Change Management

Recently I attended a course that covered the topic of “Change Management.” It is probably a sign of the times that we study how people deal with and react to change imposed by management. The course provided many useful insights into common reactions to change – but paid little attention to the necessity of the change itself. Of course change is inevitable, but Change Management is focused on “managementimposed” change – not the “Every Day is a New Day” flavor of change. The only reference to the necessity of change was simply “don’t change for the sake of change.” Or, in other words, change only when there is an obvious reason to change. And this statement (or perhaps not just a statement, but an algorithm in disguise) is where my operations research (O.R.) radar perked up.

To demonstrate what I mean, suppose there exists some objective way (e.g., a model or function) that measures how well our business processes are meeting our goals (market share, profit, etc.). We find ourselves, in a very general sense, with an optimization problem: which variables under our control should be changed and by how much? These variables include many business-oriented issues like “process effectiveness” and “organization.”

Consider the following algorithm:

Step 1: Select a one-time change in business practice.

Step 2: If the stated change can be reasonably expected to move the business closer to its goals, do it. (Read as: there is an obvious reason to change.)

Step 3: Go to Step 1.

This seems reasonable, and at the most basic level, it is – it certainly fits the requirement of “don’t change for the sake of change.” But in terms of an optimization methodology, we see we have only developed a greedy algorithm. From a purely theoretical point of view we know there are times when greedy algorithms perform well, even resulting in an optimal solution, but for complex (or most real-world) problems, the greedy algorithm often results in suboptimal solutions. Still, what are other options?

An intriguing option is Threshold Accepting [1]. The steps for this would be as follows:

Step 1: Select a one-time change in business practice.

Step 2: If the stated change seems reasonable, do it. However, if the stated change is unreasonable, still do it a certain percentage of time.

Step 3: Go to Step 1.

When your typical practitioner looks at the Threshold Accepting algorithm in the context of, say, determining a good lumber-harvesting plan, it is easy to see the usefulness. Some un-improving moves in the short term are the most productive in the long term. However, in the context of business policy and business organization, the attitude changes quickly to “don’t change for the sake of change.”

So here we have an interesting dynamic. When the top-boss enters and announces a change that seems certain to move the business away from its goals – our initial reaction is… well… one that requires the manager to take courses in Change Management. Yet optimization theory suggests we should be more open to un-improving changes with the hope of diagnosing and implementing other more-improving changes down the road. Certainly more thought needs to be put into the change – as business changes can have large impacts on the people and the bottom line. Still, perhaps there is sound O.R. reasoning not to scoff too much at “change for the sake of change.” It could be a step back that allows us to see the very large step forward.

REFERENCES

1. Dueck, G., Scheuer, T., Sept 1990, “Threshold accepting: a general-purpose optimization algorithm appearing superior to simulated annealing,” Journal of Computational Physics, Vol. 90, No. 1.