In this lesson, we will explore the concept of suboptimization. Perhaps you've heard of something called the law of unintended consequences. Sometimes we take actions to improve something, only to find that we've made something else worse. Think about the password policies that many of us have to conform to. These policies may require a password with a large number of characters, numbers, upper case letters, etc. The idea is to make these systems more secure, but the unintended consequence is that people write these passwords down, or store them in an electronic file, because they're too complex to remember. So, what is suboptimization? Dr. Deming describes this somewhat counterintuitive idea in his book, The New Economics. The idea is to make the overall system perform better, not components of that system. In fact, it may be necessary for some of those components to perform at a lower level than they're capable of. If we focus on optimizing the performance of the system components, it may actually hurt the overall performance of the system. [MUSIC] Think of the example of a team of rowers. [MUSIC] Each of these rowers is a world class athlete, but they're not all equal. If each rower were to do their individual best, that is maximize their efforts, the boat would probably zigzag and not more very efficiently. However, if each rower works in harmony with the others,the boat moves effortlessly across the river in a straight line. Sitting in the rear of the boat is a coxswain. Part of the coxswain's job is to coordinate the power and the rhythm of the rowers. The result is optimization of the system. Some rowers would obviously not work as hard as others, but the net result of everyone's efforts would be more than any one person's achievement. [MUSIC] A pack of wolves works as a system. Each has a role to play in order for them to hunt and bring down prey. But if an individual wolf wanted to optimize his own outcome, he might stand back and let the others do the hunting, and then come eat from their capture. But if wolves acted like that, they wouldn't capture any prey. And all of these wolves would starve. Let's look at some more practical example. In the book The Goal, Eli Goldratt gives the example of a decision to make machining centers in a factory more efficient, by increasing the amount of metal removed with each pass of the cutting tool. Instead of shaving a chip one millimeter thick, the process was changed so the tool removed three millimeters. This made the machining of the parts a much speedier process. And reduced the cost of operating machining centers. However, increasing the amount of metal taken off on each pass made the parts brittle, which necessitated a heat-treating operation. The increased load on the furnaces gave rise to a serious bottleneck and heat-treating. And made the plant significantly less productive and less profitable. Imagine a maintenance department. They might be directed to measure things such as cost of spare parts in inventory, cost of overtime, labor costs, and administrative cost. Their performance might be evaluated based on these measures. They're typical of cost measures that everyone manages. But what management behaviors might result? Well, in order to save the company money, the maintenance manager might look for ways of decreasing these costs. Perhaps by reducing the amount of spare parts in stock, or reducing overtime by spreading out preventive maintenance activities. This looks like a good thing. If every department lowers their cost, the company is better off, right? Of course, the problem is the department's performance is dependent on other departments. By saving costs on maintenance, the manager has caused the unplanned downtime for production machines to increase, as well as for the mean time to repair those machines to increase. These costs associated with production downtime, are much higher than the amount saved within the department. I once worked for a company in the south-eastern United States, that had a complex of three office buildings designed and built in the 1970s. These buildings were earth burned and designed so that the lighting and the body heat of employees was all that was needed. There was no heating system. It was a mild climate, so they were able to do this. The system worked very well. Later in the 1990s it was decided that we could save money by switching to more energy efficient fluorescent lights. So all of the bulbs in the build were replaced with more energy efficient fluorescent bulbs. Which because they're more energy efficient, give off less heat. As a result, employees started bringing space heaters in from home to put under their desks. And you can imagine what happened to the electric bill. In fact there were no savings. What we've seen are just a few examples of how suboptimization can hinder the performance of any complex system or organization. I suspect if you begin to think about this and look around you, you will see other examples as well. [SOUND]