Just because writing code ensconced behind ivy-covered walls at Harvard wasn’t exactly digging in an iron mine, don’t underestimate the vast challenges that faced Gates, or the courage it took for him to take them on. For starters, some experts at Intel didn’t even believe that running a high-level programming language on the 8080 chip was possible. In addition, while Roberts provided schematics for the Altair, the only working model at the time was sitting in his offices in Albuquerque, so Gates and Allen had nothing to program or test their work on.
Undeterred, the duo studied the technical specifications of the Intel central processing unit (CPU) and developed an emulator program on the PDP-10 to virtually replicate the functioning of a working Altair. Drawing on all of their reserves built up from long hours on the night shift at C-Cubed, Traf-O-Data, TRW, and Honeywell, Gates and Allen accomplished the seemingly impossible: a working copy of BASIC that Allen flew out to Albuquerque encoded on a paper punch tape.
When Roberts first beheld his box of chips and toggles turned into a useful computer by Gates and Allen’s efforts, he remembers, “I was dazzled. It was certainly impressive. The Altair was a complex system, and they had never seen it before. What they had done went a lot further than you could have reasonably expected.”43 After a brief celebration in Boston, Allen headed back to New Mexico to work out the bugs in BASIC and prepare for a coming road show to promote it. Gates joined him during the summer after his sophomore year at Harvard was complete. He would never return to college.
MITS already had preorders for thousands of machines, and the hardware team was struggling to keep up. It must have been this kind of demand for even such a rudimentary machine that confirmed Gates’s conviction that the appetite for personal computing was virtually unlimited. Soon thoughts of returning to school in the fall had faded like a sunset over the New Mexico desert. In fact, he would never complete any degree. That summer he and Allen, with legal help from Bill’s father and a local attorney, formed Micro-Soft (the hyphen was later removed)—an abbreviation of “microcomputer software”—and struck a licensing agreement with MITS for their version of BASIC. It wasn’t just a technology tour de force, but a business one as well. At only 19 years old, Gates helped craft a sophisticated agreement that anticipated many of the key issues in software licensing, including exclusivities, sublicensing, and confidentiality.
BASIC sold briskly at first in tandem with Altair shipments, but soon software volume began to drop in relation to hardware. The reason was simple: Most computer owners at the time viewed hardware as something to buy, but software as a something to copy and share.
In response Gates published “An Open Letter to Hobbyists” in early 1976,44 accusing his would-be customers of outright theft and laying out a Randian argument for the protection of intellectual property. “Who can afford to do professional work for nothing?” he asked. Without adequate compensation for valuable work done, he argued, “there is very little incentive to make this software available.” The moochers who take something for free in the short term are only hurting themselves in the long run. “One thing you do [by stealing] is prevent good software from being written.”
Over the next few years as the computer market grew, Gates realized he needed to get out from under an increasingly overwhelmed and inefficiently operated MITS to begin building software for the proliferation of new, competing hardware systems if he ever wanted to realize his dream of Microsoft on every desktop. After relocation from Albuquerque to a new permanent headquarters near his hometown in Washington, the company was beginning to build momentum. Early programmer Steve Wood explains some of the company’s early success.
We were always a year or two ahead of where the demand was really going to be. But we were generally guessing right. A lot of it was Bill’s and particularly Paul’s ability to see where some of the stuff was going to go. A lot of people are able to see things like that, but we had just an enthusiasm, a real high level of drive and ambition. There wasn’t anything we couldn’t do. Okay, so no one has done this for a personal computer before, so what? We can do it. No big deal. . . . We overcommitted ourselves. We missed deadlines. We consistently underestimated the time it would take to do a project, and we committed to too many of those projects at a time. But we always got it done.45
Gates also began recruiting the talent he needed to keep driving his corporate vision. In 1980, with around 30 employees, he reached out to old Harvard friend Steve Ballmer, who was attending Stanford Business School after a stint at product marketing giant Procter & Gamble. He would become Gates’s alter ego, business champion, marketing guru, and eventual heir to the reins of the company. He also took the job of chief recruiting officer from day one and relished the job with entrepreneurial zeal.
“There’s a standing policy here,” Ballmer said in a 1983 interview. “Whenever you meet a kick-ass guy, get him. Do we have a headcount budget? No way. There are some guys you meet only once in a lifetime; so why screw around?” The interviewer noted that Ballmer’s style “hardly complies with standard corporate protocol.” Ballmer responded, “I believe in the old cliché, ‘rules serve the company; the company doesn’t serve the rules.’”46
Ballmer’s favorite fishing spot for future employees was the undergraduate pond where he liked to spot eager tadpoles full of raw talent who could be nurtured to become stellar producers within the Microsoft ranks. The recipe was simple: Take one brilliant student with drive and initiative coupled with a background in science, math, or computers. Offer scant money up front in exchange for a dynamic, empowered, and lively work environment with the financial potential of long-term equity options. Heat in the crucible of brutal work hours, intense deadlines, and a competitive corporate culture with a drive to win. What comes out is a virtual clone of your company leader—Gates himself—a techie Marine with the killer instincts to take any hill against impossible odds.
The work ethic at Microsoft was indeed legendary. Parking spaces at the office filled from front to back with the earliest arrivals closest to the door. Gates memorized which cars belonged to which programmers and took note of who came earliest and stayed latest. It was an unwritten rule that nights and weekends were work hours regardless of how much time was spent during the week. Gates even had a contest with his programmers during out-of-town trips to see who could leave the office at the latest possible moment and still catch their flight at SeaTac, the Seattle-Tacoma International Airport.
Act I: DOS Kapital
Although now famous for the nearly ubiquitous DOS and later Windows operating systems, Microsoft nearly missed out on this most fundamental layer of the computer software hierarchy—the intellectual capital on which the Gates fortune, and thanks to Gates so many others as well, is built.
As the 1980s dawned, a company called Digital Research headed by Gary Kildall all but owned the operating system market for Intel 8080-based computers with a product called CP/M (or Control Program for Microcomputers). An InfoWorld interview in May 1981 quotes Kildall as claiming “a couple hundred thousand CP/M users out there now,” with usage tripling each year.47 In fact, Gates had partially enabled the standard by actively promoting CP/M while singularly focused on the programming language layer, including BASIC. The advantage seemed obvious at the time. After all, it was easier to code his language for one operating system than hundreds. So, for example, Microsoft licensed CP/M from Digital Research and created a hardware product called SoftCard, allowing Apple users to run CP/M-compatible programs on their otherwise incompatible computers.
Then IBM called. And that call turned out to be one of the most critical moments in the history of American business.
In 1980 Big Blue was engaging in a secretive project to rapidly launch its own personal computer based largely on off-the-shelf components and the new 16-bit Intel 8088 processor. The project was on a fast track and IBM had no time to develop its own software, so it started searching for likely providers. One morning in July, Gates got a call from IBM’s J
ack Sams asking for an introductory meeting.
“What about next week?” asked Gates.
“What about tomorrow?”48 responded Sams. Despite Gates’s youthful appearance and their seemingly diametrically opposed corporate cultures, the companies quickly found common ground in the language of technology, and the initial meeting went well. Sams returned in August to talk turkey accompanied by a corporate attorney bearing a nondisclosure agreement (NDA), which Gates signed with little fanfare.
With Gates sworn to secrecy, IBM proceeded to unveil the blueprint of the initiative, dubbed “Project Chess.” IBM needed high-level languages like BASIC, COBOL, FORTRAN, and Pascal for the new 16-bit platform. Could Microsoft create them? Absolutely, replied Gates with his typical can-do, competitive attitude.
But what about the operating system? Could IBM also sublicense the CP/M source code Gates used in his Apple SoftCard? Unfortunately, that was going to be a problem. Not only did Microsoft not have the rights to sell or license the CP/M source code to IBM, but Digital Research didn’t even have the necessary 16-bit version of the operating system to offer in the first place. If IBM wanted CP/M, it would have to talk to Kildall.
Gates set up a meeting for IBM with Digital Research, and the boys in suits flew out the next day. “The meeting was a fiasco,”49 recalls Sams. To start with, Kildall didn’t even show up. Instead, he left his wife in charge of the proceedings, and she, together with a Digital Research attorney, refused to sign IBM’s NDA. IBM badly needed an operating system for its project and was eager to do business. By flying off that night for a Caribbean vacation and by being fussy about an NDA, Kildall turned his back on the opportunity of his lifetime—and left it for Gates to seize.
IBM went back to Microsoft and dumped the problem in Gates’s lap. Gates was more than eager to take up the challenge.
Across town, a small CPU board manufacturer called Seattle Computer Products was equally frustrated with Kildall. It had been trying to cajole Digital Research into delivering CP/M in a 16-bit version for the boards it created. Without an operating system, the latest, more powerful hardware was all but useless to users. As a result, sales were slow. Tired of waiting for Kildall, an in-house programmer by the name of Tim Patterson had written his own operating system, which he dubbed 86-QDOS (for Quick & Dirty Operating System), and word of it soon began to spread.
At least it spread to neighboring Microsoft. As fate would have it, apparently IBM was unaware of it. After learning about this ready-made operating system for the 8086 chip, rather than build the OS from scratch, Microsoft struck a deal with Rod Brock at Seattle Computer to license Patterson’s program for a one-time fee of $25,000. Around this time, Brock was also approached by software distributor LifeBoat Associates—yet another player tired of trying to coax a 16-bit version of CP/M out of Kildall. He offered Brock $250,000 for the rights to 86-QDOS. Gates and Ballmer countered with an offer to buy, not just license, the code for $50,000 plus sweeteners, including a perpetual license back to Seattle Computer, free OS updates, and favorable pricing on Microsoft computer language products. Brock declined the richer LifeBoat offer and signed with Microsoft because he saw more long-term future benefit from receiving free updates and most-favored-nation status. Besides, he was a hardware guy—software was just an enabler for him to move his CPU boards.
While today it seems like the deal of the century for Gates, at the time it was a substantial risk for a small company like Microsoft in a very crowded field. “We had no idea IBM was going to sell many of these computers,” as Tim Patterson, the virtual grandfather of what would become known as MS-DOS, would later summarize. “Somehow, people seem to think we had an inkling it was going to be this big success. I certainly didn’t. So buying DOS for fifty thousand dollars was a massive gamble on Microsoft’s part.”50 As it turned out, the gamble would pay off in spades and cement Microsoft’s place in corporate history.
IBM and its eventual clones exploded onto the scene in the early 1980s. From a starting point of zero in 1981, they accounted for the majority of computers sold by 1986. Microsoft software ran on them all, generating solid revenue from every copy of DOS and BASIC shipped at a minuscule marginal cost. Gates plowed the money back into his business to create new applications and improve the operating system with future versions he’d dub Windows.
Some software industry roadkill, like Seymour Rubinstein, a former TV repairman whose once best-selling WordStar word-processing program would soon join the extinct ranks of the early software dinosaurs,51 chalked up Gates’s success in operating systems to “pure luck.” He surely spoke for the legion of second-handers following in the amoral footsteps of Rand’s characters like Dave Mitchum in Atlas Shrugged, the incompetent railroad superintendent who complained about “injustice,” “bad luck,” and “the conspiracy of big fellows, who would never give him a chance.” According to Rubinstein, “There was no foresight, no imagination, no brilliant maneuvers, just a lucky break caused by a combination of Digital Research screwing up and Seattle Computer Products having something which wasn’t very good that could be modified for IBM.”
Would he say the same about an Olympic sprinter who invested a lifetime in physical training, only to win gold after the race’s leader stumbled? For the moochers, there’s no imagination in seeing the future of computing, no foresight in building for multiple platforms, no brilliant maneuvers in striking elegant and creative deals to reach business goals. For the less competent it’s easier to chalk it up to lucky breaks than to view their own personal shortcomings in stark contrast to genuine brilliance.
Act II: As Far as Your Mind Will Take You
In John Galt’s speech at the climax of Atlas Shrugged, he says that “you must drive as far as your mind will take you—with achievement as the goal of your road.” Now, from a strong position in operating systems and computer languages, Gates was to set for the world a new distance record of how far a mind can drive.
Gates was eager to capture a share of the third layer in the computer software hierarchy: applications. With the ubiquity of Microsoft products today, such as Word, Excel, PowerPoint, and Outlook, it’s easy to forget just what a competitive fight it was in the quickly changing field loaded with frenzied competition from incumbents like WordStar, VisiCorp, Lotus, Symantec, Borland, Ashton-Tate, and MicroPro. Undeterred, Gates set out to beat them at their own game with better products, more features, lower prices, and higher quality.
For example, in 1988 Lotus 1-2-3 had 70 percent of the spreadsheet market, having itself taken the lead from once-dominant VisiCalc some years earlier. Microsoft Excel was seen as an also-ran with a mere 10 percent market share. Less than 10 years later Lotus’s share had fallen to 26 percent and Excel’s share had increased to 68 percent.52 It’s often assumed that Microsoft applications simply piggybacked off of their huge installed base of operating systems, but the truth is that these applications were separate software packages that had to compete on their own merits in a free market. Good products sold. Bad products died.
Nobody forced anyone to buy anything. Gates struck rigorous deals with computer manufacturers and distributors alike using the full weight of his business leverage to wring the best possible terms from his partners. Gordon Eubanks, one of the industry’s pioneers and past president of Symantec, said of Gates, “I do know of instances where he used his influence, but who wouldn’t? This isn’t a race where there’s a handicap. Bill doesn’t go around carrying a 100-pound sack on his back. That’s what some people think should happen.”53 But that didn’t stop resentments from smoldering in the minds of the envious—those willing to bend the police power of the state to the task of putting that 100-pound sack on Gates’s back. As Rand put it in Atlas Shrugged, those people thought “it was society’s duty to see that no competitor ever rose beyond the range of anyone who wanted to compete with him.”
To be sure, Gates played hardball. But he played by the rules—and ironically, he fueled resentments by insisting that others
abide by them as well. In one case, a deal to market Microsoft’s OS/2 network software with 3Com was moving too slowly for Gates, so he decided to bypass his partner and sell the product directly to end users—a right he’d carefully built into Microsoft’s contract with 3Com. 3Com’s Bob Metcalfe viewed it as a double-cross and counter to the spirit of their deal, though clearly within Microsoft’s rights under the contract. “The response that I got,” said Metcalfe, “was, ‘You ought to negotiate your contracts more carefully. You were stupid.’”
Alan Kessler, 3Com general manager, had a more sanguine reaction, saying, “We were adults. We were big boys and girls when we signed the contract. . . . We signed it on the belief that certain things were going to happen to the OS/2 market, they did not happen, and it was not incumbent on Microsoft to give us relief.”54 For Gates, business is war. If you weren’t prepared to abide by an agreement, don’t expect any quarter from his side.
“It is deep in the [Microsoft] culture that success is never guaranteed,”55 observed Pam Edstrom, Microsoft’s early head of public relations, neatly explaining Gates’s drive to win. It was a tough-minded approach at odds with some in the industry who viewed business more like a gentlemanly game of golf than a battle for survival. But you couldn’t argue with success. Or could you? By 1994 nine out of 10 computers sold were IBM PC compatibles and Gates was a billionaire many times over. But corporate envy had festered and spawned a government lobbying effort to restrain the juggernaut.
Yet not all competitors bought in. In 1990 a Federal Trade Commission investigator, tipped off by others in the industry, approached Scott Cook, chairman of Intuit, whose flagship Quicken program was facing stiff competition from Microsoft Money after a period of information sharing between the two companies. Cook said he told the FTC it was wasting its time investigating Microsoft. “They are pestering one of the best run companies I’ve ever seen,” he said, “a company that should be the model for American industry. . . . [W]hen you lose to Microsoft, it’s because you snooze.”56
I Am John Galt Page 18
