The president of one auto company just doesn't publicly criticize the way a competitor runs his business. But it happened early this year. I missed it, but so did lots of people, and better late than never.

The critic was James Holden, president of the North American side of DaimlerChrysler. He said this in a Detroit speech: "A nameplate is not a brand. The automakers who think they are, in our opinion, are kidding themselves to think they can build equity in 47 different brands at once."

Mr. Holden was taking a whack at General Motors Corp., because GM is the only automaker in the whole wide world that thinks this way. The GM strategy has been a disaster and helped the company lose six points of market share. Mr. Holden must be the original Mr. Nice Guy. Better to let GM continue on its misguided way and DaimlerChrysler would pick up some of the business it loses.

This is a serious problem at GM. Let me explain.

Most of the auto world has been carried away by Brand Marketing theory the past few years. The idea seems to be that if customers associate your vehicle (the brand) with certain favorable characteristics, it will sell better, like Coca-Cola and all those soaps from Procter & Gamble. Then again, if you noticed what's been happening at Coke and P&G recently, you might question the whole idea, but Detroit is always behind the marketing curve.

GM led the move to brand marketing when a cabal of directors, led by John Smale, overthrew the management. Mr. Smale, an outside director, was from P&G and the biggest believer in brand marketing.

My oft-repeated opinion is that brand marketing works when the product costs 25 cents to make while $1.75 is spent on advertising and packaging to con the public. Then it sells for $3.95, leaving lots of profit.

The new GM managers, eager to please the fellow who put them in office, embraced brand marketing. Every GM vehicle became a brand. The Chevrolet Cavalier was a brand, the Chevy Malibu was a brand, the Chevy Camaro was a brand, and so were the Chevy Monte Carlo, Chevy Lumina and the Chevy Impala. And we haven't gotten to the Chevy trucks, or all the Pontiacs, which were brands, and all the Oldsmobiles, Buicks, Cadillacs and GMC trucks. Each model was a separate brand.

The highest Ford Motor Co. marketing officials told me that GM was nuts, but they weren't Mr. Nice Guys like Jim Holden of DaimlerChrysler, who had to tell GM.

Ford and DaimlerChrysler do embrace the brand marketing theory, but for them the "division,' such as Ford, or Lincoln, or Dodge or Chrysler are the brands, not the nameplates within that grouping. Thus, Chrysler is a brand and vehicles carrying the Chrysler name are supposed to have certain similar characteristics. Actually, this is phony, too, because Ford and DaimlerChrysler build their cars and then fit the brand definitions around whatever they have built.

But at least they don't get as hung up about it as GM.

The brand marketing theory worked its way into the organization. Divisional vice presidents who knew the business were pushed out. New people, often young, often from outside the business, who had learned their trade selling diapers or pencils or dog food, were put in charge of the vehicles, or brands.

GM market share declined year after year. There were other reasons: design, engines, quality. Brand marketing was part of it. Apparently no executive wanted to offend Director Smale by saying that what sold soap didn't sell cars.

Fortunately, he retired from the board this year.

Privately, some important GM people have told me that the concept of "every nameplate a brand" is dead, and now the divisions (like Oldsmobile) would be the brands. But I haven't really seen the proof. And managements are loathe to admit they have be en wrong for eight years.

Unfortunately, there is a bigger problem: the GM structure. The new division chiefs are devoid of power. They don't decide what the cars look like, the quality, or the engineering. The brand managers, who are supposed to be powers in the new organization, often are new to the business. The real power in the system is supposed to go to the Vehicle Line Executive for each model, but I see very little sign of success here. A VLE may be doing several models, which raises lots of questions. I just don't think that has worked too well. The proof is in the steady decline of market share.

GM was reorganized almost to death in the 1980s, but the company needs another one - going back a few steps.

Maybe they'll thank Jim Holden of DaimlerChrysler for pointing it out.