David Gray on Russ Ackoff

David Gray wrote a great article in a funny place, as a comment to a really great YouTube video.  

I took it upon myself to reproduce it here because you need to read it and I do too.

Dividing work.

Division of labor, as Adam Smith pointed out in the 1700s, has the potential to increase productivity. But division of labor also leads to interdependency: every worker relies more heavily on others to do her job, and as the number of handoffs increases, so does the potential for dropped balls. As the number of divisions grows, so grows the interdependence.  

This interdependence creates a need to synchronize and coordinate the work.
Traditionally this has been the job of management and bureaucracy. We coordinate the work through measurement and control.  

As you increase the number of divisions, you also increase complexity. One of the goals of these divided and interdependent systems is making work more efficient and systems more consistent, predictable and reliable. The goal in many cases has been a perfect system; a system that’s idiot-proof. If something can be automated, you automate it. If you can’t automate it, you constrain it to the minimum possible variation. 

Of course, the more idiot-proof the system, the more you will constrain behavior, forcing people in many cases to act like idiots, against their better judgment. Even when people know there’s a better way to do something, they will often be constrained by policies and procedures that were designed to reduce variety in the system. If your system needs to solve problems that you can’t anticipate, then you have a problem, because automated systems and idiots can’t solve problems. 

In addition, although dividing work may make the system more efficient, by dividing work into ever-more specialized tasks, we also disconnect people from the meaning and purpose of what they are doing. From their small, constrained box, people can’t see the big picture so they must make decisions and act within a very limited and constrained perspective.

Standardized, interchangeable parts.

Another core idea from the industrial revolution is the concept of, interchangeable parts. Standardization does make it easier to mass-produce quality products. Standards also make it easier to connect things. 

For example, having a standard for electrical sockets assures that you can plug in any appliance, and having a single standard for web links makes it easy for any web page to link to any other. 

We run into problems, though, when we try to apply standards to things that inherently have a high degree of variety: for example, a customer service call. Customer problems come in all shapes and sizes, and even problems that might seem very similar on the surface can be subject to a lot of variability based on the context. 

We have gotten so used to the idea of standards as a good thing that we tend to apply them in the wrong places. For example, consider the idea of a “best practice.” The concept of a best practice assumes that there is one “best way” to solve a problem: that every problem can be isolated from its context, and a single best way of solving it can be described and shared. Unfortunately, this has caused a lot of problems in the business world, because it’s impossible to isolate problems from their context. 

A system is not just the sum of its parts. What makes a system work is not the parts in isolation, but the interactions between them, and the inherent tradeoffs that must be made to achieve different kinds of system performance. Standardization is something you apply to the parts of a system, not a whole. A best practice from one company, or from one part of a company, cannot necessarily be applied successfully elsewhere. 

Systems expert Russell Ackoff points out that “If we have a system of improvement that is directed at the parts, taken separately, you can be absolutely sure that the performance of the whole will not be improved.” 

Ackoff illustrates his point with the following example:
“I read in the New York Times recently that 487 kinds of automobiles are available in the United States. Let’s buy one of each and bring them into a large garage. Let’s then hire 200 of the best automotive engineers in the world and ask them to determine which car has the best engine. Suppose they come back and say the Rolls Royce has the best engine. Make a note of it. “Which one has the best transmission?” we ask them, and they go over and test and they come back and say the Mercedes does. “Which one has the best battery?” They come back and say the Buick does. And one by one, for every part required for an automobile, they tell us which is the best one available. Now we take that list, give it back to them and say, “Now remove those parts from those cars, and put them together into the best possible automobile, because now we’ll have an automobile consisting of all the best parts. What do we get? You don’t even get an automobile, for the obvious reason that the parts don’t fit. The performance of a system depends on how the parts fit, not how they act taken separately.” 

What’s true for the fit of the parts of a system is also true of the fit between a service and the context within which the service is delivered. Every interaction with a customer is different: sometimes in subtle ways, and sometimes in profound ways. Two interactions that look similar on the surface may be dramatically different, in ways that are hard to predict. 

Consider a customer walking up to an airline counter who has already been bumped from two flights and whose luggage has been lost. These previous interactions will have a major effect on the context of the current one. 

Managers think that standardization is a good thing to do. If we standardize, goes the idea, our costs will go down. But if your interactions are highly variable, as most service interactions are, then the opposite will happen. Attempts to standardize the work will make costs go up, not down. This is because standardizing the work reduces the ability of your system to absorb variety. We try to cage variety into nice neat swim lanes: For example, voice menus in an automated voice system. But when there is a lot of variety in your environment, these kinds of control systems are exactly the way to make things not work.