Accomplished Lean practitioners will tell you that Lean is ultimately a way of thinking. Training yourself to think in terms of a system makes Lean vastly easier to understand and implement.

When I began studying the works of Edwards Deming I was puzzled by explanations of systems. I couldn’t quite wrap my head around what Deming and other authors were writing. I finally came to believe that they all made the subject far too obtuse and complex. I’m convinced the fundamentals are just not that hard to understand. Let’s compare.

In “The New Economics” Deming defined a system as “a network of interdependent components that work together to try to accomplish the aim of the system.” This is a fairly standard definition of a system.

With all due respect to Dr. Deming (and the respect due Dr. Deming is great), this definition and ones like it raise more questions than they answer. And these are questions with complex answers — questions about networks, about interdependence, about the meaning of “work together”.

But the most important questions these definitions raise are basic questions about the definitions themselves: Why does a system only “try to accomplish the aim of the system”? Why does it not actually accomplish the aim of the system?

And what if the system actually accomplishes its aim? What then? Having accomplished its aim, it need no longer try. No longer needing to try, it no longer matches the definition of a system. Is it suddenly no longer a system because it did what the system was supposed to do? That doesn’t make sense.

The problem with these sorts of definitions is that they aren’t terribly robust. They can’t answer fairly obvious questions so, in my opinion, they offer little in terms of practical insights for real systems.

But everyday we deal with real systems, systems of flesh and blood, concrete and steel, paper and electronics. The outcome of a real system may not be exactly the outcome we want, but it is an outcome nevertheless, and it’s usually fairly predictable.

This all leads me to believe we need a better definition, a definition of a real system. Here’s mine:

A system is a set of components that interact to produce a particular outcome.

I find this definition much more useful.

It asserts what ought to be obvious: the particular outcome of the system is whatever the particular outcome is. If the particular outcome is unexpected, then the system you actually have isn’t the system you thought you had. If the particular outcome isn’t what you had in mind, you need a different system. Not necessarily hugely different, but certainly different.

This is a much more robust definition in terms of Lean. Looking back at this definition — a system is a set of components that interact to produce a particular outcome — it should be clear that there are only three things you can do to change the outcome of the system: modify the components, modify the interactions, or modify both.

This modifying of components and modifying of interactions to change outcomes is what Lean, Deming’s ideas, and Toyota’s methods are all about. Value stream mapping, statistics, the PDSA loops, the improvement kata, the five-whys, all are designed to help us discover, in greater detail, the nature of the the components and interactions that make up a system. The more we know about the nature of a component or interaction the more predictably we can improve the outcome of our system.

This is why understanding systems is so important to the Lean practitioner. Without an ability to think in terms of the system, any changes you make will be essentially random, instead of ordered, and you won’t be able to predictably improve outcomes. It all starts with having the right, robust definition.

Copyright 2014 by Paul G. Spring. All rights reserved.