Michael Karesh (2026) stopped by to comment on our review of Thomas Bonsall’s book about the Edsel. I think that Karesh makes an important point about General Motors’ handling of its passenger-car brands. Thus, I wanted to elevate his comment to the front page — and offer a few additional thoughts.
“GM only had too many brands for a company where at least 99% of the people had no idea about the purpose of brands or how to properly manage them. That said, the percentage has been no higher at Ford.
GM’s structure had one fatal flaw that they only started to fix when it was far too late: each brand had a separate set of dealerships that needed whatever sort of car was selling well at any given moment. So each make (except for Cadillac) ended up with every sort of car trying to cover the entire market = 4+ poorly defined brands.”
This got me thinking: Was it inevitable that GM would fail to adapt to changing times? Or were there new approaches management could have adopted much earlier?
GM’s brand hierarchy declined in four stages
Let’s start by analyzing how GM’s traditional hierarchy of brands started to lose relevance. My sense is that the first stage in this process occurred around 1953-55. This was when the low-priced Chevrolet was given many of the features that premium-priced brands once had to themselves. Those included V8 engines, automatic transmissions and a variety of power accessories.
The second stage was when smaller cars were introduced in the early-60s. This began to blur the traditional price distinctions between the brands. As product proliferation increased over the next decade, GM had an increasingly difficult time differentiating its brands.
Also see ‘In 1974 Harry Bradley anticipated the American wagon’s future’
The third stage was the rise of imports as well as vehicles with alternative packaging schemes such as trucks, mini-vans and sport-utility vehicles. Whereas in the early postwar years a brand could build its popularity mostly through styling differentiation, by the 1980s more was often needed. Consumers increasingly chose vehicles based on practical considerations. These ranged from extra cargo capacity to greater reliability.
The fourth stage was the automaker’s increasing emphasis on sharing components. A key result was that a Buick might have virtually the same size, drivetrain and method of manufacturing but would be distinguished by superficial stylistic differences. And by the 1990s, the divisions were downgraded to little more than marketing arms of the company.
How might GM have better adapted to each stage?
Now let’s return to Karesh’s point about how each brand besides Cadillac was given full lineups. In what ways might management have handled that differently in each of the four stages?
In the first stage GM might have withheld from Chevrolet more features given to premium brands. However, as discussed here, that might have undercut the brand’s sales leadership. In addition, as cars got bigger, the need for a V8 engine and such features as power brakes would increase to the point where they were considered essential. So at best this was only a temporary fix.
In the second stage GM might have withheld compacts and even mid-sized cars from at least some premium brands (go here for further discussion). Even so, that also would have only been a temporary fix once Detroit needed to begin downsizing its fleet of passenger cars in the late-70s.
In the third stage it would have been even more difficult to maintain distinct identities of the brands unless GM forced each one to specialize in functional areas. For example, only one division might have been allowed to field a minivan. Perhaps just as importantly, the autonomy of brands could have been protected by allowing each to continue managing their own assembly plants. This would have given premium brands room to emphasize higher manufacturing quality.
Most of the above steps would have more rapidly resulted in corporate centralization of decisions. Yet paradoxically, giving each brand a smaller range of more distinctive products could have better preserved their relevance in the marketplace better than what GM did in the fourth stage.
What was a better structure for GM’s dealer networks?
Karesh is correct that the biggest reason why GM didn’t better protect the integrity of its brands was because it acquiesced to dealer demands for full lineups. So that raises the question: Was there a way to restructure dealer networks so that this didn’t happen?
Perhaps one approach would have been to eliminate GM’s division structure and integrate their dealer networks into one system. This could have reduced the pressure from dealers to give individual brands full lineups. They would instead have had more flexibility to sell multiple brands that fielded a narrower range of products.
Also see ‘Bigger didn’t prove to be better for General Motors in late-70s and 80s’
This undoubtedly would have often reduced the sales of individual brands. For example, Chevrolet might not have outsold Ford nearly as frequently. By the same token, neither Buick in the 1950s, Pontiac in the 1960s and Oldsmobile in the 1970s would likely have seen their sales soar to the levels that they did.
Eliminating the division structure would have been a radical step to take, particularly in the mid-50s. And given the increasing power of dealers due to state-level franchise laws, it could have taken years for the full benefits of integrating dealer networks. Even so, over time it could have made GM a more formidable competitor. And it also might have made it harder for the government to use antitrust laws to break up the company.
What kept GM from adapting to the times?
Perhaps the biggest factor was path dependency. Management found it easier to continue making incremental changes rather than taking a step back and rethinking its basic strategy.
To make matters worse, when management did try to reform GM’s structure they tended to make ill-considered moves, such as over-centralizing the supervision of the automaker’s assembly plants. By the same token, one of the first high-level efforts to micro-manage product decisions led to the Chevrolet Vega backfiring.
Also see ‘Would GM have done better in the 1960s and 1970s under Alfred Sloan?’
So in a sense John Z. DeLorean was right to criticize such moves as over-centralization. However, he was arguably wrong that maintaining the relative autonomy of divisions was the best option. Times were changing and GM’s structure — which was developed way back in the 1920s by Alfred Sloan — was becoming obsolete.
What I find fascinating is that a succession of GM executives was unable to adapt despite the fact that for many years this company lacked for nothing in the way of resources. It could hire the finest talent and back it up with the best technical knowledge that money could buy. That brings us to the moral of this story: Contrary to what is often said about the U.S. auto industry, bigger has not always been better.
Share your reactions to this post with a comment below or a note to the editor.
RE:SOURCES
- Auto editors of Consumer Guide; 1993, 2006. Encyclopedia of American Cars. Publications International, Lincolnwood, IL.
- Karesh, Michael; 2026. Comment in “Thomas Bonsall’s Edsel book is nuanced but suffers from groupthink.” Posted March 25.
- Wright, J. Patrick; 1979. On A Clear Day You Can See General Motors. Wright Enterprises, Grosse Pointe, MI.
ADVERTISING & BROCHURES
- oldcaradvertising.com: General Motors (1955)
- oldcarbrochures.org: Oldsmobile Silhouette (1993)
GM dealer consolidation was practiced to a degree in Canada. In Canada there were no separate divisions per se but rather a single entity operating as “GM of Canada”. The dealer network was essentially split in two as “Chevrolet/Oldsmobile” and “Pontiac (a Chevrolet wearing a Pontic body)/GMC/Buick”. Cadillacs were sold by a select number of dealers operating in both groups. This scheme seemed to work successful for many years until the GM brand offerings were rationalized in the early 2000s. Now most GM dealers in Canada are “Chevrolet/Buick/GMC” (oddly, GMC is redundant with Chevrolet trucks being sold in the same store) again, with Cadillac being added for select dealers.
Perhaps Chevrolet and Pontiac should have kept to a smaller A body and not moved to the B body for 1959. Largest Chevrolet and Pontiac should have been roughly a mid-sized car. Buick and Oldsmobile should have had senior compacts, and Chevrolet and Pontiac would have had subcompacts and compacts. Of course, the full-size Chevrolets and Pontiacs were huge sellers in the 60’s.
There was a huge appetite for full-size cars in the 60s even though the midsize ‘64 Chevelle was very comparable in size to the then full-size tri-five Chevrolets. It would be unimaginable that GM would introduce the ‘64 Chevelle as the “new” Impala. Much of the public still “needed” a full-size automobile even though the standard for such had grown immensely over a ten year period. Look at the outcome of the downsized Plymouth and Dodge of 1962. GM couldn’t have handed over the sale low priced, full-size cars to Ford and Chrysler (after 1964). To me, the real blurring of the lines between brands came not so much from the size of the vehicles (Chevrolet in full size form was always at least marginally smaller than its up the ladder siblings), but rather the upgrading of the traditional low priced three models (i.e. Ford LTD, Chevrolet Caprice, Plymouth VIP) so that they were encroaching on and even surpassing their upmarket competitors in terms of luxury and style. This was the ultimate manifestation of stage one described in the article.
IIRC Dodge panicked in 62 and rushed the 880 into production, broadly a Chrysler Newport with a Dodge front clip. GM tried the same stunt with Pontiac in the late 70s (?) dropping the full sized Pontiac, and almost immediately was forced to bring in the Pontiac Parisienne, a badge engineered Caprice previously sold only in Canada. Just out of curiosity Dan, are the Canadian Chevy trucks concentrated on pickups and SUVs while GMC handles the larger commercial type trucks?
No, the full line of both Chevy and GMC trucks are sold at the same dealerships. I think that there is strong brand loyalty for both nameplates so GM Canada doesn’t want to risk alienating either camp. My son recently purchased a new GMC Yukon XL but he would have taken a Chevy Suburban if it was a better deal.
I should have added the last real attempt GM made to differentiate their products was the 1961 new compacts. Chevy got the Chevy II, a conventional compact. Pontiac went for a transaxle and IRS, Olds got an all aluminum very small block. Buick got the same engine, along with a unique version of the larger car’s Dyna-Flow transmission.
That’s all true. At the same time, the pricing of GM’s premium-priced compacts started to blur the line between the divisions. In addition, the Y-body cars arguably shared too much sheetmetal to be adequately distinctive. This hinted at what was to become an all-too-common practice in the 1970s and 1980s.
Here’s my breakdown of what I would’ve done with GM in my little fantasy world. Chevrolet would have a full lineup of vehicles, from subcompact to full-sized, sporty, minivans, trucks, vans, and SUVs. Theyre the mainstream, entry-level, family brand and the big money maker. There’d be no more Pontiac, just SS sporty versions of some Chevys for those who have a need for speed. Olds and Buick were interchangeable. Kill one of those two, keep the other for the more upscale market. Let’s say Olds survives. They would not offer a full lineup. Only a mid-size and full-sized vehicle, possibly a convertible, maybe a mid-size and/or full-sized SUV. Cadillac would be for only the most affluent, discerning customers. A rival to the European brands (Mercedes, Jaguar, Bentley). Maybe the wreath and crest would offer only a couple of exclusive, super expensive models, not shared with the rest of the GM family. No GMC. It’s redundant. No Hummer, no Saturn, no GEO. Hummer served no purpose and they were hideous. And Chevy (in my world) sells like hotcakes, no need for GEO’s captive imports orthe upstart Saturn. All of these vehicles would be sold at the GM Superstore where one would be located in larger towns and cities across the U.S. and Canada. That’s my little fantasy.
Sounds good Johnathan. What timeline are you thinking? I’d keep Hummer but more of a Jeep Laredo competitor and I mean Laredo only. No Tonka Toy H2 and H3. BTW does GM buy SAAB and Subaru in this scenario, and if so do they also throw away the advantages of owning tnhem?
For purposes of brevity I will concentrate on the ”failed” brands of GM. Pontiac Oldsmobile primarily. The awful irony is that these 2 were the rip roaring favorites of the 1960’s and 1970’s. So what happened?
GM in general was becoming too indistinguishable in its different brands. Reduncancy seemed obvious to many buyers. ”Lets see what Buick with the same recipe ”.
Thats when more individualization should ‘ve come into focus. Pontiac had a great Grand Priz in 1969 ~ 1972 -possibly thier best. That shouldve been thier flagship – not the much less appealing Grand villes and such. Who needed those when you can simply cobble up a 4 door version of the Grand Prix – and the big car market is covered. The Firebird was another big seller – expand on that line a bit. Cancel most of the Tempest line – keep 1 or 2 for the seniors. A compact – Ventura – is useful for 1st timers or students.
Oldsmobiles – The Cutlass was the Gold standard of the industry. How did Olds screw up their golden goose? That was mistake number 1. Keep the Delta 88 with its great ”get out of my way ” twin grille. American classics – drop OLDs 98 and lost in the wilderness Toronado.
Cadillac should ‘ve made second = gen Seville a lower prices compact car – similar to the later CTS. For the young bloods – which are always badly needed.
A better question than how General Motors’ five-division empire might have survived is this: how on earth did it last as long as it did?
The answer isn’t brilliance. It’s timing. Pure, dumb luck wrapped in a booming postwar economy and a field of competitors busy stepping on their own rakes.
The Independents? They didn’t just fade away, they conducted a long, slow, assisted suicide. Ford spent money like a man trying to impress a date who’d already left the restaurant, chasing GM with the Continental Division and the Edsel—two monumentally expensive lessons in how not to read the room. Chrysler, meanwhile, staggered from one crisis to the next, occasionally producing a genuinely good car almost by accident, like a blind hog finding an acorn.
Against that backdrop, GM didn’t have to be perfect. It just had to be less wrong.
In the 1950s and 60s, GM could afford its internal overlap, its brand confusion, and its quiet cannibalism. The tide was high enough to float all five boats, even if they were bumping into each other in the harbor.
Ford couldn’t make a five-division structure work. Lincoln and Mercury wandered through a barren identity desert, changing direction so often they wore grooves in the sand. Chrysler, not to be outdone, decided that if GM had five divisions, then five divisions must be the secret sauce. So they tried it.
DeSoto became redundant almost immediately. Dodge and Plymouth turned on each other like two dogs fighting over the same bone. Imperial was spun off, then pulled back in when the math stopped working. In the end, Plymouth took the bullet so Dodge could limp forward.
Neither Ford nor Chrysler could make the model work. And truth be told, GM couldn’t either. They just took longer to admit it.
The problem was baked in: too much overlap, too much redundancy, too many factories, too many layers of management all needing to be fed. It was a five-headed beast with an appetite that never stopped growing, even as the quality of what it produced became… negotiable.
And while Detroit was busy arguing with itself, the imports showed up—quiet, efficient, and built like someone actually cared. Within a couple of decades, they didn’t just take a bite out of the market. They took half the plate.
So what did GM do?
Naturally, they added more brands.
Hummer. Saturn. Saab.
Because when a man is drowning, the obvious solution is to put on a heavier coat.
Hummer was pure ego on wheels. “Let’s take a Tahoe, square it off, make it look like it just came back from a desert war, and charge a king’s ransom.” The idea being that men with something to prove would line up. Turns out, the men with the most to prove were also the ones least able to afford it. They settled for lift kits and loud exhausts on used Silverados.
Saturn was, in theory, what Chevrolet should have been all along: simple, honest, well-built transportation. But instead of fixing Chevrolet, GM created an entirely new brand to do the job. It was like building a second kitchen because you burned toast in the first one.
And Saab? That was pure vanity. Ford bought Volvo, and GM couldn’t stand not having a Scandinavian conversation piece of their own. So they picked up Saab like a souvenir—less for what it was, more for what it said.
And then there’s GMC. A brand that, to this day, exists as a slightly nicer Chevrolet wearing a different tie, asking for a little more money and hoping nobody looks too closely.
If common sense were lard, GM couldn’t have greased a skillet.
Eventually, the math caught up. Bankruptcy followed. The government stepped in. By then, GM had achieved a rare and unfortunate balance: too big to fail, and too clumsy to succeed without help.
Meanwhile, the answer had been sitting in plain sight.
Two divisions.
One for the masses. One for aspiration.
Ford and Lincoln. Toyota and Lexus. Honda and Acura. Nissan and Infiniti. Volkswagen and Audi.
A broad, bread-and-butter brand that sells everything from practical compacts to work trucks. And a halo brand that sells the dream—same bones underneath, but dressed up with softer leather, shinier trim, quieter cabins, and a badge that tells the neighbors you’ve arrived.
Because let’s be honest: folks spending six figures on a luxury car don’t want to step past a row of bargain sedans and work trucks to get there. They want a separate door. A different smell in the showroom. A sense that they’ve crossed into another tax bracket just by walking in.
Underneath, it’s often the same machinery. Same platforms, same engineering, same bones. But perception carries a premium, and that premium prints money.
It’s genius in its simplicity. Fewer brands. Less overlap. Greater scale. Higher margins where it counts.
So how did GM make five divisions work for as long as it did?
A roaring economy. Weak competition. And just enough distance between perception and reality to keep the illusion alive.
Until one day, it wasn’t.
And like every small-town parade float that looks fine from the bleachers but falls apart up close, the truth became impossible to ignore.
The emperor didn’t just have no clothes.
He’d been naked for years.
On the Seville- Are you thinking along the lines on a BMW 3 series? Something maybe on an Opel platform?
From what I understand, Toyota manages to maintain 5 retail channels in Japan (including Lexus), Nissan 3, and Honda and Mazda 2 channels. If GM had remained as dominant in its home market as Toyota has at home, perhaps 5 channels might still be viable.
Very eloquent and spot-on!
Lots of speculation, reminding one of the endless what-ifs of the Packard (owned several), Nash, Hudson and Studebaker folks. A topic of fascination most certainly. The postwar era was one of constant change and turmoil in society that even prominent historians can still disagree on. Auto-execs of the time were trying, often with limited success, to understand the transportation markets of a new and unprecedented era characterized by the cheapest energy in world history and groundbreaking prosperity in the US. Mistakes were made, to understate the result.
I’ll agree with DeLorean in that the dissolution of unique brand autonomy was a primary mistake. Olds is a case in point: there had always been a unique pride in Lansing in the quality of product manufacturing at Oldsmobile, often rated as one of GMs best plants, a pride which was diluted and largely lost upon Smith’s reorganization. With GM Assembly becoming the responsible manufacturing entity throughout the corporation at the worst time, one of increased product regulation and economic distress. The resulting astronomical costs dictated changes, but up to a point the historical inter-divisional competition had been largely beneficial. However it became clear that simplifying product design and development for efficiency made perfect sense and had to be done… why have every division with it’s own engines and varied transmissions when being selective and using only the best of each could produce better results. For example why not develop a better lightweight automatic plus use THM 350/400 as transmission choices, paired with one good ohc 4, a Pontiac ohc 6, and SBC V8 for use with the lower echelon makes, and Olds low & high deck V8s (imo the best ohv GM V8) for all upper makes, V8s then transitioning to modern aluminium ohc FI versions by the mid ’70s. Eliminating by the ’60s the superfluous duplicative versions of 5 unique divisional powertrains would have provided savings needed for the resources to properly design and develop shared corporate-wide drivetrains (ala Mopar) and perhaps would have allowed GM to avoid the engine and transmission debacles of the ’80 and beyond that greatly sullied GM’s image.
Summed up: GM should have kept unique divisional character, styling, and control of assembly (they had the resources but squandered them with idiotic re-orgs, buying sprees, and Saturn flights of fancy), simplified each division’s product line (no Cimmarons! no Caprice! no Calais!) in order to better retain Sloan’s divisional rank and status, and centralize R&D wherever it made sense. IMO Roger Smith has a lot to answer for.