Waste, Want, and the failure of the Toyota Production System

The fundamental genius of the Toyota Production System is bringing the focus on the customer into consideration for any and every process. The Toyota Production System does not always have the most convenient tool to directly address a particular problem, but many other methodologies lack the fundamental centering on the customer. Six Sigma is a set of tools to reduce variation, for example, but it does not intrinsically ask you whether the product you reliably produce is what your customer wants. Six Sigma will cheerfully coexist with the waste of over-processing.

When the Toyota Production System was first being hailed as the panacea for American manufacturing, the common stereotype was that American manufacturing was designed to produce as much as possible, regardless of whether anyone wanted it. American manufacturing managers care about machine utilization – the more, the better! And this attitude has not entirely gone away; or rather, it has largely not gone away at all, and only gone underground. This Monday a friend told me that his father, and old-timer at a local factory, preferred working on the third shift, away from all the nonsense about Toyota production, where he could produce three times as much as the first shift.

People of all occupations speak in short-hand, so for a man to boast of his production quantity does not necessarily prove he is overproducing, or producing with any more defects than his first-shift neighbor. And those on the first shift, where “Toyota production” is blamed for slowing things down, may be trying to drape Toyota-inspired charts and graphs over the same attitudes and goals and incentives as they’ve always had—which can only result in more inefficiency, not less. “Overproduction” shifts from being solely a problem of the commercial product and becomes the production of charts that nobody wants, needs, uses or cares about, charts which are always updated when a manager from Corporate is scheduled to visit but never updated when there’s a push on to hit the end of month production target. But someone involved in this story has definitely lost track of what Toyota Production System is fundamentally about.

I was once involved with a major event, featuring a Genuine Actual Real Toyota Engineer, meant to apply Toyota Production principles to one of the production lines in a century-old factory that had second and third generation employees. One of these old salts, Tom, was responsible for scheduling production on the line—not his first job at the company, and not his first year in that job.

Tom didn’t take very kindly to being told how to do his job, but he didn’t rely on the strength of his own opinions. He got data – the orders for the product that ran on that line – and he charted it. He printed the chart on the plotter so that it was three feet tall, so that he could show on the chart the enormous spike in demand he’d recently dealt with. It was quite out of the bounds of the typical run rate, and made a terrific illustration.

The Genuine Actual Real Toyota Engineer from Japan thought that it must be a stock order, and through his interpreter he asked about that. “No,” Tom said, “it’s not an order for stock; it is an order from a customer, a shippable order. If we ship it we can invoice for it and get paid.” The Engineer was amazed, shaking his head and muttering “Very difficult, very difficult.”

As it happens, they were both right. As Tom said, it was a shippable, invoiceable order, and Tom’s boss, and his boss, and all the bosses all the way up the multinational diversified corporate ladder wanted him to build, ship, and recognize revenue, as soon as possible but at least in time to be in the next quarterly investor update. But it was stock, too; stock to be held by a distributor, stock that would not tie up our capital but would sit in the “customer’s” warehouse for months being gradually sold off; stock purchased at a pretty discount because it allowed a sales person to save the quarter.

American business managers caught on quickly to Toyota’s anti-inventory message, because inventory ties up capital and that makes investors fidget. Inventory is bad! Down with inventory!

But the Toyota Production System also emphasizes takt time, which is to say the amount of time it takes to complete a step in the manufacturing process – or better put, the amount of time it takes to complete every step in the manufacturing process, so that the whole entire process moves ahead in synch, like a clock, to the tempo of the takt time. To conventional thinking, this is arbitrary and silly; some days you have to work harder and faster than others, that’s just the way it is. Slowing everything down so it will look pretty and synchronized makes as much sense as telling soldiers who are being shot at to walk slowly and not muss their uniforms.

Who, though, is the enemy? Under the Toyota Production System, the enemy is waste in all of its forms, including producing more than is wanted or producing defective products that are not wanted. Defects are unexpected, of course, and unexpected things happen when there is a lot of change, a lot of variation, too much going on to keep track of it all. If we know exactly how long a step in the process should take, than we know that if it took more or less time something changed. The change could be a good thing, increased efficiency, in which case we should learn from it and apply that improvement wherever we can, but it could also be a bad thing: a missed step, a defect.

The pell-mell rush to ship everything possible in time to invoice it in this fiscal quarter is the right approach if this quarter is the only quarter that matters, if we never have to improve our quality or productivity or efficiency, if God is going to stop time and give us all eternal rewards based on our production numbers this month. But, if you take a slightly longer view, and consider it more important that you know how to get better, then takt time becomes important because it is yet another way in which variation—and the likelihood of defects, of time and effort spent on unwanted results—becomes visible. If those metaphoric bullets are coming from the CEO and the shareholders toward anyone who didn’t help make the profit target for the month, takt time will not save you. If the enemy is waste, then takt time is not a silly ineffectual dress rehearsal; it is the scrupulous elimination of any possibility that the enemy could be hidden from view.

The Genuine Actual Real Toyota Engineer wanted Tom to run the production line at the same speed (takt time), producing at the same steady level (Heijunka). Tom proved it was impossible to do that given the variability in the customer demand, so the Genuine Actual Real Toyota Engineer was stumped. He couldn’t tell Tom to disregard the customer demand as that would violate the fundamental principles of the Toyota Production System.

But what did that customer really want? As a distributor, they were not going to put the product immediately to use if it all showed up on one day. They would be reselling gradually over time. The customer wanted the discount they were able to wrangle out of their sales representative by placing such a larger order. With the discount in hand, they would have been perfectly happy to receive the products at a steady pace over a much longer period of time.

What did the sales rep want? He wanted the commission from booking the deal. It was a large enough deal that probably his manager had promised him some special extra compensation if he landed it. Nothing in the incentives or instructions offered to that sales rep made it his responsibility to learn when the customer actually needed to have the product in hand, so nothing in the incentives offered to the customer encouraged them to spread the deliveries out over time.

Nobody in the room that day with Tom – nobody in the building—had the authority to fix the problem. The root cause of that problematic order spike was that the sales rep did not know what his customer wanted, and in fact did not even know who his customer was.

“The Customer Is Always Right!” has become so hackneyed that it is hard to find it in use today without irony. But it is not wrong; it is right enough as far as it goes. It is simply insufficient, and fails to properly define who the customer is. The customer is the one who is willing to pay for the product you offer. If you run a hair salon and someone calls asking for pizza, that person is not your customer. If you run a car factory and someone calls asking for pizza, that person is not your customer. And if you run a car factory that can produce a car in four days and someone calls asking for a car in four seconds, that person is not your customer.

When someone with money asks for something you cannot deliver, that person is not your customer.

There are potential customers who will ask for something which may be difficult to deliver, and then the business managers must decide if it is worth the risk of failure and the extra cost to satisfy the challenging delivery; they must decide whether to accept the customer. A business that never adapts to meet new requests will not change as its customers change, and eventually all the customers it once had will no longer exist. But money on offer does not by itself define a customer. It must be matched to a capability to provide.

A capability for which nobody will pay is not good business (it is overproduction), but payment for something you are not capable of providing is also not good business (the result will be defective).

The order that the sales rep booked for an excessive quantity of product from Tom’s line could not be met by the date on the order. Since the capacity to provide was not available, the order should have been refused, or negotiated to something achievable.

But the order that was booked did not adequately represent what the customer wanted to pay for. The customer understood that the order would not likely all arrive by the date requested, but there was no penalty, surcharge or inconvenience for requesting it for immediate delivery. If it had been delivered all at once it would probably have caused a storage problem in the distributor’s warehouse (although very likely there was no penalty or negative feedback that would reach the person placing the order, any more than for the sales rep). The low price that the buyer wanted would have no adverse effect on the factory operations, while the short delivery time adversely affected a large number of other customers. The customer who booked the larger order would not have been willing to pay a surcharge for the harm to the rest of the factory’s business, but, being offered the convenience without any consequence, they were happy to take it.

“Customers” will take many things for free, but true customers pay for what they get. This crucial distinction is often hard to keep in view during new product development; Henry Ford is reported to have said that his customers would only have asked for a faster horse before he gave them the choice of a car. Those customers are using “horse” as a word to mean “fast, reliable, low-cost transportation,” and eventually “car” became a better term for that than “horse.”

When I was a product manager in a forklift company we had an inside joke that if you asked a forklift user what he wanted, he would tell you he wanted a radio and a place for his drink; and if you asked the forklift operator’s boss he would tell you that operators are not allowed drinks or radios while they work. The customer, as it turns out, is not the person who drives the forklift, but the person who pays for it. A certain amount of comfort for the operator reduces operator fatigue, improves productivity, and is good for everyone. But a forklift operator’s supervisor wants pallets to be moved in a safe, clean, efficient and attentive manner, and a forklift operator wants to watch football and forget he is at work.

It turns out, though, that forklift operator supervisors are not much more help. They will tell you they want the forklift to break less; but that is, like a faster horse, a trifle obvious, and if your warranty and service data is any good you already know what is breaking and don’t need to be told by the supervisor.

While the supervisor is telling you that he would like the forklifts to break down less often, however, you might notice that the operators are standing idly by their trucks, waiting for the third-party computer system to boot back up. It takes a while to boot up but must do so after every break, because while on break the operator unplugs the forklift from its only power source (a giant lead acid battery) in order to recharge it. If you could eliminate the need to unplug the battery (and power off the computer) in order to charge it, you would save the forklift supervisor down time on every single shift after every single break, far more time than he is losing to the relatively infrequent breakdowns. What he wants is not what he is saying.

To follow the Toyota Production System in the development of new products, it is necessary to apply the Toyota Production System analysis to your customer. For what activity is your customer paid? If you sell to a consumer and your customer is not paid for using your product, how is he rewarded? What is the actual cause of pleasure or reward, and what is getting in the way of achieving it? When you understand what the customer actually wants, you will understand what the customer will actually pay for. There is very little point in asking him what he wants (you’ll probably hear about faster horses and free pizza), but if you need to you might ask why he is doing as he is—you might even ask him five times to get to the root of it.

But you must know who your customer is – who you wish your customer was, to become more like what they want, and who is the customer that will pay you today for what you can do today. And this is why so many companies have struggled to implement the Toyota Production System: they don’t know who their customers are.

At the factory where Tom worked, we thought the customer was the end user of the product, or at least the distributor who ordered it from us. But we were wrong. The business was not run to satisfy that customer. That customer would not care at all whether the factory shipped the product this month or that month, this fiscal quarter or that fiscal quarter. The all-out effort to maximize shipments in a fiscal period was done for the actual customer, the one who cared about that result: and that is the shareholder.

As it turns out, the shareholder pays the CEO.

The shareholder may pay the CEO indirectly, through the board of directors, by means of the shareholder vote; or the shareholder may pay the CEO directly, in the form of stock options and grants. As the stock value goes up, the CEO gains wealth. What activity will make the stock price go up? That becomes the value-add activity.

And now in the factory things become very confused for the factory worker, because he is told that Toyota Production System is “customer focused” and will make the “customer” happy, but he is reprimanded if he does not make the shareholder happy. And some of the things he is told to do for Toyota Production System seem to make sense for the “customer,” like never make any defective products, but some of the things he is told do not seem to make any difference to the “customer” at all – such as that it is very important to ship on June 30 and not July 1. And so it seems like Toyota Production System is a bunch of nonsense that doesn’t actually make his boss happy, and so he treats it like nonsense, and before too long it goes away. And then another CEO wonders why Toyota Production System doesn’t work for his company, and he finds something different to try to promise his shareholders as a panacea.

And that CEO will never understand that the Toyota Production System does not consider inventory wasteful because it ties up your capital. If you can get paid to tie up your capital in inventory, by all means do it! (And thus warehousing exists as a profitable business.) Toyota Production System views inventory as waste because it hides waste. Inventory is a time buffer, and a time buffer exists to hide a variation in process time, and variation in process time exists because the worker does not know what to do or because the worker needs to fix a defect. And that is a much more serious problem than the negligible amount of overhead expense allocated to storing inventory.

The Toyota Production System is not a panacea. It is fundamentally designed to improve processes, and not everything worth doing is a process. But the Toyota Production System has “failed” far more often from being misunderstood, misconstrued and misapplied than it has from any real deficiency in its basic philosophy, which is to understand what your customer wants, and do that.

[Note: Tom is an alias.]

Excuses Enabling Waste

I first began learning about “Lean Manufacturing” in 2006, and since then I’ve had many opportunities to see this philosophy misunderstood and misapplied. Most recently I saw a variation on basic Lean terminology in a class meant to teach Lean Six Sigma principles for process improvement.

I’ve never been formally credentialed in Lean Manufacturing and it presently seemed like a good time to cross it off the list. I am taking an online program mainly consisting of recorded lectures, which means I did not have the opportunity to object when the professor started explaining the concept of “Value-Enabling Activity.”

In classical Lean Manufacturing (substantially derived from the Toyota Production System), an activity is classified as either Value Adding (VA) or Non-Value Adding (NVA). The simple rule of thumb is: would a customer reasonably pay you more money to do more of this activity?

In the case the professor was discussing, the setting was health care. Time spent in the waiting room was classified as Non-Value Adding, since nobody wants more time in the waiting room. Time spent with the physician was classified as Value Adding, since that’s the essential service that people are seeking.

But for check-in activity the professor had a category previously unknown to me, “Value-Enabling”. The idea is that the health care provider cannot safely provide care without going through the check-in activity—checking for medical allergies, other health conditions, and so on. Even though it would be hard to find a customer willing to pay more for more time spent on check-in, this activity is considered essential and not properly described as “waste,” the other term for Non-Value Adding activity.

This is a basic perversion of the meaning and purpose of classifying activity as Value Adding or waste.

While it may be true that with today’s laws, customs, and technology there might not be at present any way to eliminate the check-in activity, the fact that the patient do not enjoy this activity and would happily be rid of it should never be forgotten. That is the entire purpose of classifying activities very simply as either value-adding or waste.

It is no defense to say that the activity is necessary. It is currently necessary to ship cars from the car factory to the customer – and typical to ship them not to the customer, but to the dealership. Transportation is absolutely one of the seven wastes. Transporting the car from the factory to the dealership is waste, because no customer would pay extra to have the car shipped back and forth several more times. It is necessary but unwanted – just like the waste you flush away every day.

A reflective person will realize that, really, the only way to have no waste is if the product materializes immediately upon the customer’s first realization that they want it. Anything else involves delays and is wasteful.

More specifically, consider the case of a chunk of metal being shaped into a widget. Modern CNC (Computer Numerical Control) machines can automatically switch between several different tools to shape a part. The process of switching between tools is Motion, and therefore waste.

In most analyses, the time a chunk of metal spends in a machine is all Value-Add time. Many people would balk at the idea of being so hyper-sensitive as to classify the time switching tooling as waste. Nevertheless it is true that if you somehow had one tool which could do it all, and therefore did not need to switch tools, the process would take less time; your machine would be more efficient and therefore more productive.

The correct reason not to worry about the time a machine spends automatically changing its tooling is that this small amount of time may not be where your real problems lie. And this is perfectly consistent with the principles of Lean Manufacturing; no practitioner of Lean Manufacturing would suggest chasing a waste of a few seconds when there are problems costing you hours of time you could resolve. There is absolutely no shame in working on the biggest problems first; the only way it might not be best to work on the biggest problem is if you are not actually capable of influencing it. Otherwise, always go for the biggest problem. If it’s too complex, figure out the major causes and pick the most significant; repeat as needed until you’ve got a problem you can work on.

Going back to the health care scenario, check-in time might currently be necessary. But what if a quick scan of your retina were enough to bring up your complete medical record and verify that it is really you? This would virtually eliminate the Non-Value time and leave only the valuable part of the check-in process.

Someone might object that in this particular case, there may be a potentially viable technology solution, but that, in general, there are some things that just have to be done in order to get to the good stuff, and those things are Value-Enabling. But that category allows you to justify and excuse activity your customer dislikes, turning a blind eye toward your own inefficiency and becoming systematically indifferent to what your customer wants.

Consider that it is still necessary today for products to be shipped from the factory to the consumer. If everybody just shrugged and said “Can’t be helped,” Amazon would not be the force in commerce that it is today. On the contrary, Amazon has patented anticipatory shipping – shipping you something before you tell them you want it, so that the shipping time gets even closer to zero. Service technicians do this all the time, keeping commonly needed replacement parts in their vans, but Amazon is the first major company to realize that this basic concept could be extended across the entire range of things ordinary retail consumers might want shipped to their house.

It’s not clear whether Amazon has actually implemented this idea yet. But, with delivery drones and delivery lockers and delivery to your trunk, it is very clear that Amazon hasn’t declared the time it takes to delivery an order “Value Enabling” and thus beyond criticism. Transportation is waste. Get rid of it whenever possible.

There is no business or process anywhere in the world that has achieved fully zero waste. Some of the causes of waste are laws of physics and some of the causes are laws of governments. It should never bother you to have some part of your process described as waste; if your feelings are hurt, you’ve misunderstood the message completely. And it’s easy to test: if you are right and it’s not waste, charge more for it; do it more often; increase the amount of it you have in proportion to the other parts of your process. If it’s actually valuable you’ll make more money. But if it will increase your costs – then yes, Virginia, it is waste.