There are a lot of misconceptions and, shall we say, incomplete information about how Bill Gates became rich. But it’s an interesting and important story, so let’s cut through some of the misconceptions and the conspiracy theorizing of Gates and his early years.
Bill Gates made his money through software, first programming languages, and then operating systems. By getting into position to sell one or both products to virtually every computer manufacturer, Gates became a millionaire by age 23 and a billionaire by age 31.
The myth of Bill Gates
A fair amount of myth surrounds any famous person, and Gates is no exception. He is the young prodigy who never bothered to finish college and outsmarted IBM to become a billionaire, swindling a more talented rival in the process.
The real story is more complicated than that. Myths typically happen because people like explanations that fit on the back of a napkin, but if you want to understand Bill Gates, it requires more analysis than what fits on a napkin.
How Bill Gates became a millionaire
The other problem with understanding Bill Gates is the realization that his wealth came in stages. He was hardly a nobody when he met with IBM. He was already rich at that point.
And that story begins at the dawn of personal computing. In 1974, a company called MITS started offering a personal computer that you could build yourself. But it wasn’t like building a computer today. They sent you a box of parts, but nothing was pre-assembled. Before you could plug the circuit boards together, you had to solder all of the components onto those circuit boards, and then you could put the computer together.
That may not sound like a very appealing product, but MITS couldn’t keep it in stock. There was a long waiting list to get one.
Bill Gates and Paul Allen recognized the potential. Their families were wealthy enough that they had access to computers in high school in the early 1970s, but there was no such thing as a home computer. This was uncharted territory. They knew this home computer would be popular, but it didn’t have any software. Their bright idea was to create software for this new computer that would make it easier for its owners to write their own software for it. The product would be an implementation of a programming language called Basic. It was a language that you could learn in an afternoon, but it took years to master.
They were in the right place at the right time with the right idea, but there was a problem. They couldn’t get one of these machines.
Developing on hardware sight unseen
Gates and Allen had access to a large business computer through their school. So they wrote an emulator for this new system that ran on the computer they had access to. Or, should I say, Paul Allen wrote the emulator. Gates worked on the product, then used Allen’s emulator to test his code once it was ready.
And even that is an oversimplification. What set Microsoft’s implementation of this programming language apart was that it could do complex math with decimal points. Gates wasn’t looking forward to writing that part, and another friend overheard the conversation and volunteered to work on that part of the project, telling Gates he had done something like that before and could do it again.
By the time they finished, they had a paper tape containing their code. It ran fine on the emulator, but up to that point, it had never run on the real hardware. For that matter, they had never seen the real hardware except in photographs.
Once it was ready, Paul Allen flew to MITS to demonstrate the product, carrying the paper tape with him. When he arrived on site, he loaded the tape into the machine, loaded their code, and tried it out on real hardware for the very first time. It worked, and MITS quickly worked out an agreement to distribute the software to its owners.
Microsoft’s stroke of genius: Licensing, not selling rights
Microsoft Basic immediately became the de facto standard for small computers.
Crucially, Microsoft did not give MITS exclusive rights to the product. Microsoft reserved the right to port the product to other machines.
As you might imagine, other companies noticed MITS and its early success and decided they wanted a piece of that market. So other computers followed, some of them compatible with the MITS computer, and some not. And it wasn’t long before several companies decided it might be a good idea to try selling already assembled computers that the purchaser could take out of the box and plug in without having to know how to solder. Those computers were ready by 1977, and all of them either shipped with Microsoft Basic, or were running Microsoft Basic soon after.
By 1978, at the age of 23, Gates was a millionaire.
We’ve all seen that famous 1977 mugshot of Bill Gates where he looks like he’s barely a teenager and has a goofy grin on his face. He was older than he looked, and within a few months, he would be a millionaire. Because there were dozens of companies making computers to sell to consumers, and without a product he controlled, those computers were essentially useless. It didn’t matter whose computer you bought, Bill Gates and Paul Allen made money off it.
Bill Gates and the college myth
Bill Gates famously went to Harvard and never finished. Paul Allen talked him into dropping out to form Microsoft with him. So there is a popular narrative that Bill Gates is an example of a billionaire who didn’t go to college.
But college isn’t just about what you learn in the classroom. It is also about the conversations you have, whether that is in between classes, or after hours entirely.
If you go to college and you only learn from what some instructor says in the classroom, you did it wrong. You also meet crucial people who challenge the way you think, and those relationships can become crucial in unpredictable ways.
Gates met his collaborators at school. College also provided Gates and Allen access to computers at a time when they couldn’t get them on their own. No amount of money was going to get them higher up on that waiting list. Even then, that was just a single computer. The exposure to the bigger systems allowed Gates and Allen to think beyond just what that simple computer could do in 1975.
There was another crucial relationship Gates formed at Harvard. He met a business student by the name of Steve Ballmer.
Steve Ballmer is a complicated person. In many ways he is a goofy parody of himself, and his tenure as Microsoft’s second CEO had its ups and downs. But Ballmer was a shrewd and ruthless businessman. Gates was also a shrewd and ruthless thinker, but he didn’t have the business background that Ballmer did. He saw how the two of them complimented one another, and if Gates had not met Ballmer, I’m not so sure he would have become a billionaire. At the very least, Ballmer made it happen a lot faster than it otherwise would have. Because when IBM came calling in 1980, it was Steve Ballmer who Bill Gates called to ask for advice.
Ballmer had three words to say to Gates: buy a suit.
How Bill Gates became a billionaire
There is more myth around how Bill Gates became a billionaire than the rest of his story. IBM was developing a personal computer. They needed an operating system for this personal computer. IBM also needed a programming language. The obvious choice was Microsoft Basic. They also knew that Microsoft resold a popular operating system called CP/M. They hoped they might be able to buy both things from the same supplier.
Gates of course was more than happy to supply his Basic programming language to IBM, but he didn’t have the rights to resell CP/M to them. Even if he did, what he had couldn’t run on the CPU IBM wanted to use. Ironically, the best Gates could do was sell them an operating system called Xenix, which is a form of Unix. But it couldn’t run on the CPU IBM wanted to use either.
Gates referred IBM to Gary Kildall and his company Digital Research, the original producers of CP/M.
According to the myth, Gary Kildall wouldn’t make time to meet with IBM, decided to go flying instead, and IBM came back to Gates, who seized on the opportunity and provided IBM with an operating system.
What really happened to IBM and Digital Research
The only public comment Bill Gates ever made about IBM and Digital Research was, “Gary went flying.” And the story grew taller from there. History is written by the victors
It didn’t help that Gary Kildall never told the story himself. He died in 1993, and even his unfinished memoir left out a lot of details. His ex-wife, Dorothy McEwen, was involved in the negotiations, but was unable to recall most of the details.
So for years all we had were bits and pieces that Gary Kildall told his friends, which mostly came down to IBM wanting Digital Research to sign a non-disclosure agreement that made them uncomfortable and seemed one-sided. Kildall had meetings with other companies that day and IBM arrived early, but he did indeed meet with them himself. But they didn’t come to an agreement.
In 2016, former Digital Research employee Tom Rolander filled in the missing details. The major disagreement was over licensing. Digital Research was working on a version of CP/M the would run on the Intel 8088 CPU that IBM wanted to use. IBM wanted to license that operating system, but they didn’t want to call it CP/M. They wanted to call it PC DOS. And rather than paying a royalty, they wanted to pay up front rather than paying a licensing fee.
Digital Research didn’t like that idea, because it suggested that the product was separate and incompatible with every other operating system on the market. They thought selling the product under a different name through IBM would be bad for their own brand. And they thought making an exception for IBM would mean every other company would want the same deal.
In all fairness, neither company had much of an idea where they stood with one another. Each thought they had a deal, but on terms much closer to what they wanted.
Back to Bill Gates and Microsoft
Once IBM figured out that they weren’t going to get exactly what they wanted from Digital Research, they went and talked to Bill Gates again. And the difference between Bill Gates and Gary Kildall that ended up making all of the difference was that, at least in 1980, Bill Gates was willing to give IBM anything they wanted.
Kildall admired IBM but didn’t want to make exceptions for anyone, including them. Gates, being the underdog for the last time in his life, was willing to make an exception.
The problem was, Microsoft wasn’t really an operating system company. Fortunately for Gates, he knew people. There was a small company in Seattle called Seattle Computer Products who built a computer that used an Intel 8086 CPU, similar and compatible with the CPU IBM wanted to use. Digital Research was dragging their feet on porting CP/M to that CPU. So one of their employees, Tim Patterson, wrote a CP/M clone that would run on the 8086.
Microsoft licensed this operating system, then bought it outright for $50,000. Then they turned around and licensed it to IBM for much more. But they also retained the right to license it to others. After SCP realized they’d done the equivalent of selling Manhattan Island for $20 worth of trinkets, they sued, eventually settling for $1 million.
There have been rumors for decades about secret code that only Kildall himself understood being present in DOS, and also a persistent rumor that there was an Easter egg that printed Kildall’s name in early versions of DOS, and that either one would prove DOS violated CP/M’s copyright. Any knowledge of any secret code died with Kildall, and the people who claimed to know about the Easter egg never gave enough detail for anyone else to replicate it.
It makes for a great story, but it may never be anything more than a story. It is also possible that some stories got mixed together. Gates planted and Easter egg in Microsoft Basic, which makes me wonder if some people may have gotten some wires crossed and combined some unrelated stories. What Kildall did instead of suing is yet another story: DR DOS.
But back to Gates. Arguably it wasn’t the IBM deal that made Gates a billionaire. It was what IBM left out of the deal.
The clones
Nothing in the agreements gave IBM exclusivity. They got to sell the Microsoft products under their own name, but Microsoft had the right to do what they wanted on their own.
Whether Gates anticipated the possibility of IBM clones or whether he thought being able to sell the operating system to other companies was just a good idea because Digital Research was doing it, or because he just licensed DOS because that’s what he did with Basic, we may never know for sure.
But whether intentionally or unintentionally, he repeated the formula he’d followed with Microsoft Basic. Only this time it was with a product that proved to be even more valuable. IBM sold a product called PC DOS, leaving Microsoft free to sell the virtually identical MS-DOS.
IBM made it very difficult to clone its PC by putting certain critical functions on a ROM chip on the motherboard. This ROM was copyrighted. And then IBM published the source code for that ROM in the form of a book and included it with the computer. If anyone cloned the ROM, IBM could say they copied the code IBM published. IBM lawyers knew it was difficult to prove you had not done something.
That’s one reason politicians will sometimes make outrageous accusations against their opponent. Proving a negative is incredibly difficult.
It didn’t take long at all for clones to appear. Predictably, IBM went after them and was able to shut most of them down at first. Microsoft made some money regardless. If a cloner failed, Microsoft made a little money. If they succeeded, Microsoft stood to make a lot of money. But they had no legal liability themselves.
The television series Halt and Catch Fire dramatizes those early efforts to create an IBM compatible PC. Their depiction of reverse engineering the IBM ROM is fairly close to how it happened. The detail the show got wrong was no computer company was try to reverse engineer the operating system. You could just license that from Microsoft and save yourself a lot of trouble.
Compaq cloned the ROM and kept their clone to themselves. But other companies soon followed, and sold their ROM to other computer companies. So you could build an IBM-compatible computer by buying an Intel CPU, the rest of the chipset from Intel and NEC, a ROM BIOS from a company like Phoenix, and an operating system from Microsoft. At that point, IBM couldn’t control the flood.
By 1986 or 87, almost every computer company not named Apple was either selling an IBM compatible computer, or was working on one. And not only was IBM selling more than a million computers a year, but their most successful competitor was selling 2 million units per year. Microsoft derived some financial benefit no matter what computer anyone bought. And for the fastest growing segment of the market, Microsoft was making a lot per sale.
The time was right to go public with an IPO. Microsoft’s IPO was incredibly successful. By some accounts the IPO immediately made Gates a billionaire on paper, and by others, it took until 1987 for it to happen.
On paper, Bill Gates was a billionaire at age 31.
Gates’ legacy is something I’ve gone through before, so there’s little reason to rehash it here. But some of the details of how he got where he is today can be a bit confusing and aren’t widely known. It wasn’t dumb luck, and it wasn’t a conspiracy. Everything about Gates is more complicated than that.
David Farquhar is a computer security professional, entrepreneur, and author. He started his career as a part-time computer technician in 1994, worked his way up to system administrator by 1997, and has specialized in vulnerability management since 2013. He invests in real estate on the side and his hobbies include O gauge trains, baseball cards, and retro computers and video games. A University of Missouri graduate, he holds CISSP and Security+ certifications. He lives in St. Louis with his family.
Great post!
On the question of easter eggs in CP/M also in DOS or other questions about if there was wholesale copying. The source of both are now available and a forensic analysis showed that there was API copying, which may or may not be an IP problem, but nothing more than that. This speech has bad audio but covers that. The link is to the corresponding white paper too:
https://www.youtube.com/watch?v=GEzQx5ApsMg
https://www.zeidmanconsulting.com/DOS_comparisons/
Thanks. I am aware of Zeidman’s analysis, but he only tested one possible scenario, that of source code theft. The more likely scenario would have been working from a memory dump or a disassembly of CP/M 80, which his methodology would be unlikely to turn up. Given that CP/M was written mostly in PL/M and DOS was written mostly in 8086 assembler, the memory dump or disassembly scenario is the more likely. I always found his observation that the API was the same a bit humorous. That was part of IBM’s original design specification. The intent all along was for it to be easy to port 8 bit CP/M apps to the IBM PC.
I covered that a few years back at https://dfarq.homeip.net/did-microsoft-steal-dos-from-cpm/ .
The problem with Pournelle’s claim is that he never disclosed it. He said he had the key sequence written down but couldn’t locate the paper or notebook he’d written it in. The possibility remains that someday, someone will find it in his papers. But unless or until that happens, or someone else discovers it, it’s just a rumor.
So I am inclined to think that under the copyright laws in effect at the time, DOS was in the clear. Zeidman’s methodology has problems, but Pournelle’s story has a bigger problem.