But Toyota’s precursors, including Ford and General Motors, all used “batch and queue” systems, as Dr. Womack and Mr. Jones called them. They were all geared for mass production. They made large quantities of every automobile they produced, depending on economies of scale to lower their costs, and therefore kept large inventories on hand of every part they needed. Toyota plants, by contrast, produced a varied stream of goods flowing rapidly down the line, at a cost below those of their competitors. This seemingly impossible feat was achieved through a series of interrelated innovations, which evolved gradually from the 1930s on. Toyota’s process designers judged every step according to the value it provided customers; if they saw no customer value in it, they discarded it, no matter how beloved it might be to finance, engineering, or any other organizational function. With this principle of eliminating muda implemented and supported from the top, workers at each station could be entrusted to control their local operations — not in a random way, but in harmony with the operation as a whole.
For example, each station in a Toyota plant produced only as many parts, whether fenders, windshield wipers, or assembled engines, as its “customer” at the next station called for. Each plant as a whole produced only the vehicles that a dealer had specifically ordered, either in response to a purchaser’s request (which was typically fulfilled within 10 days) or from the dealers’ on-the-ground awareness of customer preferences. This was known at Toyota as the “pull” principle, a term adopted by Dr. Womack and Mr. Jones. “Whenever we drive by a car dealer,” they wrote in Lean Thinking, “our first thought is always the same: ‘Look at all that muda, the vast lot of cars already made which no one wants…. Why did the factory build [them] in advance of customer pull?’”
Toyota also had developed its principle of “flow”: the smooth movement of work from each step to the next, with as few breaks in the sequence as possible. Instead of being controlled from above, the plant moved according to signals sent forward and back from each part of the factory to its internal customers and suppliers. The whole system regulated itself through “takt time,” an expression that Toyota borrowed from German musicians: a metronome-like beat that was paced to match the daily demand for particular components. There were no quality inspectors; workers verified each part’s reliability before it left the station. There were few, if any, storehouses; Toyota had invented the “just-in-time” delivery system to ensure that its suppliers delivered parts as soon as they were required, directly to the places that needed them. And there were no labor problems. Toyota had worked out a groundbreaking agreement in 1946, exchanging lifetime employment for flexibility: Workers could be redeployed or retrained at any time.
Toyota’s customer-based concept of value (“only assets that attract customers are considered valuable”) turned many conventional practices upside down. For example, the automaker saw typical adversarial supply chain relationships as intrinsically wasteful and expensive, because companies were preoccupied with outwitting each other instead of serving customers. Donald L. Runkle, the vice chairman of Delphi Corporation (responsible for purchasing, and a long-standing lean production champion at Delphi and GM) compares these relationships to poker games. “Each side holds their cards close to the vest, and they often negotiate by bluffing,” he says. “You don’t talk about costs very much, because that might show your cards, and you might lose some advantage.”
Moreover, as Mr. Jones noted in a recent e-mail newsletter, the practice of squeezing suppliers tends to push them into self-defeating efforts that produce savings only in the short run: “Without any fundamental changes, there is a limit to how much margin there is left to squeeze.”