Even Microsoft Started from Nothing

We are all born small, but given the right genes and a rich environment, even the tiniest can grow into giants. On November 26, 1976, the Office of the Secretary of State of New Mexico awarded a trademark to a four-man Albuquerque startup headquartered in a two-bedroom apartment. In the 30 years since the day its name was first registered, Microsoft has grown to be the highest-earning technology company in the world.

The decade of Microsoft’s birth, quite recent by most standards, is antiquity in the computer age. Yet a stereotype survives of the basic disco-era computer enthusiast: an unkempt, college-age male, hunched, fever-eyed, over a box of tangled wires and circuits in his garage. He may never make a dime from his efforts; he just wants to play around with the machine, find out what it can do, and share the results with his friends, computer hobbyists around the world.

Bill Gates never fit that image. Sure, the cofounder of Microsoft may have looked the part: shaggy, uncombed hair, disheveled clothes. But from the moment he saw his first computer, when he was 13, the boy who drew up contracts on the playground to rent his sister’s baseball mitt was trying to make money off a new toy.

Gates already stood out at Lakeside, his Seattle private school. The son of a wealthy, high-powered lawyer, he was the smartest kid in class, and he knew it. Obsessive and fiercely competitive, he had a photographic memory and a knack for math and science and would chortle at students who failed to master physics concepts in one try. But he didn’t understand true passion until he started eighth grade in 1968. Over the summer Lakeside had bought a teletype machine linked by a phone line to a mainframe computer downtown. Gates soon became one of a cluster of proto-techies who jostled to spend all their free time punching BASIC commands onto rolls of yellow paper tape.

Frequently a chubby sophomore stood at the teletype breathing down Gates’s neck. His name was Paul Allen. He and Gates became fast friends in the newly christened “computer room.” “We both were fascinated with the different possibilities of what you could do with computers,” Allen recalled. “It was a vast area of knowledge we were trying to absorb.” But as much fun as he had programming ticktacktoe and lunar landing games, Gates, a born capitalist, saw more lucrative possibilities. He, Allen, and two other classmates formed the Lakeside Programmers Group in the fall of 1970. Their sole mission: to make money. “I was the mover,” Gates later said. “I was the guy who said, ‘Let’s call the real world and try to sell something to it.’”

The first opportunity came while he was still in eighth grade. The Computer Center Corporation, a Seattle company that sold time on a giant PDP-10 mainframe, wanted to make sure its new hardware and software could withstand a full load of users. Gates and his friends were hired to try their best to crash the system. (Decades later, millions of Windows 95 buyers would pay him to do essentially the same job.) In return, they got all the free computer time they wanted. Gates often lingered at the CCC teletypes past midnight, learning the ins and outs of hardware, software, and computer languages. “It was when we got free time at C-Cubed that we really got into computers,” he remembered. ”I mean, then I became hard core. It was day and night.”

Gates was just finishing junior high; by the time he graduated from high school he had earned a tidy sum. A Portland computer-timesharing company gave the Lakeside Programmers Group $10,000 of free computer time in 1970 as payment for a payroll program. Junior year, Gates netted $4,200 for computerizing Lakeside’s class scheduling. Then he and Allen started a company, Traf-O-Data, to sell municipal governments traffic-flow-analysis software. (They made about $20,000.) Finally when the defense contractor TRW offered the pair $4 an hour to help program the new computers at its Vancouver, Washington, power plant, Gates took off the second trimester of his senior year to work full-time 160 miles from home. He returned in the spring—wallet, mind, and ego bulging—to graduate with his class.

Despite his youth, his work with Allen in the early 1970s was definitely not kid stuff. They produced some fairly advanced programming, with ever higher levels of professionalism. But nothing they did then changed the world; that would have to wait until college. In December 1974 Gates was midway through his sophomore year at Harvard; Allen worked close by at Honeywell in Boston. Walking across Harvard Square one day on his way to visit Gates, Allen noticed the new Popular Electronics at a newsstand. “PROJECT BREAKTHROUGH!” the cover declared, over a picture of a rectangular, switch-studded metal box. “World’s First Minicomputer Kit to Rival Commercial Models.” The minicomputer in question was the Altair 8080, the first affordable personal computer marketed to the general public.

Allen sprinted the rest of the way to Gates’s dorm. Here, finally, was their chance. If they could develop a programming language for this computer, they could set the standard for—and maybe even control—a promising new industry. (These motivations may seem too easy to ascribe to two college-age kids in hindsight, but recall that Gates often boasted that he would make his first million by age 25.) Nobody had started a personal-computer software company yet, but only one could be first. “We realized that the revolution might happen without us,” Gates said. “After we saw that article, there was no question of where our life would focus.”

Thousands of computer hobbyists around the world got just as excited. Orders deluged Micro Instrumentations & Telemetry Systems, the Albuquerque company that sold the Altair. “Computers were things that were housed in big buildings and took up several floors and had a staff to maintain them and a priesthood to watch over them,” said Eddie Curry, the executive vice president of MITS. “A large part of the success of the Altair and the microcomputers that followed was the desire of people just to own one. It didn’t really matter if they could do anything.” That was lucky, because there wasn’t much you could do with an Altair anyway. It had no monitor or keyboard, and it understood only binary code. To program it, the user had to toggle the switches on its front panel hundreds or thousands of times. One wrong flip could derail an entire sequence.

But were the Altair to come with a programming language, buyers could use it to enter commands on a teletype. Over Christmas vacation, Gates and Allen sent MITS a letter on Traf-O-Data stationery. “We have available a BASIC language interpreter that runs on MCS-8080 series computers. We are interested in selling copies of this software to hobbyists through you,” they wrote. “If you are interested, please contact us.” Ed Roberts, the president of MITS, had heard several similar claims, but no one had come through with a finished language. He bit anyway. When he phoned the number on the letterhead, though, he reached the home of one of Gates and Allen’s Lakeside friends—the ”official” Traf-O-Data headquarters. The boy’s mother answered. She had no idea what Roberts was talking about.

Back at school, Gates finally got through to MITS. “We’ve got this BASIC, and can we come out and talk to you?” he asked. “Well, we don’t have enough memory cards for the machine yet,” Roberts said. “Why don’t you come out in a month?” Gates and Allen were ecstatic, except for two snags: They hadn’t actually started on the language, and they didn’t have an Altair to write it on. Allen quickly modified an existing program to simulate the Altair’s microchip, and Gates started filling notebooks with lines of code.

They had decided to adapt BASIC, an easy-to-learn language originally developed by John Kemeny and Thomas Kurtz at Dartmouth in 1963, for the Altair’s chip and small (4 kilobytes at most) memory. Gates skipped class, meals, and sleep to finish it in time. A month later Allen—who, as the older of the two, looked more like an adult—was flying to Albuquerque with a roll of paper tape. Somewhere over the Midwest, icy terror suddenly stabbed him in the gut: He had forgotten to write the program that would tell the computer to recognize the tape. Jotting furiously on a single leaf of paper, he finished it by the time the plane landed.

At the airport, Allen met the burly, six-foot-four-inch Roberts, who motioned him toward an old pickup truck. They drove to MITS headquarters, sandwiched in a strip mall between a massage parlor and a laundromat. Inside, crowded among a herd of exhausted new employees hired to fill the backlog of orders, Allen saw his first real Altair. “It was a squat sheet-metal box that looked one cut above military surplus,” wrote Stephen Manes and Paul Andrews in Gates: How Microsoft’s Mogul Reinvented an Industry—and Made Himself the Richest Man in America, “but . . . the nondescript box held the most memory of any microcomputer in the world.” Luckily enough, the machine was hooked up to a tape reader, so Allen wouldn’t have to input BASIC manually, a job requiring 30,000 toggles. He did, however, have to key in the program he had written on the plane. As Allen fumbled with the tiny switches, Roberts and Bill Yates, the Altair’s designers, chuckled behind him. Whatever this kid had come up with couldn’t possibly work. Allen silently agreed. The program had never been run on a real Altair. If Gates had made one single mistake, all their labor would amount to nothing. As the paper tape ran through the reader, Allen crossed his fingers.

Suddenly, the teletype started clacking. “MEMORY SIZE?”

“Hey, it typed something,” Yates said.

“Oh, my God, it actually works on a real computer!” Allen thought. He entered the memory size.

“READY,” the teletype printed.

“PRINT 2 + 2,” Allen typed.

“4,” the computer responded.

In that moment, Gates and Allen turned an insular hobby world into an industry. By June, they had both relocated to Albuquerque. MITS hired Allen as director of software, while Gates spent his summer vacation perfecting BASIC. They signed a ten-year contract granting MITS exclusive license to the language and the right to sublicense it to third parties. In return Gates and Allen got a royalty (between $30 and $60, depending on the BASIC version) for every copy sold and the promise that MITS would promote their program. When purchased with an Altair, BASIC sold for $75; on its own, it cost upward of $500. Software had been invented decades before, but only now did it become a product.

As thousands of orders for Altairs poured in—more than MITS could handle, in fact—BASIC was selling somewhere down in the hundreds. Soon Gates figured out why. Members of the Homebrew Computer Club in Silicon Valley had duplicated a tape of Gates’s program. Now countless copies were floating around America’s garages, basements, and dorm rooms. MITS BASIC had become the world’s first pirated software. Many Altair buyers felt no qualms accepting, and even copying, a pirated program. Sure, mainframe software was rented, and at high prices. But most hobbyists took a more populist approach. You paid for the hardware—the chips and wiring—and the software you shared.

Gates changed all that. This, more than BASIC or MS-DOS or Windows, was his most important innovation. “Bill Gates had recognized what Roberts and all the others had not: that with the advent of cheap, personal computers, software could and should come to the fore as the principal driving agent in computing,” wrote the historian Paul Ceruzzi. “And only by charging money for it—even though it had originally been free—could that happen.”

Back at Harvard for his junior year, Gates fired off a scathing letter to Computer Notesmagazine: “Most of you steal your software. . . . Who cares if the people who worked on it get paid? . . . One thing you do do is prevent good software from being written. Who can afford to do professional work for nothing? What hobbyist can put 3 man-years into programming, finding all bugs, documenting his product and distribute for free? The fact is, no one besides us has invested a lot of money in hobby software.”

Of course, Gates’s position proved somewhat hypocritical. He hadn’t paid the original developers of BASIC, after all; how was his innovation protected intellectual property and theirs not? The debate would rage for years. (Gates revisited it in the 1980s, when Apple sued him for ripping off the Macintosh graphic interface for Windows.) What Gates couldn’t see was that the pirated paper tape was making BASIC the industry standard. The illegal copies had spread to computer rooms all across the country. New hardware makers, inspired by demand for the Altair, were springing up all the time, and they all wanted to do business with a company headquartered in a dingy Albuquerque apartment.

Gates closed his letter with a dash of optimism that, uncharacteristically, underestimated his future: “Nothing would please me more than being able to hire ten programmers and deluge the hobby market with good software.” He signed the letter “Bill Gates, General Partner, Micro-Soft.” A portmanteau of microcomputer and software, the name had been coined sometime the previous summer. (In the licensing contract with MITS, Gates and Allen had been referred to as simply the “Licensors.”) Roberts had been the first to use the name in print in October 1975, in a letter to Computer Notes magazine defending the decision to charge for BASIC.

Microsoft started hiring new programmers in the spring of 1976. Later that year it moved into its first real offices, on the eighth floor of a bank building, and on November 26 the Office of the Secretary of the State of New Mexico registered “Microsoft” as a trademark. Gates dropped out of Harvard two months later. Microsoft was looking more and more like a real business. But one thing was holding it back—its contract with the floundering MITS. Roberts had never been able to keep up with demand for the Altair, and in 1977 he sold the company to the Pertec Computer Corporation. The Pertec brass knew that BASIC, not the increasingly obsolete Altair, was MITS’s real asset, and they refused to license it to other manufacturers. Gates and Allen used that fact to wriggle out of the contract, which had strictly stated that the Licensee must use its “best efforts” to sublicense their program.

After that, nothing—not competitors, not market slumps, not the Justice Department—could slow the most successful startup in U.S. history. Microsoft made $16,500 in 1975. By 1978, its sales topped $1 million. Last year, its thirtieth anniversary in business, it grossed $39.8 billion.