Eliyahu M. Goldratt and Jeff Cox – The Goal

If the goal is to make money, then (putting it in terms Jonah might have used), an action that moves us toward making money is productive. And an action that takes away from making money is non-productive.

Can I assume that making people work and making money are the same thing?

So this is the goal: To make money by increasing net profit, while simultaneously increasing return on investment, and simultaneously increasing cash flow.

“They’re measurements which express the goal of making money perfectly well, but which also permit you to develop operational rules for running your plant,” he says. “There are three of them. Their names are throughput, inventory and operational expense.”

“Throughput,” he says, “is the rate at which the system generates money through sales.”

“Inventory is all the money that the system has invested in purchasing things which it intends to sell.”

Operational expense is all the money the system spends in order to turn inventory into throughput.”

“Just remember we are always talking about the organization as a whole—not about the manufacturing department, or about one plant, or about one department within the plant. We are not concerned with local optimums.”

So the way to express the goal is this? Increase throughput while simultaneously reducing both inventory and operating expense.

“Throughput is the money coming in. Inventory is the money currently inside the system. And operational expense is the money we have to pay out to make throughput happen. One measurement for the incoming money, one for the money still stuck inside, and one for the money going out.”

“Any money we’ve lost is operational expense; any investment that we can sell is inventory.”

If it’s knowledge, say, which gives us a new manufacturing process, something that helps turn inventory into throughput, then the knowledge is operational expense. If we intend to sell the knowledge, as in the case of a patent or a technology license, then it’s inventory. But if the knowledge pertains to a product which UniCo itself will build, it’s like a machine—an investment to make money which will depreciate in value as time goes on. And, again, the investment that can be sold is inventory; the depreciation is operational expense.

“A plant in which everyone is working all the time is very inefficient.”

“Alex, the goal is not to reduce operational expense by itself. The goal is not to improve one measurement in isolation. The goal is to reduce operational expense and reduce inventory while simultaneously increasing throughput,” says Jonah.

“One phenomenon is called ‘dependent events.’ Do you know what I mean by that term? I mean that an event, or a series of events, must take place before another can begin … the subsequent event depends upon the ones prior to

big deal occurs when dependent events are in combination with another phenomenon called ‘statistical fluctuations,

These types of information vary from one instance to the next. They are subject to statistical fluctuations.”

“Alex, if I simply told you what to do, ultimately you would fail. You have to gain the understanding for yourself in order to make the rules work,”

Ron is setting the pace. Every time someone moves slower than Ron, the line lengthens. It wouldn’t even have to be as obvious as when Dave slowed down. If one of the boys takes a step that’s half an inch shorter than the one Ron took, the length of the whole line could be affected.

Once I’ve closed the gap between us, I can’t go any faster than the rate at which the kid in front of me is going. And he ultimately can’t go any faster than the kid in front of him. And so on up the line to Ron. Which means that, except for Ron, each of our speeds depends upon the speeds of those in front of us in the line.

What’s happening isn’t an averaging out of the fluctuations in our various speeds, but an accumulation of the fluctuations. And mostly it’s an accumulation of slowness—because dependency limits the opportunities for higher fluctuations.

You can look at it this way, too: Herbie is advancing at his own speed, which happens to be slower than my potential speed. But because of dependency, my maximum speed is the rate at which Herbie is walking. My rate is throughput. Herbie’s rate governs mine. So Herbie really is determining the maximum throughput.

Everybody stays together behind Herbie. I’ve gone to the back of the line so I can keep tabs, and I keep waiting for the gaps to appear, but they don’t.

“Listen, if you guys want to go faster, then you have to figure out a way to let Herbie go faster,” I tell them.

“What you have to do next, Alex, is distinguish between two types of resources in your plant. One type is what I call a bottleneck resource. The other is, very simply, a non-bottleneck resource.”

“A bottleneck,” Jonah continues, “is any resource whose capacity is equal to or less than the demand placed upon it. And a non-bottleneck is any resource whose capacity is greater than the demand placed on it.

“Yes, but as you already know, you should not balance capacity with demand. What you need to do instead is balance the flow of product through the plant with demand from the market. This, in fact, is the first of nine rules that express the relationships between bottlenecks and non-bottlenecks and how you should manage your plant.

Speaking fundamentally, the bottleneck flow should be on a par with demand.”

“No, bottlenecks are not necessarily bad—or good,” says Jonah, “they are simply a reality. What I am suggesting is that where they exist, you must then use them to control the flow through the system and into the market.”

“Don’t you see?” I ask them. “If we’ve got a Herbie, it’s probably going to have a huge pile of work-in-process sitting in front of it.”

“Okay, if we can’t do anything to change their position in the sequence, then maybe we can increase their capacities. We’ll make them into non-bottlenecks.

“Very perceptive of you. Make sure the bottleneck works only on good parts by weeding out the ones that are defective. If you scrap a part before it reaches the bottleneck, all you have lost is a scrapped part. But if you scrap the part after it’s passed the bottleneck, you have lost time that cannot be recovered.”

“Whatever the bottlenecks produce in an hour is the equivalent of what the plant produces in an hour. So … an hour lost at a bottleneck is an hour lost for the entire system.”

“First, make sure the bottlenecks’ time is not wasted,” he says. “How is the time of a bottleneck wasted? One way is for it to be sitting idle during a lunch break. Another is for it to be processing parts which are already defective—or which will become defective through a careless worker or poor process control. A third way to waste a bottleneck’s time is to make it work on parts you don’t need.”Read more at location 3012   • Delete this highlight

Note: increase bottenevk capacity any way you can Edit

make the bottlenecks work only on what will contribute to throughput today… not nine months from now,”

The other way you increase bottleneck capacity is to take some of the load off the bottlenecks and give it to non-bottlenecks.”

So by continuing to run the non-bottleneck parts, this guy was actually interfering with our ability to get an order out the door and make money.” “But he didn’t know any better,” says Bob. “Exactly. He couldn’t distinguish between an important batch of parts and an unimportant one,” I say. “Why not?” “Nobody told him.” “Until you came along,” I say. “But you can’t be everywhere, and this same kind of thing is going to happen again. So how do we communicate to everybody in the plant which parts are important?” “I guess we need some kind of system,” says Bob.

With the bottlenecks more productive now, our throughput has gone up and our backlog is declining. But making the bottlenecks more productive has put more demand on the other work centers. If the demand on another work center has gone above one hundred percent, then we’ve created a new bottleneck.

“And that is the problem,” says Jonah. “Because what happens to those extra hours of production from Y? Well, that inventory has to go somewhere. Y is faster than X. And by keeping Y active, the flow of parts to X must be greater than the flow of parts leaving X. Which means …” He walks over to the work-in-process mountain and makes a sweeping gesture. “You end up with all this in front of the X machine,” he says. “And when you’re pushing in more material than the system can convert into throughput, what are you getting?” “Excess inventory,” says Stacey.

By definition, Y has excess capacity,” says Jonah. “So if you work Y to the maximum, you once again get excess inventory. And this time you end up, not with excess work-inprocess, but with excess finished goods. The constraint here is not in production. The constraint is marketing’s ability to sell.”

Stacey points out immediately that in no case does Y ever determine throughput for the system. Whenever it’s possible to activate Y above the level of X, doing so results only in excess inventory, not in greater throughput. “Yes, and if we follow that thought to a logical conclusion,” says Jonah, “we can form a simple rule which will be true in every case: the level of utilization of a non-bottleneck is not determined by its own potential, but by some other constraint in the system.”

The numbers are meaningless unless they are based upon the constraints of the system. With enough raw materials, you can keep one worker busy from now until retirement. But should you do it? Not if you want to make money.”

“And the implication of these rules is that we must not seek to optimize every resource in the system,” says Jonah. “A system of local optimums is not an optimum system at all; it is a very inefficient system.”

“What we really have to do is just keep the kid at the front of the line from walking faster than Herbie. If we can do that, then everybody will stay together.

What you have to do is find a way to release the material for the red parts according to the rate at which the bottlenecks need material—and strictly at that rate.”

If Ralph can determine a schedule for releasing red-tag materials based on the bottlenecks, he can also determine a schedule for final assembly. Once he knows when the bottleneck parts will reach final assembly, he can calculate backwards and determine the release of the non-bottleneck materials along each of their routes. In this way, the bottlenecks will be determining the release of all the materials in the plant.

“If we cut our batch sizes in half, then I guess that at any one time we’d have half the workin-process on the floor. I guess that means we’d only need half the investment in work-in-process to keep the plant working. If we could work it out with our vendors, we could conceivably cut all our inventories in half, and by cutting our inventories in half, we reduce the amount of cash tied up at any one time, which eases the pressure on cash flow.”

If we reduce batch sizes by half, we also reduce by half the time it will take to process a batch. That means we reduce queue and wait by half as well. Reduce those by half, and we reduce by about half the total time parts spend in the plant. Reduce the time parts spend in the plant, and…. “Our total lead time condenses,” I explain. “And with less time spent sitting in a pile, the speed of the flow of parts increases.”

It’s perfectly okay to have more setups on non-bottlenecks, because all we’re doing is cutting into time the machines would spend being idle.

STEP 1. Identify the system’s bottlenecks. (After all it wasn’t too difficult to identify the oven and the NCX10 as the bottlenecks of the plant.) STEP 2. Decide how to exploit the bottlenecks. (That was fun. Realizing that those machines should not take a lunch break, etc.) STEP 3. Subordinate everything else to the above decision. (Making sure that everything marches to the tune of the constraints. The red and green tags.) STEP 4. Elevate the system’s bottlenecks. (Bringing back the old Zmegma, switching back to old, less “effective” routings….) STEP 5. If, in a previous step, a bottleneck has been broken go back to step 1.


In essence the Kanban system directs each work center when and what to produce but, more importantly, it directs when not to produce.

Adhering to the flow concept mandates the abolishment of local efficiencies.

Rather, he insisted that the setups required are not cast in stone, that the processes can be modified to drastically reduce the setup time required. He led the efforts to develop and implement setup reduction techniques that eventually reduced all setup times in Toyota to be, at most, just a few minutes.10 It is no wonder that Lean is now strongly associated with small batches and setup reduction techniques.

At the initiation of the Kanban system, to achieve reasonable throughput, Ohno had to start with many containers each holding a non-negligible quantity of a particular part. Gradually, Ohno reduced the number of containers and then the quantities in each container. If the flow was not noticeably disturbed, then the reduction of the number of containers and quantities per container continued. When the flow was disturbed the Five Why’s method was used to pinpoint the root cause. It had to be fixed before the quantities could be further reduced. It took time but the end result was a remarkable improvement in productivity.

In summary, both Ford and Ohno followed four concepts (from now on we’ll refer to them as the concepts of flow): Improving flow (or equivalently lead time) is a primary objective of operations. This primary objective should be translated into a practical mechanism that guides the operation when not to produce (preventsoverproduction). Ford used space; Ohno used inventory. Local efficiencies must be abolished. A focusing process to balance flow must be in place. Ford used direct observation. Ohno used the gradual reduction of the number of containers and then gradual reduction of parts per container.

Since the Kanban system takes time to implement, its assumption is that the environment is relatively stable—that the processes and the products do not change significantly for a considerable length of time.

A second aspect of the stability required by TPS is stability in demand over time per product.

But, the most demanding aspect of the stability required by TPS is stability in total load placed by the orders on the various types of resources.

A good starting point for improving flow will be to choose the time buffer to be equal to half the current lead time; such a choice will ensure that the company will find itself somewhere on the plateau of the graph.

Per batch, the time that has passed since its release is tracked. If less than one third of the time buffer has passed the priority color is green, if more than one-third but less than two-thirds the priority color is yellow, if more than two thirds the color is red, if the due date has passed the color is black. Blacks have higher priority than reds, etc. If two batches have the same color, to try and decide which one should be worked on first is an excellent example of trying to be more accurate than the noise.

It is essential to strengthen the tie between sales and operations—that is the real challenge. A system must be put in place to ensure that every due-date commitment is given only according to the yet unallocated capacity of the bottleneck.