Hard Drive: Bill Gates and the Making of the Microsoft Empire (16 page)

Smith had a bachelor of science degree in marketing and an M.B.A. in finance. While in the service he had taught electronics in Colorado for roughly ten years. In January of 1977, Smith went to work for Tektronix, a $1.2 billion Oregon company that made engineering equipment. Smith was hired to help the company start a microcomputer development group. When the Tektronix group needed computer software languages, consultant Adam Osborne recommended Gates and Microsoft. Smith flew to Albuquerque to negotiate a license for FORTRAN, which Steve Wood was developing. Later, Smith negotiated with Microsoft to supply Tektronix with Pascal. (Named for the seventeenth century mathematician and philosopher Blaise Pascal, Pascal handles complex programs faster than BASIC and is often the computer language of choice for business programmers.)

Around Tektronix, Smith was regarded as something of a maverick. In late 1978, Smith told Wood that in another six months he would be ready to leave Tektronix. Microsoft, Smith said, was the kind of company he wanted to work for.

“It became apparent to me,” Smith explained, “. . . that the guys at Microsoft really had their act together, but may not have really known it themselves. It wasn’t obvious to me that they understood how to leverage the business. They had the contracts with Apple, with Radio Shack, but it wasn’t clear they were turning it into an actual business. ... I was amazed at how successful they were without what appeared to be any real business management.”

Smith recognized that Microsoft was a company oozing with entrepreneurial talent. He decided that if a business expert such as himself joined the company, Microsoft could be even more successful. “From my point of view,” he said, “they were just lucky, and if I got involved, we could probably add some discipline on the business side. It didn’t take me very long to realize they really weren’t lucky. Gates really understood what he was doing.”

That realization came to Smith not long after he started working for Microsoft, when Gates came into Smith’s office one day and shouted at him, “How can you possibly take this much time working on this contract? Just get it done!”

Recalled Smith of that short but educational meeting with his much younger boss: “I think what I realized was that I needed to focus, that the money and the opportunities were simply there, [and I needed] to close contracts with customers. So I focused on personal performance over management. Initially, I was dealing more with management issues, as a guy with an academic background coming out of a large company. But it only took me a couple of meetings to realize that personal performance was what mattered.”

In hiring Smith, Gates was hoping to bring more credibility for Microsoft to the negotiating table. At 34 years old, Smith was ten years older than most of the people working for the company, with the exception of Bob O’Rear. Gates was only 23 and looked more like 17, which he recognized was a liability in many business circumstances. Gates told Smith during the job interview that he was seeking someone who
looked
older.

Smith was the first nontechnical person with the company to do OEM sales, which mostly had been handled by Gates. Smith had only been on the job a few weeks when he checked Microsoft’s bank balance and discovered ten
certificates
of deposit for a hundred thousand dollars each. After that, he never worried about his marketing budget.

When it came to making deals for Microsoft, what Gates may have lacked in age and experience he made up for in
self-confidence
and business acumen. Usually, when OEM customers came to Microsoft, they were ushered into Smith’s office. Gates would come in a short while later. Smith would be dressed in a suit; Gates would often be wearing whatever he had on from the day before after pulling an all-nighter in the office. “They always thought he was one of my tech support guys, until I introduced him,” Smith said. “I think they were impressed by his youthfulness and his technical knowledge, and frankly, as soon as he started talking the credibility was there. The credibility wasn’t there until he started discussing what Microsoft could do, and what BASIC could do.”

In the fall of 1979, Gates and Smith made a trip to a division of the Xerox Corp. in Dallas, Texas, to negotiate a licensing agreement for Stand-alone Disk BASIC. This BASIC performed many of the functions normally handled by the computer’s operating system, and that’s what Xerox was looking for. At the time, Xerox was working on Project Surf, a joint venture with a company called Convergent Technologies, to develop a personal computer. It was a big deal, the biggest deal Microsoft had taken a shot at. When Gates walked into the Xerox conference room, he was taken aback by the august surroundings. Seated around a huge table were more than a dozen of the company’s executives and technical people, dressed in expensive suits and ties. Gates leaned over to Smith and whispered, “This is the biggest conference table I’ve ever seen.” Smith whispered back that there was an even bigger table at Tektronix.

The meeting was a bit of a revelation for Gates. Prior to this time, many of the computer companies he had dealt with were little start-up operations, manned by pie-eyed visionaries like Gates himself. They related. Now, Gates was dealing with some of the biggest corporations in America.

But if that caused his heart to race, no one around the table could tell. Within seconds of launching into his presentation,

Gates had the attention of the room, charming the room with his knowledge of the product and complete grasp of the work at hand. Years later, a person who had attended the meeting as a Xerox executive said Gates was “cool as a cucumber. I thought he did this every day for a living.”

It turned out to be one of the easier deals to cut, although Project Surf never produced a marketable computer. Before the day was out, Microsoft and Xerox had signed a $150,000 contract—roughly double the size of Microsoft’s previously largest contract.

Gates hated to lose business deals. He told Smith that when you lost a deal, you lost it twice—you didn’t get the money, and the other company did. If the deal were worth $50,000, then you essentially lost $100,000 because that was the difference between what you could have had and what someone else walked away with. “We had one rule that we both agreed on,” recalled Smith. “We would never lose the deal.”

For all his native acumen, Gates was also rough-edged and inexperienced in the art of the deal. Gates’ style was to browbeat customers until they wilted and acquiesced. But he often was so intense in negotiating sessions that he would push too hard and actually jeopardize the deal. A former top Microsoft: executive said there was “almost a viciousness” to the intensity Gates displayed when trying to secure a deal with another customer. “I’ve seen Gates lose a deal in negotiations because of it,” the executive said. “Most sales guys get more gentle. He gets mad at a customer.”

After working together for a while, Gates and Smith developed a negotiating strategy to offset their shortcomings. Gates was cast as the bad cop, the one with final authority to agree to a deal. The more experienced Smith was the good cop, the one who handled the actual negotiations. He would try to coax every last dime out of customers, and if talks stalled over a penny, he would plead helplessness, claim he could do no better than the offer on the table—Gates, after all, was the guy calling the shots.

The tactic, and the fact that Gates was loath to let any contract get away, was productive. But there were potential pitfalls. Gates and Smith once blew an important deal with an Intel subsidiary after Gates let his temper flare unchecked. Smith and Gates, in separate offices at Microsoft headquarters, had gotten Intel on the phone for a conference call. In the course of the conversation, Gates became upset over a bad reference Microsoft had received from another division within Intel—something about Microsoft being late with delivery. Normally Smith would break into the conversation and smooth things out before tempers got out of hand.

Smith, however, was called away from the phone without Gates’ knowledge, and before he could intervene, the Intel representative had been chewed over by Gates pretty thoroughly. When the shouting was over, the company decided to buy software from other sources. It could have been a costly lesson, except that the Intel subsidiary eventually disbanded and after six months of rebuilding bridges, Microsoft won back the contract. Microsoft eventually licensed all its languages to Intel.

Despite his occasional temper tantrums, Gates was a great salesperson. Microsoft overcommitted itself and set unrealistic deadlines, but as far as Gates was concerned, it was more important to get the sale and worry about the consequences later. And Gates had supreme confidence that he could handle those consequences.

The performance of Microsoft’s small cadre of programmers was ragged in the company’s early years. Deadlines were often missed, products weren’t always well designed, and contracts had to be revised due to unforeseen obstacles or delays. What sustained the company was not Gates’ ability to write programs, to “crank code,” despite his vaunted grasp of detail and of the nuances of communicating with slivers of silicon. Gates sustained Microsoft through tireless salesmanship. For several years, he alone made the cold calls and haggled, cajoled, browbeat, and harangued the hardware makers of the emerging personal computer industry, convincing them to buy Microsoft’s services and products.

He was the best kind of salesman there is: he knew the product, and he believed in it. Moreover, he approached every client with the zealotry of a true believer, from the day he first articulated the Microsoft mantra: “A computer on every desktop, and Microsoft software in every computer.”

Executives from Japanese or European computer manufacturers would visit Microsoft’s unprepossessing offices in Bellevue, Washington, often intending to strike fear into Gates because his programmers were dangerously behind in delivering on one contract or another. Silver-haired men of weight arrived in their power suits with their pencil-packing entourage ready to lay down the law. There to greet them would be Gates, looking 17-years-old, wearing whatever pizza-stained T-shirt he had had on the day before, dirt clouding the lenses of his glasses, unshaven after an all-nighter at the office.

Before they could fully digest the appearance of the boy whom they had not long ago entrusted with a part of the company’s future, before the subject of breach of contract could be suggested, Gates would launch a counteroffensive.

“Bill would sort of paint the picture of where Microsoft was and what the important issues were at the moment and what was driving him, and then somehow go from there into how that was going to lead into new developments,” said a former Microsoft programmer. “He’d say, ‘We feel real bad about what’s late and what hasn’t been done, and here’s what we can do for you,’ and then he’d paint a grandiose picture about what Microsoft could do, the Gatesian vision.”

“What it would amount to was selling them more stuff. ‘We’re sorry we haven’t come through on this smaller delivery, but look, we’ll get you in on this huge part of our vision, and that will put you in an even better position.’ What started out as a complaint about a smaller deal would turn into the sale of a much bigger contract.” He was confident, sometimes brazen. He could be impressed. He just couldn’t be fazed.

As Microsoft headed into the Eighties, software development tools such as BASIC were the company’s bread-and-butter products. Every few weeks or so, Kay Nishi flew first class between Tokyo and Seattle, bringing in more business. Gates, who always flew business or coach, didn’t particularly like the high air fares Nishi was charging to Microsoft, but Nishi was more than making up for his lavish spending habits.

“We were selling BASIC to the Japanese like crazy,” said Marc McDonald. “We were selling Disk BASIC for Japanese machines, and that typically would bring us anywhere from $150,000 to $200,000 a pop.”

Microsoft had ended 1979 with about $4 million in annual sales, with much of the money coming from BASIC. But BASIC represented only one layer of the essential software needed on every personal computer. At the top end are application programs such as word processing, which turn the computer into a useful machine for the public. Languages such as BASIC are the middle layer, providing programmers with the tools they need to develop software. The bottom layer of software, without which the computer is virtually useless, is known as the operating system. This is a low-level language that actually runs the computer, performing the most elementary tasks required of the microprocessor in the arcane world of machine language. If the user wants to record data onto the computer’s disk, for example, the operating system finds free disk space.

During the early years of the personal computer revolution, there was no standard operating system. Almost every new machine that came on the market used a different system to control the microprocessor. But gradually one operating system known as Control Program for Microcomputers, or CP/M, became something of an industry standard by 1979. It was developed by Gary Kildall of Digital Research.

In February 1980, Microsoft entered the operating systems market, too. Gates negotiated with AT&T and acquired the license for a standard version of its UNIX operating system, to be sold under the Microsoft name of XENIX. This multi-user operating system had became popular in the 1960s when Bell Labs made it available on university computer systems. Microsoft adapted its version for the new 16-bit microcomputers. “We were actually introducing XENIX while we were still negotiating the contract with AT&T,” recalled a Microsoft manager. “But we had a lot of confidence that we had the contract.”

Every personal computer needed applications, a high-level language like BASIC, and an operating system. Microsoft now had its fingers in all three layers of the personal computer software cake. Only the company’s languages had become an industry standard, but Gates had high hopes for XENIX. As things turned out, Microsoft’s operating system
would
become the industry standard, but this system would not be XENIX.

In acquiring a license to distribute UNIX, Microsoft had gotten its hands on an operating system that someone else had already spent the money to develop. Any computer company could license UNIX from AT&T, and a lot had already done just that. But Microsoft was the first to focus on microcomputers and did so through a master distribution agreement. Gates received a sliding-scale discount from AT&T so Microsoft could aggregate all the royalties coming in from the other companies it sublicensed XENIX to. Microsoft did not even spend a lot of time or money “porting”, or adapting, XENIX to other computers. It hired a small California company known as Santa Cruz Operation to do that. They performed much of the technical work, and all Microsoft did was pass down part of the royalty money. It was a good deal for Santa Cruz, a better one for AT&T, and a great one for Microsoft, which now had an operating system and was making money with very little investment of capital.

One of Microsoft’s first customers for XENIX was a software company known as 3Com Corporation, which had been founded in 1979 by Bob Metcalfe in Santa Clara, California. Metcalfe wanted to use XENIX for his company’s initial product, networking software called TCP/IP that linked together personal computers.

Gates and Metcalfe had met each other for the first time in Chicago in late 1979, a short while before they did business together, when they appeared at an industry soiree to talk about the future of personal computers before executives of Sears Roebuck & Co. The function was organized by a crusading industry visionary named Blair Newman, whom Gates knew from Harvard.

During his talk that day in Chicago, Newman suggested that Sears establish high-tech stores across the land to sell computerized home appliances, which could all be controlled electronically by what he called a “Home Bus.” (In this case “bus” referred to the path along which digital information was transmitted between a computer’s microprocessor and hardware.) Newman continued to pitch his Home Bus idea at the Consumer Electronics Show in Las Vegas in January of 1980.

Newman was brilliant but impatient and unstable. He led a tragic, troubled, drug-addicted life until killing himself in 1990. He had entered Harvard Business School in 1975 at age 28, with the highest test score of anyone ever admitted. When Gates wasn’t monopolizing the school’s PDP-10 to further enhance BASIC for the Altair, Newman was using the computer to study a Las Vegas hotel owned by Hughes, a project which later landed him a job with Hughes’ holding company, the Summa Corporation. In 1979, after working briefly as a consultant for Apple Computer, Newman formed Microtype, a company that planned to produce computerized typewriters. Newman persuaded Gates and others to invest about a quarter of a million dollars in the business. Gates and Steve Ballmer joined the board, as did Metcalfe, as a favor to Gates. But Microtype, which was operating right down the road from Metcalfe’s 3Com, soon went defunct. Shortly before the company went under,