Birth of a Behemoth

Microsoft wasn’t always Microsoft. And when I say that, I mean more than just that its original name was Micro-Soft. The heartwarming tale of two high school friends who left college to found the world’s most successful company has been told many times and has inspired many imi­tators; the dot-com world was overrun with them.

An uncanny knack for launching the right product at the right time took it part of the way. In fact, some contend that marketing is the only area in which Microsoft excels. But, you may be wondering, why are we talking about product quality and marketing skills? What of increasing returns, positive feedback, and virtuous cycles? What of tipping to a stan­dard and lock-in? Didn’t Microsoft’s Windows simply become the de facto standard for PC operating systems? Why wouldn’t that alone be enough to guarantee Microsoft’s dominance and to relegate all others to remote corners of the computing world, if not to outright oblivion?

All fine questions. But before we even get to network economics, we must answer two more basic questions—how a software monopolist emerged from the pack, and why it was Microsoft. Competitive markets are supposed to be robust. It should be virtually impossible to cheat your way into becoming a monopolist. The Bork-inspired Chicago School even revolutionized antitrust law by arguing that virtually all markets would regulate themselves through the competitive process, and that antitrust enforcement should be extremely rare. And yet, by the late 1990s, Robert Bork himself had come to believe that Microsoft had beaten the system to become a marauding monopolist.

Long before any of those events occurred, Microsoft became a monop­olist by being the strongest player in a market protected by reasonably high barriers. Economists are well aware of the human penchant for monopolization. They know that every entrant wants to compete, that every competitor wants to join an oligopolistic cartel, and that every oligopolist wants to destroy his competitors to become a monopolist. But they also know that only markets protected by barriers to entry are prone to monopolization. In the absence of such barriers, a monopolist’s attempt to raise prices and to extract added profits from consumers will invite new entrants. The relationship among entry barriers, switching costs, and profitability doomed the dot-coms. But software has at least two natural protective barriers, and both are highest in platforms.

The first barrier derives from the nature of software itself, and from the difficulty inherent in profiting from the sale of a product that lends itself to unauthorized reproduction—the zero-price prediction that haunts the information sector. A software company must invest time and effort in software development, but once the first copy goes out the door, competitors can copy it and circulate it at virtually no cost. The barrier thus arises because the software business appears to be inherently unprofitable. This “unprofitability” barrier presented itself in the early days of software, gained resolution to some extent, but continues to rematerialize; outright software piracy remains a huge issue in the inter­national arena, and the outcry over P2P file swapping continues to roil the entertainment industry. This barrier is also fundamental, so anyone interested in profiting from an information product must devise a way around it.

The second natural barrier in software markets, and the one that’s both of greater significance today and of more direct relevance to Microsoft’s position as a monopolist, lies in the network nature of the operating systems market. But that’s a barrier for a mature industry. There may be no way to reach the network barrier to entry until we’ve resolved the unprofitability barrier. How did this industry ever develope? Why didn’t rampant piracy stop it dead in its tracks? After all, how could anyone— much less a start-up software firm—police and enforce IP rights that are so easy to ignore?

A historical detour may provide some answers. Back in the 1950s, 60s, and 70s, the way to make money in the computer business was to sell computers: big, expensive, precision equipment that only governments, universities, and big businesses could afford. These machines often filled an entire room, and access to that “computer room” was strictly con­trolled. Most users—and becoming an authorized user was far from trivial—wrote their programs on paper. Then, confident that the program would work, a users would present her work to a keypunch operator who would help her produce a stack of coded punch cards.

Now, this little blast from the past may bring fits of nostalgia to some, peals of laughter to others, but it’s unlikely to make anyone jump up and scream: Man, if I were around back then, I would have started a soft­ware company! And by and large, nobody did. The task of writing soft­ware fell to the computer companies, who saw it primarily as a component of the computing systems that they were selling. Their prod­ucts consisted of many working parts; some were hardware and others software. Occasionally, a hardware manufacturer might allow a small company—or even a couple of college kids—to develop software under contract. For the most part, software was just a component of a com­puting system.

But technology marches on, and by the mid-1970s the machines had become smaller and less expensive, the programming community had grown beyond its original retinue of well-trained nerds with company- issued pocket protectors, and the wonderful world of punch cards had begun to fade into the era of the dumb terminal. All of a sudden, programmers could sit at individual computer screens and observe their interactions with the still-hidden computing machine. A few companies, such as Radio Shack, even made small stand-alone “personal” comput­ing machines. Suddenly, we had developed a class of software users— people who might even be willing to buy software that didn’t come with their machines.

Of course, no one could have seen this niche as a decent way to make money. After all, everyone who owned a computer knew how to write his own programs. Hobbyists bought computers to write code; that’s what made them hobbyists. And those who wanted to use others’ pro­grams could copy them easily.

No decent-sized company was likely to dedicate the resources neces­sary to engage in such a risky line of business. The prospects for prof­itability just seemed to be too slim. And so the challenge fell to a couple of kids. Back in 1975, when keypunches, terminals, and personal com­puters coexisted, Paul Allen showed Bill Gates a copy of Popular Elec­tronics with a cover story describing the MITS Altair 8800, a $360 computing kit. The two crafted code exciting enough to earn them a con­tract writing Altair software. And so, in the balance of risks taken in the name of innovation, Gates dropped out of Harvard to found Micro-Soft, Allen moonlighted at Micro-Soft while keeping his job at MITS, and MITS ponied up the princely sum of $3,000, agreed to share the royal­ties with the kids and their little startup, and (I guess) looked the other way as Allen moonlighted.¹

At the beginning, MITS appeared to have the best of the deal. Micro­Soft’s software became reasonably popular reasonably quickly, and it probably helped improve Altair sales. That made MITS happy, but it didn’t really help the young Mr. Gates; Micro-soft was having a hard time collecting its royalties. Much to Gates’s chagrin, few people were willing to pay for software that they could easily copy from their friends. Fortunately for the future of his company, Gates wasn’t bashful. On February 3, 1976, Gates penned an open letter challenging the govern­ing ethos of the small-but-growing hobbyist community.

To me, the most critical thing in the hobby market right now is the lack of good software courses, books and software itself. Without good software and an

owner who understands programming, a hobby computer is wasted. Will quality software be written for the hobby market?

Almost a year ago, Paul Allen and myself, expecting the hobby market to expand, hired Monte Davidoff and developed Altair BASIC. Though the initial work took only two months, the three of us have spent most of the last year documenting, improving and adding features to BASIC....

The feedback we have gotten from the hundreds of people who say they are using BASIC has all been positive. Two surprising things are apparent, however, 1) Most of these “users” never bought BASIC (less than 10% of all Altair owners have bought BASIC), and 2) The amount of royalties we have received from sales to hobbyists makes the time spent on Altair BASIC worth less than $2 an hour.

Why is this? As the majority of hobbyists must be aware, most of you steal your software. Hardware must be paid for, but software is something to share. Who cares if the people who worked on it get paid? Is this fair?... One thing you do is prevent good software from being written Most directly, the thing

you do is theft.²

Gates’s lament is a wonderfully articulate expression of a fundamen­tal problem. He had inadvertently developed a business model that would later inspire an entire generation of dot-coms: Build your market first and worry about profits later. Unlike his followers on the Internet, though, Gates recognized the model’s folly. He knew that if he could collect royalties from only one out of every ten “customers,” he would have to charge an exorbitant royalty rate just to break even—and that, high price in turn, would drive even more hobbyists to theft. How could anyone expect to make money in such a cockamamy business?

While it was tough for a couple of years, the young company flowered beyond its maiden contract with Altair, lost its hyphen, moved to Washington State, and built itself up to a couple of dozen employees and a few million in annual revenues. While a few other software companies managed to spring up, the business model still looked shaky. Software was just too hard to steal. In 1981, Microsoft got its big break. IBM, with the help of Intel, was about to foray into the world of personal comput­ing. But IBM had designed all of its software to power mainframes. It was unsuitable for a machine designed for home users, hobbyists, and small businesses. IBM needed an operating system to power its new machine. In one of those occasional fateful decisions that determine the course of industrial history, IBM outsourced the task and licensed MS- DOS (Microsoft Disk Operating System) to serve as its operating system.

Microsoft suddenly had a captive audience—everyone whose IBM com­puter came equipped with MS-DOS—and captives can’t be thieves. An entire world of IBM customers would have no opportunity to avoid paying Microsoft’s royalties. Microsoft was thus able to clear that first fateful barrier of easy theft; its guaranteed royalties were large enough to both recoup its fixed development costs and to turn a profit. Any further sales that it could make before rampant piracy set in were gravy.

So that explains Microsoft. How did anyone else enter this nascent software industry? Though IBM was the biggest of the hardware man­ufacturers, it was far from the only one. At that time, consumers con­figured their own systems. They were able to choose from among a number of hardware companies, and the hardware choice restricted their operating systems choices. MS-DOS was popular, but it wasn’t even the only version of DOS, and DOS wasn’t the only operating system. The operating system choice, in turn, delimited the range of available soft- ware—at least some of which came preloaded onto the system. Once again, software sold with hardware provided software companies at least some security. The software companies also began offering documenta­tion and services to their paying clients, effectively expanding their product offerings beyond the easily stolen software itself. And then, to a large extent, the community ethos changed. Consumers came to share Gates’s belief that unauthorized copying of software was theft. At least many American consumers came to share that belief. In other parts of the world, different beliefs dominate even today. By some estimates, for example, piracy accounts for over ninety percent of all software in China—roughly the same proportion of the market that motivated Gates’s open letter.

And so, the software industry cleared its first barrier and emerged to play a small role in a computing world still dominated by hardware. And Microsoft, through its partnership with IBM, stood poised to become the biggest player in that new industry.