It must be painful for product planners weighing their options. On one hand is the small, practical car, sure to lose money because it is often priced for first-time buyers. On the other is a roomy, macho sport/utility vehicle (SUV) loaded with curb appeal and options that customers want, generating profits that can exceed the entire cost to buy a small car.

What's an automaker to do? There is, in fact, only so much room on American roads (not to mention its garages and wallets) for SUVs, so it seems inevitable that the craze will end. When the economy begins to slide and gas prices climb, buyers could be ready to trade in their Grand Cherokees and Navigators for something more affordable.

With its latest manufacturing strategy, General Motors Corp. is betting on the future of small cars, which surely is difficult for a company that loses $1,000 on every small car it builds.

The strategy used to be codenamed "Yellowstone," but GM has dropped the moniker because it has "served its purpose," says Mark Hogan, vice president and general manager-Small Car Group. Mr. Hogan talks at length about the strategy at a recent manufacturing conference in Detroit, and he makes it clear that GM wants to employ "co-design modularity" to take advantage of the technological innovations of top suppliers.

Rather than GM's traditional approach with "building to print," the automaker wants suppliers to be integrated with full responsibility for their modules "with little GM participation." The goal is to eliminate parts and optimize materials.

For example, Mr. Hogan refers to the cockpit for the Epsilon program, which is the replacement for today's Chevrolet Malibu, Pontiac Grand Am, Oldsmobile Alero, Saturn LS and Opel Vectra. Co-design modularity reduces the number of parts delivered to the assembly plant from 75 to 1. In addition, total cost reduction is 8%, and quality is expected to jump 53%.

Even before Epsilon, Mr. Hogan says co-design modularity was tapped as a likely strategy for GM's Delta (Cavalier/Sunfire/Opel Astra/Saturn S-series) small car program. Together, Delta and Epsilon account for about 4 million units a year, and in both cases, suppliers will provide most vehicle systems in 15 key modules. But he says the platforms will not fully utilize the strategy because some design aspects already were frozen.

"But I can assure you on future programs we will start co-design modularity further up front in the vehicle development process," he says. "It's the way we're going to do business."

But the concept is held up by a significant labor relations hurdle. The United Auto Workers union will focus heavily on supplier outsourcing in upcoming contract talks as merely a ploy to eliminate union jobs.

Mr. Hogan says sourcing for Delta is "virtually 100%" complete, while Epsilon is 70% sourced. It's unclear whether all of the suppliers have agreed to union representation. "Most of these companies have representation with a variety of different unions," he says.

Mr. Hogan says suppliers will be chosen based on "concept competition" and will have full access to GM research during the program. Ultimately, Mr. Hogan says, GM hopes to forever bury the "brute force and tire iron" cost-cutting methods of the early 1990s and replace it with true partnership.

"After all, they're living in our house now," he says of suppliers. "When one member of the house becomes sick, it really affects everyone .. GM has a stake in ensuring our suppliers are successful."

While GM appears eager to delegate work to suppliers, Mr. Hogan identifies three areas likely to remain core to the automaker: body-in-white, powertrain and painting.