Apple Didn't Refuse to License the Mac. One Man Did — to Protect a 55% Margin.

On June 25, 1985, two Microsoft executives mailed Apple a roadmap for winning the personal computer war — and it was a roadmap to give the Mac away. Bill Gates and Jeff Raikes urged Apple to license Mac technology to three to five major manufacturers — AT&T, DEC, HP, Motorola — and let the platform spread the way Microsoft's would soon spread on every beige box in America.¹ The memo went to John Sculley and Jean-Louis Gassée.¹ One of them wanted to do it. The other killed it. For thirty years, almost everyone has had the wrong man in the wrong chair.

The official story is that Sculley, the soft-drink executive who pushed Jobs out, was too closed-minded to open the platform — and that his stubbornness handed the desktop to Windows. Nearly every word of that is backwards. Sculley was the one trying to open it.

The licensing champion was the man history blames: the CEO was shopping the platform around while someone else quietly killed every deal

Sculley didn't resist licensing in 1985. He chased it. He sent a vice president, Chuck Berger, out to drum up interest, and the interest was real — AT&T's chief executive personally contacted Sculley about bundling Mac software onto Unix workstations.⁴¹⁰ This is a CEO actively shopping his platform to the largest companies in the country. The man who stopped him was Jean-Louis Gassée, the newly installed head of Macintosh, and his argument was not philosophy. It was arithmetic: licensing the OS to other hardware makers would surrender Apple's hardware margins, which ran above 55%.⁴ By the winter of 1985, Gassée had killed every deal and Sculley dropped the plans.⁴ The closed Mac was not an act of ideology. It was an act of margin protection — and it was made by the head of the hardware business, not the CEO.

Here is the thesis, and it is uncomfortable for both camps. Apple stayed closed not because Sculley was rigid, but because Gassée was right about the money — Apple was a hardware company that happened to make an operating system, and every machine someone else built was a machine Apple didn't sell. The decision looks like a blunder only if you assume Apple's real product was the software. It wasn't. The real product was the box, and the box carried a margin no licensee would ever pay Apple to give up.

What licensing would have cannibalized: follow the dollar and the open strategy only pays if the buyers were never yours to begin with

To see the trap, follow the dollar. When Apple sells a Mac, it captures the entire 55%-plus margin on the hardware.⁴ When Apple licenses the OS instead, it captures a fee per machine — and someone else captures the hardware. That trade only works if the licensee sells to people Apple could never reach. If the licensee instead sells to people who would have bought a Mac, Apple has swapped a fat hardware margin for a thin software fee on its own customer. Gassée's bet was that the second thing would happen. A decade later, Apple ran the experiment and found out.

The clone program proved the fear — at a $50-per-machine price: Apple finally ran the experiment, and the nightmare arrived precisely on schedule

Apple finally reversed course. In December 1994 it announced a clone program with Power Computing as the first licensee, and Radius shipped the first authorized clone on March 27, 1995. The terms: Mac System 7.x and the ROMs, licensed at a flat $50 per machine.⁵ On paper, free money. In practice, a leak. Apple's CFO found the strategy cost the company money outright — the $50 fee didn't come close to recouping the hardware sale it replaced — and CFO Fred Anderson told Wired that 'somewhere around 99 percent' of clone sales went to the existing customer base.¹² That is Gassée's nightmare arriving exactly on schedule. The clones weren't expanding the Mac universe. They were selling the same Mac buyers a cheaper box, and Apple was collecting fifty dollars to lose its best customers.

Steve Jobs ended it the way he ended most things — abruptly and completely. He shipped Mac OS 8 on July 26, 1997, an OS the clone vendors were not licensed to sell, stranding them on the old version. Then Apple bought Power Computing's core assets for $100 million in Apple stock that September.⁶ He didn't cancel the program; he made it impossible to continue, then absorbed the largest survivor. The bet on premium hardware economics, suspended for two and a half years, was back on.

Wasn't the closed platform what cost Apple the war?: the standard autopsy names the wrong cause of death

The strongest objection is the one history settled on: Windows won the desktop precisely because it ran everywhere, and Apple, by staying closed, doomed itself to a single-digit share and a near-bankruptcy. That is true as far as it goes — and it is the wrong autopsy. Even Jobs, announcing the Power Computing buyout, said clones could only amount to a tiny percentage of overall Mac OS sales, and that Mac OS computer sales had declined by almost 20% over the prior two years.¹³ The collapse was a demand problem, not a licensing problem. And when Apple did open up, the clones didn't grow the platform; they fed on it.⁷ So the refusal wasn't a strategic error that lost the war. It was a defensible — if ultimately losing — bet that Apple's money lived in the hardware, and the one time Apple bet the other way, the numbers proved Gassée right and the company nearly died doing it.

There is one more irony buried in the same year the licensing fight began. Under pressure from Microsoft, which threatened to stop building Mac applications, Sculley signed a deal on November 22, 1985 granting Microsoft a license to the Mac's visual displays.⁸ When Apple later sued, the courts found that agreement covered 179 of the 189 interface elements Apple contested, and the rest weren't copyrightable.² The Ninth Circuit affirmed in September 1994, ruling Apple could not get patent-like protection for the idea of a graphical interface, and the Supreme Court declined to hear it in February 1995.³ Apple wouldn't license the look to hardware makers who'd pay for it — but had already licensed it, under duress, to the one competitor best positioned to exploit it.

The legend needs a villain who was too closed-minded to see the future. The record offers something stranger and more useful: a CEO who fought to open the platform and a hardware chief who closed it on arithmetic — and who was vindicated by the very experiment that proved him a loser. Apple's refusal to license the Mac wasn't a failure of vision. It was a correct read of where its money lived, held a decade too long against a market that had stopped caring. The bet wasn't wrong. The world was.