This essay is adapted from Lean Solutions by James P. Womack and Daniel T. Jones, to be published Tuesday, Oct. 11, by Free Press, an imprint of Simon & Schuster Inc.
Most of us have been conditioned to think that car buyers decide what they want in an instant, and then expect to obtain what they want in the next instant. But is instantaneous response always the customer's desire? And do consumers really make their decisions on the spur of the moment?
Or do they plan ahead? And could they be equally satisfied receiving the product at some point in the future, particularly if that point was certain and could reduce their total cost of consumption, while getting exactly the product they wanted by collaborating with a provider on the process?
In our case - and we expect that we are typical - we don't think very far ahead about mundane items. The right kind of toothbrush will be available in the drugstore whenever we need it.
But we do think far in advance about big-ticket items such as cars. A decision-making algorithm is always grinding away in the back of our head, and we know some ways in advance what we are going to ask for when we do ask. But today we have no one to talk to until the actual moment of commitment to a specific product at a specific price.
Imagine going into a car dealership and saying that you will need a new mid-sized SUV of a given description, with the following color and equipment, about a year from now, when your current vehicle reaches an age and mileage that makes you worry about its reliability.
And imagine further that you offer to make a commitment right now - with some flexibility on the delivery date to suit the automaker's production capacity - in return for a significant discount.
Who could you find to talk to? Possibly the security guard, after you're labeled an unstable personality, but certainly not the salesman. No dealer is currently geared to talk with you about the long term - only about the instantaneous short term of making the actual deal.
In the past few years, there has been an enormous amount of discussion in the global auto industry about how to address the presumed desire of most customers to get exactly the product they want right now.
Taking inspiration from Dell Computer's practice of making every product to order, practically every automaker has embarked on a "three-day car" or a "five-day car" initiative, to discover how to make the precise vehicle each customer really wants quickly to a confirmed order.
The reward for the industry would be twofold. Actual transaction prices - what the customer really pays for the vehicle - could be firmed up by eliminating discounts to move unpopular models and option combinations built to inaccurate forecasts.
And carrying costs could be eliminated on the 60-day period for new vehicles on dealer lots - currently worth about $60 billion - that has been a constant in the North American auto industry for more than 80 years.
What do the companies have to show for their efforts? Very little. With current designs, car companies must add more than 1,000 parts and components to each vehicle as it comes down the final assembly line.
What's more, the number of copies of a given model that car companies are able to sell each year, even in global markets, continues to shrink. So manufacturers have found it necessary to run several completely different vehicles down many of their assembly lines to achieve economies of scale.
Each of these vehicles comes in several body styles, with many options and several trim levels, and in a wide range of colors. This creates a multitude of choices for many of the 1,000 fit points on each vehicle. It greatly complicates the job of getting the right materials to the right assembly station at the right time.
As a result, even a lean producer such as Toyota has discovered that to make the assembly process work, it must lock its schedule in place 10 days in advance of actually building a specific car.
Car companies must add to this delay the time it takes to get an order from a customer through the dealer to the car company and worked into the production schedule, which usually takes several weeks. And then they must add the time required to get the vehicle cost-effectively from the factory to the dealer (for "preparation") and on to the customer, which is about a week.
Simple math shows that it is not possible today to promise a customer delivery of a custom-ordered vehicle in less than a month, even when it is manufactured in the region of sale. And if the vehicle must travel by boat between regions, it is necessary to add two to four more weeks. At least Toyota is likely to deliver the vehicle on the promised day, because of the rigor with which the company manages its order and production process.
By contrast, the typical European luxury-vehicle manufacturer doesn't offer custom-order vehicles in markets outside of the region of production. And it delivers made-to-order vehicles to customers on or before the exact date promised - typically six to eight weeks after the sale - only about 20 percent of the time.
Doubtless, progress will continue to be made in compressing the scheduling process. We hope that suppliers will gradually relocate closer to assemblers within the region of final assembly. As a result, the total time required from customer order to delivery will surely shrink somewhat in the years ahead.
But the prospect of truly making every vehicle to confirmed customer order, and delivering it within three to five days, is practically nonexistent at any time in the foreseeable future.
However, if a pure get-it-right-now-for-every-customer model can't work, perhaps a different approach can. To consider the possibilities, let's start with three physical facts:
1. Current automotive production systems can deal with small amounts of get-it-for-me-now demand, provided that production slots for last-minute orders are planned in advance and don't exceed a certain fraction of production.
2. Getting these made-to-order products to the customer quickly will still cost car companies more, no matter how well production is planned. This is because of the need to juggle the schedule, expedite missing parts from suppliers and expedite the finished product to the consumer.
3. Making all other vehicles to customer order and getting them to the customer slowly will cost the provider less, because the provider can carefully plan its output and arrange for supplies at just the right time.
This would make it possible for providers to practice what Toyota calls heijunka - leveling demand by both total volume and mix over extended periods. As a result, the whole production process could run smoothly at a steady pace, rather than responding continuously to sudden shifts in aggregate demand and mix.
Continuously changing both the total level of output and the mix of products causes all sorts of costs. These include overtime, "just-in-case" inventories, excessive equipment wear because there is no time for preventive maintenance, and excessive capacity on average to meet the spikes in demand. These costs can be largely avoided with some simple leveling.
Now let's make the leap from current practice. Suppose dealerships had no finished units. None, that is, but a demonstrator for each type of product on offer and a few loaner vehicles, perhaps low-mileage used vehicles coming back on trades.
And suppose they offered the customer two choices:
The ability to plan ahead produces major savings for the provider on those vehicles. And it doesn't inconvenience the customer, so long as the new vehicle arrives when it is promised and a loaner vehicle can be provided if the current vehicle suddenly ceases to work in the meantime.
Equally important, this slot system - with plan-ahead and immediate-delivery slots moving through the same provision system - permits car companies to satisfy both got-to-have-it-now and plan-ahead customers.
It can do this without the customer steering, special deals, unrealistic delivery dates and expediting that get-it-for-me-now companies such as Dell currently face by offering their customers only one option.
What's more, it permits the provider to build every product to customer order, while saving large amounts of costs by level-scheduling its production facilities and getting parts without the need for expediting.
If this approach were fully implemented in the auto industry, everything would be built to order. There would be no finished-unit inventories, with $60 billion of carrying costs, and no rebates on vehicles built to the wrong forecast.