Google’s new plan to operate as a technology holding company called Alphabet raises many questions about how such a structure should ideally work.

A good place to look for insight is the experience of a highly innovative company from the last century: General Motors. A very similar idea to what Google envisions is spelled out in “My Years with General Motors,” the 1963 memoir of legendary executive Alfred P. Sloan Jr.

We know GM as a crushing bureaucracy, but when Sloan stepped down as chairman in the mid-1950s his GM stood atop the brand-new Fortune 500 ranking by such a margin that there was really no comparing it with the oil company in second place. It wasn’t America’s most successful car company. It was America’s most successful company.

What should be interesting is how Sloan achieved that, interesting even to innovators as renowned as Google co-founders Larry Page and Sergey Brin. They have either figured out already, or soon will, that there’s really just one good way to manage a big company with lots of business units. That’s Sloan’s way.

Not that either has been known for embracing traditional management. In the early 2000s, they even tried getting rid of managers altogether, because talking with bosses and reviewing projects seemed to get in the way of the real work of the bright engineers.

This experiment didn’t last very long. As described in a Harvard Business Review article, what killed it was Page spending all day, every day fielding questions about expense reports, refereeing battles between engineers and everything else that gets to a manager’s inbox.

It’s easy to imagine Sloan just chuckling at such a notion.

He knew about start-up culture, too. Hot start-ups were what made up the car business a hundred years ago. Had magazines like Fast Company existed in that era, it seems safe to assume that Buick Motor Co. would’ve been as celebrated as Google, so flat and nimble in its management that it didn’t even need an organization chart.

But a flimsy management structure doesn’t work very long. It sure didn’t for GM. By the time Sloan became an officer at the end of World War I, GM had turned itself into a loose confederation of car companies that competed with each other, spent money like sailors on shore leave and pretty much ignored the CEO presiding over it all, the colorful GM founder William Durant.

Durant had already been forced out once as CEO for his irresponsible spending. In his second go-around as GM CEO, he was still a brilliant visionary, and he still hadn’t the first clue on how to actually manage the company.

What Sloan described was simply chaos. Decades later, Sloan still found it unbelievable that GM’s corporate treasurer had to go, hat in hand, to negotiate with the finance staff of the Buick division for enough cash for the corporate parent to pay its taxes and other bills.

It was, as Sloan dryly observed, “decentralization with a vengeance.”

Durant was soon fired again and the new CEO, Pierre S. du Pont, found himself with a copy of a memorandum Sloan had written in late 1919 called the Organization Study.

Don’t be fooled by the dull title. Sloan’s memo should be considered a foundational document in American business management, treated a little like the way constitutional lawyers treat the Magna Carta.

Sloan had figured out the right balance between making all the decisions in headquarters and letting divisional presidents have a free hand. Too much control from the top stifles innovation. Not enough leads to chaos.

Sloan, who became the CEO in 1923 and held that job until 1946, recognized that divisional leaders had to act like entrepreneurs. The boss at Chevrolet needed to feel genuinely empowered to do everything possible to grab market share from Ford.

On the other hand, managers had to know how they fit with each other and the parent company. They had to make products that didn’t compete with each other’s in the market, and they had to freely share technology. If they couldn’t agree, Sloan was happy to have the CEO step in.

Maybe the most important thing Sloan did was get control of the checkbook. GM would only spend money on things that generated a demonstrated return. That put all the divisions on a level playing field, competing for that capital with their best ideas.

Over at Ford Motor Co., Henry Ford had no interest in Sloan’s kind of thinking. Ford once wrote that “the Ford factories and enterprises have no organization, no specific duties attaching to any position, no line of succession or of authority, very few titles, and no conferences.”

That does sound a little like the Google Larry Page envisioned in 2001. One result was that Ford slipped from dominating the industry in the early 1920s to having just half of GM’s market share by the mid-1930s.

Sloan pointed out in his memoir that Ford shared with Sloan’s old boss, Durant, one troubling weakness. As he put it, “they injected their personalities, their genius, so to speak, as a subjective factor into their operations without the discipline of management by method and objective facts.”

Sloan wrote that he wasn’t much of a book reader, but there wouldn’t have been anything to read on modern management in 1920 had he gone looking for it. What he and colleagues did to build a rock solid organization they did with their own thinking. He noted that much of it must look pretty basic to managers trained in business schools.

Now, more than 50 years later, there’s plenty in his book that seems to be well past its sell-by date. No 2015 book on management is likely to have a chapter called “Coordination by Committee.” Then there are his detailed organization charts, page after page of them.

But even something as old-fashioned as an org chart popped up in the reaction to the news that Google was going to transform itself into Alphabet. Technology journalists speculated about what exactly a detailed Alphabet org chart will look like.

We know it will have one. The names on the boxes will be different, but it will look just like the ones Sloan used to run General Motors.