An Unlikely Beginning

It’s worth noting that the primary issue discussed at our first serious meeting wasn’t what product we were going to market, or who was going to be the boss, or even how much money we wanted to make. The core discussions centered on what kind of environment we wanted to create and work in—one that would be fair and make sense. We didn’t know it at the time, of course, but that environment would promote values at the heart of a culture that enabled our unprecedented success. But that lay far in the future. The formidable path to starting a company from scratch stretched out ahead of us.

It wasn’t by chance that the three of us decided to start a company. Jim and I were both electrical engineers and had worked together on several different projects at Texas Instruments (TI) since 1972. I met Bill in 1977 when we both worked for TI in Austin. He impressed me with his marketing knowledge and insight. When I was asked to head a project to get TI into the Winchester disk business in 1980, I picked Jim and Bill to help me figure out what made sense for TI. During that project, the three of us were exposed to the venture capital culture of Silicon Valley, and it was a real eye-opener. We discovered, too, that we worked together well as a team, with Jim as engineering manager, Bill as marketing manager, and me as Product Customer Center (PCC) manager.

WE DIDN’T MEET AGAIN for three weeks. By then, our frustrations with our jobs had grown and we began to work on our business plan more intensely. None of us had ever created a real business plan, so I visited Houston’s public library and checked out several books describing the different parts of a business plan we would have to present. We decided that our first product would be an add-in Winchester disk and controller board for the IBM PC. Each of us worked on a section, carefully following the outlines presented. When it came to projecting shipment units and revenue, we were completely at a loss. I eventually made a guess so we could finish the document.

I decided to contact L. J. Sevin, a venture capitalist in Dallas I had met through Portia Isaacson, a well-known industry consultant who had visited my group in Houston in October 1981. At the end of her visit, I offered to drive her to the airport so I could ask her about venture capital. She mentioned that Sevin was a partner in a new venture firm. I called Isaacson and asked her to set up the introduction. After I sent Sevin a copy of our plan, he showed it to his partner, Ben Rosen, who was based in New York. They said they would take a look at it and get back to us.

Jim and I were scheduled to attend one of the computer industry’s largest trade shows, Comdex, in Las Vegas in mid-November. When we learned that Sevin and Rosen were going to be there, we set up a meeting in their hotel suite. We hadn’t gotten any feedback from them about our business plan or their interest in investing before we met with them, so we hoped to get some news. All they had to say was that they were still looking at it. But getting caught up in the furor of the PC market at Comdex led us to decide to resign from TI so we could focus on starting our company. We still hoped that Sevin and Rosen would decide to fund us, but we didn’t want to wait any longer to get started.

So we were on the verge of resigning when we ran into an unexpected problem. Bill’s wife, Maura, was very close to giving birth, and he was concerned about losing insurance coverage. We decided to wait for the baby to be born before leaving our jobs. After Bill and Maura’s daughter was born on December 2, there was a minor complication and Bill decided to wait to see how her condition turned out. But Jim and I, not wanting to wait any longer, resigned.

It was on the morning of December 4, 1981, that Jim and I each submitted a short letter of resignation. We gave the company a month’s notice. Our department managers tried to talk us into staying, but after a while, seeing that we were committed to leaving, they terminated our employment on December 14.

While we were still employed at TI, Jim, Bill, and I were careful to not work on our new company at the office. Often we would go to Jim’s house during our lunch break and work on various issues. On December 9, we were working during lunch at Jim’s house and were just about ready to head back to the office when the telephone rang. It was Sevin.

He had bad news. They had talked to VenRock, a well-known venture firm, about our idea. While the people at VenRock liked it, they had a conflict due to another investment. Sevin and Rosen had also talked to Kleiner Perkins, another well-known firm, and they were negative on the idea itself. Based on that information, he said they had decided to pass. Sevin, however, made it clear that he and Rosen liked what they had seen from us and would be interested in looking at a different product idea in the future.

We were very disappointed but, strangely, not discouraged. We discussed the possibility of changing our direction and staying with TI, but neither of us wanted to give up on what we had just started. We decided to put the add-in Winchester disk product idea aside and look for other, more interesting possibilities. We realized we had been very constrained in thinking of ideas that did not conflict with TI. As soon as we were no longer employed there, there were a lot fewer constraints on what we could consider.

We decided to each look into different product areas and find a more interesting idea on which to base our start-up. We were careful to only let Bill look at areas that we believed were not a conflict with his employer, but Jim and I were free to look wherever we chose.

We picked “Gateway Technology” for the temporary name of our new company, and on December 14 reserved that name in the DBA (Doing Business As) records of the Harris County Courthouse. Then we opened a bank account and deposited $2,000, the sum of $1,000 investments from both Jim and me. That made it seem like we were really starting a company. We just didn’t have a product idea yet.

Jim and I met almost daily to discuss our progress. We had a definite time frame to come up with an idea and get venture funding, or we would have to find jobs with existing companies. Our excitement about starting a company was tempered by the knowledge that we had set aside only six months of living expenses for our families.

Throughout December and early January, the three of us considered dozens of product ideas. Some were discarded right away; others we dug into more deeply. Then, on the morning of January 8, something unusual happened.

Sitting alone in my breakfast room drinking coffee, I was turning over the portable computer market in my mind and writing notes. I liked the fact that a portable wouldn’t directly compete with IBM’s desktop. I was looking for user needs that weren’t being met and ways we could differentiate ourselves from our competitors, particularly focusing on displays, storage, and styling. I had ideas about all three areas, but I couldn’t shake my biggest concern. No matter how good the product might be, a startup wasn’t going to be able to persuade software companies to adapt the most popular software applications to run on it. Without software, the product had absolutely no chance of success.

Suddenly, I was struck by a remarkable idea that was so simple and obvious it sent a chill down my spine. What if we could make a portable computer run the software written for the IBM PC? Could we design a rugged portable computer with a nine-inch display and professional styling and guarantee the best software would be available for it? The idea was so exciting, I couldn’t believe I hadn’t seen it before.

Then my engineering instincts kicked in. If it was so obvious, there must be some fatal flaw I was missing. I began listing the negatives:

WHY NOT DO THIS PRODUCT?

1. It’s already being done.

2. Product is too heavy.

3. Can’t build it cheap enough.

4. Can’t get computer stores to carry it.

Nothing on the list seemed insurmountable. After a while, I called Jim to get his reaction. He liked the idea, but suggested we continue looking at other ideas too. We also decided not to mention the idea to Bill yet, since it might cause a conflict with his job at TI.

Then I went back to my notes and wrote:

PLAN OF ATTACK

1. Key issue, “Is it already being done?”

2. Talk to Ben, Portia.

3. If feedback OK, proceed toward business plan.

I had a strong feeling that this portable computer idea was the one.

Jim and I asked Sevin and Rosen to meet with us the next time they were in Houston. They agreed to meet on January 20. We had planned to go over several potential products and get their feedback on which idea looked the most promising, but the portable computer idea continued to gain momentum. So we decided to put together a short business plan and focus the meeting on convincing Sevin and Rosen to invest in it.

Writing the product description didn’t take very long, but Jim and I felt words weren’t enough to communicate the concept of professional styling. We decided we needed a sketch to make it clear how different our product would be from the leading portable computer in the market, the Osborne I.

Jim called Ted Papajohn, an industrial designer recently retired from TI, and asked him to draw the sketch for us. A meeting was set for 10:00 A.M. on Monday, January 18, at the ComputerLand store on Westheimer, west of downtown Houston. I showed Ted both the IBM PC and the Osborne I. The idea, I told him, was to make our product look and operate like the IBM PC. While we would arrange the display, keyboard, and floppy disk drives in a manner similar to the Osborne I, we definitely didn’t want it to look like the Osborne I, which I referred to as “Army surplus.”

The three of us didn’t want to risk being overheard, so we walked across the street to the House of Pies restaurant for a cup of coffee. We asked for a table near the rear of the restaurant, but it really didn’t matter since no one else was there at the time. After ordering, we realized we hadn’t brought any paper or pencils with us. The waitress offered her pencil, but didn’t have any paper. No problem—the placemats were made of paper. We turned one over and Ted began to sketch the product we had described.

When we were finished, Ted asked us when we needed the drawing. “Tomorrow,” I replied, because we had a meeting in two days. Ted grumbled a bit, but said he would do the best he could.

The next day, January 19, I turned thirty-seven, but barely noticed. We were in the heat of the battle, and I was totally focused on creating a business plan that would convince Sevin and Rosen, or some other venture capitalist, to fund us to develop this product. That afternoon, Jim and I went to Ted’s house to look at his drawing. We liked what he had done, so he agreed to finish detailing the drawing. We could pick it up the next morning.

Our appointment with Sevin and Rosen was set for 2:00 P.M. in a room at the Hilton Inn on North Beltway 8 that Jim had reserved for the meeting. On his way there, he picked up the finished drawing from Ted and stopped at a Kwik Kopy store to have several color copies of the sketch duplicated. Then he met me at the Hilton Inn to grab a quick lunch and prepare for the meeting.

After the meeting we met back at Jim’s house to discuss our next move. We felt pretty good about the meeting, but our recent experience taught us to not count on Sevin and Rosen. We decided to set up a meeting with Lovett, Mitchell, and Webb, an investment firm based in Houston. We also decided to flesh out the business plan to include more detail about manufacturing, cost estimates, marketing, and distribution. We were pretty sure any other investor would want to see that kind of detail.

The following Monday, Rosen called me to say he and Sevin were still positive on the portable idea and were continuing to investigate it. I was disappointed they weren’t ready to invest, but encouraged to hear that they were still interested. Meanwhile, Jim and I continued to work on different parts of the business plan so we could go into more detail the next time we had an investor meeting.

The next morning, I started focusing on compatibility issues. As I thought about the way the software worked and how it dealt with the different parts of the computer, I made a list of the potential sources of incompatibilities. They included: the floppy controller, display controller, printer interface, async communications interface, keyboard, sound/ speaker, memory controller, and the ROM BIOS. When I looked at the list, I realized I had included every part of the computer! It finally hit me that every single part of the computer would have to be designed to function exactly like the IBM PC or software applications might not work. This is going to be harder than it first seemed, I thought.

By Friday afternoon, Jim and I had most of the additional materials for our business plan ready. We had a meeting the following Monday with Jim Callier of Lovett, Mitchell, and Webb. We made copies of the plan, which had grown from its original four pages to fifteen. I was starting to get a little concerned, since we still hadn’t heard from Sevin or Rosen.

At 9:00 A.M. on Monday, February 1, Jim and I sat down in Callier’s office and began to describe our product idea. I went over the key product features first, and then Jim talked about how he planned to make the product more rugged than other PCs. The meeting lasted about an hour, but we could tell that this was an area Callier didn’t know much about. As we left, he said he would get back to us soon.

When we got to Jim’s house, there was a message for us to call Sevin. I called him back immediately. He was going to be in Houston the next day and wanted to meet with us at the airport, since he had several meetings planned. We booked a room at the Host International Hotel and began to prepare.

At 1:00 P.M. the next day, we met with Sevin and he had some good news for us. He and Rosen wanted to invest in our company to develop a portable PC—but there was a caveat. Since he and Rosen were new at the venture capital business, they needed another firm to agree to invest along with them. He wanted us to go to San Francisco and meet with Kleiner Perkins Caufield & Byers. And he wanted us to have the meeting the next Monday. After thinking for a minute, I said, “So let me get this straight. You want to invest in us, but you won’t invest unless Kleiner agrees to do so?” Sevin replied that that was correct.

That wasn’t what we wanted to hear, but we were excited to have the opportunity to meet with such a well-known firm. Immediately, we went back to work adding more content to the business plan. Jim created detailed schedules for the development and manufacturing phases. I thought that we might need to present our plan to several people, so I began making slides for an overhead projector. I figured this would be a more formal presentation than those we had done so far. We planned to start with a description of the product, then discuss schedules, and finally present projections of development costs and other financials.

“So let me get this straight. You want to invest in us, but you won’t invest unless Kleiner agrees to do so?”

We contacted Bill and told him the good news. We also said it was time for him to get on board. The train was pulling out of the station and he needed to be on it. Bill replied that he was ready to go. His infant daughter’s health issue had cleared up. He would resign the next day.

Bill gave TI a month’s notice, but hoped it might terminate his employment sooner. His supervisor said he had a week to finalize his projects and hand them off. Unfortunately, that meant he would miss the meeting with Kleiner Perkins.

Jim and I flew to San Francisco on Sunday morning. We checked in at the Hyatt Regency Embarcadero, adjacent to the Embarcadero office building where Kleiner Perkins was located. In the afternoon, we walked around the area and talked about the meeting. The breeze was cold, but I had a warm feeling about where this was heading. We both felt we were ready.

At 9:00 A.M. on Monday, February 8, Jim and I walked into the reception area. A few minutes later, a boyish-looking John Doerr walked out to greet us and took us back to a meeting room. I was surprised that Doerr looked so young. I was starting to feel old at thirty-seven.

Doerr introduced us to Jim Lally and Brook Byers, two partners who would also be in the meeting. I began going through the slides using an overhead projector, but I didn’t get far before Doerr interrupted me with a question. The meeting became one long series of questions, with me using my slides only to address some concerns. Much of the time was spent discussing competition and how we planned to differentiate our product from others. The meeting lasted almost three hours.

Afterward, Doerr walked us back out to the lobby. He was very positive and said he liked our idea a lot, but it would be a few days before his partners could meet to discuss an investment. Jim and I were both emotionally drained after the intense questioning. We felt we had done a good job as we headed for the airport.

On Tuesday, February 9, Bill became free of TI and began to work intensely on the marketing plan. He met with Jim and me to go over where everything stood. We all agreed we needed his marketing plan ASAP.

On Wednesday, I received a call from Sevin and Rosen. They were willing to lead the initial investment round of $1.5 million, with $750,000 coming from them, $500,000 from Kleiner Perkins, and $250,000 from a third firm, L. F. Rothschild, Unterberg, Towbin. The investors would own 55 percent of the company, the three founders 25 percent, and other key employees 20 percent. I could barely contain my excitement. Then Sevin said that while he and Rosen were happy with Jim and me, they really didn’t know Bill that well. He wanted to meet with him to determine if he was capable of doing the marketing job.

The next morning, Bill and I flew to Dallas to meet with Sevin. The meeting started amicably, but soon turned heated as Sevin grilled Bill on all sorts of issues. Bill believed we should sell our portables through computer dealers and planned to set up a dealer council within four to six weeks. He also strongly emphasized the need to compete on features and quality, not price. Then Sevin fired more questions at Bill. When would he hire his first employee? How many people would he need to hire? What about the European market? Would the company quibble over warranty on abused units? What about charging for service? Who is the competition?

When Bill mentioned that Osborne was the main competition, Sevin wanted to know why he hadn’t mentioned IBM first. Then he wanted to know why Apple wasn’t considered major competition. I felt that Bill had handled all his questions well, but I had never seen this side of Sevin before. I was a little concerned.

Bill and I flew back to Houston that afternoon, and that night Sevin called. He said that he and Rosen didn’t think that Bill could handle the marketing job. They wanted to look for someone else. I thought about this for a moment. I knew Jim and I could work with someone else, but the three of us had worked closely on starting up TI’s Winchester disk business. I had a lot of confidence in Bill’s ability and judgment. I finally told Sevin I was going to stay with Bill, and hoped that they could get comfortable with that decision.

Sevin tossed the issue around for a while, but eventually agreed to go along with my decision. Then he shifted to other details. He told me the three founders could decide among ourselves how to divide up our 25 percent share of the company. He also said the board of directors would consist of himself, Rosen, John Doerr, and me, and we would meet at least monthly if not more often. We agreed to meet in Dallas the following Monday to sign documents and officially start the company.

Jim, Bill, and I met on Friday to finalize a few remaining details. We decided to split our 25 percent ownership evenly. Jim and I had discussed privately whether Bill should have the same equity as us, since he had not been involved in the development of the business plan or the process of selling the idea to the investors. We quickly decided that Bill’s contribution was going to be just as important to our success as ours, so he should have the same ownership percentage, 8.33 percent. We also decided to give our early employees an amount of stock based on our expectations of their contribution. In a start-up, every employee has an important role to play, and we wanted everyone to share in our success.

The last issue was finalizing each of our roles in the company. Since I had five years of experience as the general manager of three different Product-Customer Centers (PCCs) at TI, I became president and CEO, Jim vice president of engineering, and Bill vice president of marketing, the same roles we had played when we worked together previously.

The three of us flew to Dallas on Monday morning and met Sevin and Rosen at the law offices of Merlin Samples. There were a lot of documents to sign, but we were so excited the time went by quickly. That day was a national holiday and the courthouse wasn’t open, so papers couldn’t be filed. Gateway Technology, Inc., was officially incorporated on the following day, February 16, 1982.

With money in the bank, we hit the ground running. We rented office space in the Allied Cypress Bank building on Jones Road in northwest Houston, but it was going to take several weeks to complete the remodeling. Another part of the building had been vacated recently, and the rental agent offered to let our new company use that space until our office was completed. There hadn’t been time to buy furniture, so we brought in folding chairs and tables to begin working. We had a single phone line installed with one phone on a very long cord that we all shared.

But the main issue was hiring engineers. We didn’t want to get our former employer upset, so we decided not to actively recruit current TI employees—but many did come to us. Most of the people we had worked with before we left TI were aware we were planning on starting a company, and many had contacted one or more of us to let us know they wanted to apply for a job when we got started. We also decided not to tell applicants what our product was until we offered them a job. We had been trained well by TI about the importance of keeping our new product plans secret.

In addition, we decided we shouldn’t offer jobs at the same time to all the engineers we wanted to hire. We picked Steve Ullrich and Ken Roberts for our first offers, because they needed to start analyzing the IBM PC. Both Ullrich and Roberts were electrical engineers in their early thirties and had joined TI right out of college. Jim had worked with them on various electronic design projects over an eight-year period. At the time Jim left TI, they were designing a Winchester disk peripheral product to go along with TI’s minicomputer products. There was a strong mutual respect and trust among them. I also knew, and thought highly of, both of them.

We were planning on waiting a few days to make our third offer to Gary Stimac, another electrical engineer who had joined TI in 1972. Stimac initially was a software programmer who had worked for me on the design of the TI intelligent terminal Model 742, one of the first products to use the 8008, Intel’s initial 8-bit microprocessor. I thought he was one of the best programmers I had ever met. Recently, he had been working with Jim on TI’s Winchester disk project. When Stimac heard that Ullrich and Roberts had turned in their resignations, he decided he couldn’t wait any longer and turned his in as well. He hadn’t received an offer from us yet, and he didn’t realize he was messing up our plan to spread out our hiring, but it didn’t really matter. Immediately, the new company doubled from three to six employees.

At that moment, we were six people with a huge challenge. We needed an actual IBM PC to work with. Stimac was immediately dispatched to Dallas to buy one. He returned with the PC and the technical reference manual that contained detailed specifications on creating hardware and software products to work with it. Stimac, who was assigned the task of creating the ROM BIOS for our computer, discovered the entire IBM BIOS ROM code was printed in the manual. His immediate reaction was “Great, my job’s been done for me! This is going to be easy.”

He couldn’t have been further from the reality that lay ahead.

Over the next few weeks, our team grew rapidly with the addition of four key engineers, three of whom were hired from TI. They were: Steve Flannigan, software; John Reilly, plastics; Walt Russell, aluminum chassis; and Bill Bray, power supply. Russell was working independently at the time, but we had known him from TI. They all gathered around the IBM PC we bought and, along with Stimac, Ullrich, and Roberts, began dissecting the computer. While I watched them work, I realized that we had put together an all-star team. These men were the best in their fields over all others I had met in my career.

As soon as Flannigan came on board, he and Stimac worked with an attorney to plan a legal way to reverse engineer the BIOS ROM of the IBM PC. To be safe legally, they learned that anyone who looked at IBM’s code couldn’t write any of our BIOS code. Stimac had unknowingly contaminated himself by looking at the code printed in the IBM manual, so his first job became writing a specification for our BIOS. Then Flannigan took his spec and began writing our code, which left Stimac with the job of working with Microsoft on MS-DOS and BASIC. We had discovered that the version of MS-DOS Microsoft was selling to everyone at the time wasn’t compatible with PC DOS, so we decided I needed to set up a meeting with Bill Gates, Microsoft’s CEO, and ask him for a version of MS-DOS that was totally compatible with IBM’s PC DOS. We didn’t know it yet, but we had just run into a potentially fatal flaw in the product idea our new company was based on.