The secret to success in auto manufacturing today is flexibility: building more than one platform on the same assembly line.

That sounds reasonable, but it is terribly expensive to revamp a plant with flexible tooling if flexibility isn’t already baked into your process, like it is at Toyota and Honda. There is a cheaper alternative.

Everyone wants to cut manufacturing costs. In Detroit, too often that means squeezing suppliers. But the most important costs in building a car are the tooling and the factory itself. The more cars you build in a factory, the lower the fixed cost per unit. That's why flexibility has become so important. Production runs are getting smaller all the time. I remember when Chevy and Ford would make 1 million units of one model. Today the best-selling car is Toyota's Camry, with 432,000 U.S. sales last year.

Truck runs are higher, but they may come down, too, as model proliferation grows. Building more than one platform is one way of getting more production out of a factory, but building more vehicles off the same platform is easier. This used to be Detroit's specialty. Over time, it has been forgotten or confused with badge engineering, which is building different versions of the same vehicle with a few superficial changes.

Chrysler has the right idea. Its Brampton, Ont., Canada, plant builds LX platform cars: the Chrysler 300, Dodge Magnum wagon and the Dodge Charger. These are distinct designs off the same platform. Last year Chrysler built 320,000 of these cars at the Canadian plant, working three shifts. That makes money, big time.

GM and Ford are doing it, too, to some extent. GM assembles three distinct vehicles, the Cadillac CTS, STS and SRX, all off GM’s Sigma platform, in Lansing, MI.

Ford builds three similar vehicles in one plant in Hermosillo, Mexico (closer to badge engineering, unfortunately): the Ford Fusion, Mercury Milan and Lincoln Zephyr.

To work properly, the technique requires constant refreshing of same-platform variations to keep the volume high.

If LX sales slow for example, and they inevitably will, Chrysler quickly has to offer new variations, such as the Dodge Challenger and Chrysler Imperial concepts shown off at the Detroit auto show. It probably will. It is hard to understand why auto makers are not doing it more.

Ford said it wouldn’t build a Mercury version of the Ford Freestyle cross/utility vehicle. Why not? It would add to Mercury sales and add volume at its Chicago assembly plant, lowering overhead costs per unit. And imagine if Ford could build a new rear-drive platform and run a new Crown Victoria, Mercury Grand Marquis and Lincoln Town Car off the same line in one plant. That would be enough for a highly profitable 3-shift operation. More models, fewer platforms. That’s the trick.

Jerry Flint is a columnist for, and a former senior editor of, Forbes magazine.