For decades, the global vehicle platform has been the Holy Grail at General Motors and Ford Motor Co. Yet neither carmaker has quite managed to pull it off.
GM is taking another pass at the idea, with a couple of novel twists. It will try harder to enforce commonality, making sure its engineering centers around the world don't stray from the basic architecture when designing cars for local markets.
But there is danger in overdoing the discipline.
GM struck out in its last attempt to go global - the Delta front-wheel-drive small-car program. Engineers around the world had differing ideas. "Everybody did their own thing," said one GM engineering executive.
So much differentiation crept in that vehicles based on the same platform couldn't be built in the same plant. And local markets did not get enough specific attention when some of Delta's details were conceived. So GM ended up with the worst of both worlds: not enough commonality and not enough individuality.
Ford has run into similar problems during the past 25 years as it strove repeatedly to engineer a "world car."
After a couple of Ford global projects went astray in the 1980s, the company made sure Europe's Mondeo and North America's Contour in the early 1990s were truly twinned. But the European-sized Contour wasn't big enough for U.S. customers. Ford lost out in the crucial mid-range car segment - something from which it is still trying to recover.
GM says global architectures are no excuse for making cookie-cutter cars. Yet neither can regional product developers veer off course without good cause. It is silly to allow small variations that undermine efficiency yet don't add value.
That's where GM may have hit upon a good idea: its "supreme court." A panel of top executive arbiters will decide whether changes sought by European or North American engineers are justified.
But the court is untested. And it faces an old conundrum - one that has plagued Ford and GM for years: When is it right to think global and when is the time to think local? That was the rub for Ford's ill-fated "Ford 2000" strategy and GM's world expansionism in the 1990s.
It is still a quandary, but there is one principle. While sharing platforms, GM must keep local customers in mind.
GM says it already is achieving big cost saving with its new strategy, but the true success of a global platform will be measured in the market. That means that sometimes regional engineers will have to do their own thing.