It was in most of the trade magazines and on websites over the last few weeks, but the mainstream media didn’t give it much coverage. Nonetheless, it seems like a pretty big deal: Cadillac is separating itself from the financially struggling General Motors brand.

What’s interesting is that it seemed to work so well just a few years ago. In fact, as we’ve gone through brand architecture exercises with clients, it was the prime example we pointed to when discussing the merits of dual branding the parent company with a “division,” “enterprise,” “business unit,” “sub-brand” — whatever certain organizations call it. Many clients have asked us over the years whether they should try to brand both themselves (often the sub-brand or business unit) and their parent company or umbrella brand, which in some cases has its own existing brand, and our clients simply want to communicate the connection. In other cases, though, neither may have a recognizable brand, which would obviously make things tougher.

Developing a brand and delivering upon a brand promise — one at a time — is tough enough. Trying two at a time, tied closely together, is altogether different. And, what if one doesn’t perform well? Do the long-term benefits of dual branding firmly outweigh the risk of one brand hurting the other?

History and resources also come into play. Cadillac and GM have both, and could pull off top-of-mind awareness and an audience affinity to two brands. Not everyone has the history or resources.

There are, of course, other examples — Nabisco and Oreo (and most of its other products), Proctor and Gamble and Crest toothpaste (P&G is a little more silent, but most people still connect the two). But GM and Cadillac were different. There was no disputing the brand equity of both (at least until a couple of years ago), and they went to great lengths to make the connection. It seemed to work.

They made it look easy, and many companies would aspire to do the same thing. But, of course it’s not easy. It takes years to build recognition, and even longer to gain equity. And it takes exposure — a lot of exposure, which translates to huge investments in advertising, public relations, and other marketing communications; point-of-purchase signage; product labeling, and on, and on. Not only is/was the Cadillac brand everywhere, but GM was right there with it: in the ads, on the dealership, on and in the cars. It took a long time to do, and it may take even longer to undo — not only distancing Cadillac from GM, but also Cadillac from the other GM brands.

“There is a lot of pressure on Cadillac this year because it took such a beating last year,” said a dealer from Ann Arbor and Troy, Michigan. “As Cadillac dealers, we didn’t like being lumped in with the other GM brands, especially when they threw us into the Red Tag sale. We felt it cheapened the brand.”

And thus the risk.

So now we’ll see changing dealership signage, the removal of logos from cars, and certainly new advertising and PR communications. Eventually, we may even see a more aggressive campaign designed to try to make us forget the tie completely … as if there were never a relationship between Cadillac and GM at all.

Not only is dual branding not as easy as it looks, it may not be beneficial in the long run. In some cases, it can work. Clearly, in other cases, it doesn’t. And now, we need to identify the newest example of dual branding that does work. We definitely have a new one for dual branding that didn’t … at least, long-term.

Read the entire BusinessWeek article.