scenarios. This is where we need to abstract the complexity into simple models like the famous (or infamous) 2×2 matrices. We need to make empirical decisions quickly because the environment is changing at the same time. Decision models need to be simple and adequate – a criteria difficult to achieve, but still better than trying to stop the world while we analyze our way into paralysis.

Good management needs to manage complexity by instituting structure – enough to be affordable – and by developing abstracted, simple decision models – enough to be useful.