Startup: A Silicon Valley Adventure

by S. Jerrold Kaplan

  “But speculative engineering schedules have no business as part of a contract! We may need to change things around on a moment’s notice. For that matter, Sue may need to. Do you want us to have to call our lawyers and execute a contractual addendum each time?”

  “You’ll have to see Sue about that,” Kalb said.

  As the negotiations dragged on, I came to understand that GO was laboring under the cloud of what IBM saw as an unfortunate misstep: its own operating-systems agreements with Microsoft. IBM believed that Microsoft was a monster of IBM’s own creation, an arrogant, undeserving upstart that had temporarily gained control of a key piece of IBM software. Although IBM couldn’t seem to get the better of Microsoft until their current agreement expired, it could certainly avoid making the same mistake again with GO. If IBM was going to be instrumental to GO’s success, it surely wanted a high degree of control.

  Kalb introduced a series of contractual principles that sounded like articles from the Declaration of Independence, but in fact were little more than a polite way of saying that GO’s existence would henceforth continue at the discretion of IBM. There was the principle of linkage, for example: any infraction of any contract with any party, no matter how minor—such as missing a lease payment on a copier—gave IBM the right, under the security agreement, to call its loans and seize our intellectual property. Worst of all was the principle of freedom of action: regardless of what we were compelled to do for IBM, it would not agree to any specific restrictions on its future behavior, such as engaging in competitive projects, abandoning support for our operating system, or courting our customers.

  By late October of 1989, we were burning more than a million dollars a month, in part owing to the cost of supporting IBM. At that rate, the loans—after we finished collecting them—would only get us through February or March of 1990, which was barely long enough to raise money if the IBM deal fell through and I started immediately with other investors. It was now essential that I close the agreement without delay, whether I liked the terms or not.

  Just when I thought we had everything nearly settled, we turned our attention to the details of the intellectual-property security agreement. I was no stranger to legal work, but nothing in my experience prepared me for what was to follow. This innocent-sounding idea, which I assumed would merely be a clause in the loan agreement, turned out to be the most complex free-standing legal document I had ever encountered. Hundreds of single-spaced pages in length, it described in microscopic detail a convoluted tango of onerous procedures. We were obliged to deposit to an escrow account a copy of every program, design document, drawing, sketch, specification, rendering, screen art, chip design, schematic, parts list, example program, tutorial, slide, programming note, erratum, guideline, help screen—whether handwritten, printed, or electronic—along with any copyrights or patents, until the loans were fully repaid. In addition to the materials for the escrow, the IBM lawyers insisted that we send them a list every quarter of the names and home addresses of all personnel who worked on the project, identifying their specific contributions, accompanied by a “certificate of originality” signed by each person, attesting that they actually did the work themselves. To amuse myself, I would occasionally invent imaginary types of intellectual property, which the lawyers would proceed to vehemently argue must be included in the agreement.

  I reviewed the escrow requirements with Robert. “Jesus, you’ve got to be kidding,” he said. “What the hell do you think we should do?”

  I knew there was no way to hammer this back into something realistic, but I was also confident that it wasn’t ever going to matter. “Once every so often you should send a backup disk and the documentation over to the escrow agent,” I told him. “Who knows? Maybe someday we’ll have a disaster and be happy we saved all this stuff.” I couldn’t have guessed that we were only days from such a disaster.

  Despite my cavalier attitude, I was actually quite worried about the tar pit we had fallen into. I decided to call David Liddle, our new director, first thing the next Monday for his opinion. He returned my call the following afternoon from a pay phone in the basement of a hotel in Palo Alto, about ten miles south of our office. Ironically, he was just finishing up an all-day meeting with some people from IBM.

  He listened carefully to my concerns, then gave a characteristically colorful response. “It’s the sorcerer’s apprentice.”

  “Huh? You mean like Mickey Mouse in Fantasia?”

  “Yep. There’s no one responsible in charge, so things just keep getting out of hand on their own.”

  “What should I do?”

  “We need to get an adult to supervise the process,” David said. “I’ll call Jim Cannavino and see if he’s willing to step in and straighten things out.” Jim Cannavino was the general manager of a now defunct IBM division known as PS-LOB—personal systems line of business. David had become acquainted with Cannavino because David’s company, Metaphor, also had a partnership with IBM.

  “Thanks, David, I really appreciate your help, as always.”

  There was a long silence. Then he said, “Do you feel that?”

  At first I wondered what he meant. It sounded as though he was questioning my sincerity.

  “Feel what?” I said.

  “We’re having an earthquake.”

  “Nope, don’t feel it up here.” Then my chair started to vibrate ever so slightly, as though someone were rolling a loaded cart down the hallway. I instinctively looked at the plant on my windowsill, to see if the leaves were fluttering. Once or twice a year we had a perceptible earthquake, which was easily missed except for this subtle telltale sign. “Yep, now I can feel it.”

  “Wow, this is a really big earthquake,” he said.

  “No it’s not,” I replied.

  Then the line went dead. I figured he would call back in a minute, but then realized what was happening. The quake’s epicenter was to the south, and the waves were rushing northward. But David, in the hotel’s basement, stood on firm ground, and I was on the fourteenth floor of an office tower built on landfill.

  I grabbed the corners of my desktop to brace myself just as the first shock wave hit. It felt like a surreal explosion: no sound, no source, just a sudden jolt that pitched the entire building forward about a yard, as though it had been whacked on the other side with an enormous rubber mallet. This massive steel and concrete monument, topped with a postmodern pediment like a tombstone, leaned toward the bay at an impossible angle. From the floor-to-ceiling window in front of me, I could look down at the receding slope of the lower stories, which left me floating in air just outside where my office used to be. I was so startled by this sight that I didn’t think to move.

  Then the building lurched back the other way, jerking my chair from under me and rolling out the side drawers of my desk. I stood up, still clutching the desktop for support. There was no deep throaty rumble like you see in the movies, no hysterical screaming extras running for cover. There was only a horrible metallic groan that reverberated up and down the structure as the enormous steel skeleton danced to keep its balance. The building swung back and forth three or four more times, each time leaning a little farther and rolling the rubber base of my chair around like a bumper car until I managed to snare it with my left leg.

  On the street several hundred feet below, I could see the ground swell up and down, bouncing the helpless cars around like corks in open ocean. The streetlamps, arranged in a graceful row of slender aluminum shafts with outstretched arms, whipped to and fro like a chorus line.

  Then, as abruptly as it had started, the mayhem stopped. The scenery was frozen again, but there was one problem: I was still thrust precariously out toward the bay. For one crystalline moment, I hung suspended in space and time, wondering if by some seismic serendipity the entire building might actually fall over, ending my existence. As if to answer my question, the structure crept back to its neutral position with a deep, soulful gasp.

  I let go of my desk and look
ed around. Miraculously, there was no obvious damage to my office, though the power was out. I ran out the door and called down the two adjacent hallways, “Is anybody hurt?” People were emerging from their cubicles and offices, dusting off the snowy white powder that had flaked from the ceiling tiles.

  “Everyone’s OK over here,” Kevin called from his office at the adjacent corner of the floor. “All clear on this side,” someone yelled from the other direction.

  Robert emerged from his office, which was next to mine. “Wow” was all he said.

  “The server—check the server!” Kevin yelled from down the hall.

  “Oh shit, the server!” Robert echoed. This was the nerve center of our local area network, the main electronic repository of all our work. If it was damaged, we could be seriously set back. Not only would we lose everything since our last backup, but it could take weeks to reconstruct its directories and files. Kevin beat us to the small utility room where the server was housed. The folding table on which it sat was still standing, but covered with fallen manuals and other assorted debris. The screen that normally displays its status was blank, even though it had switched over to backup battery power. Kevin examined it with the professional touch of a physician. “It’s OK,” he announced. The assembled crowd gave high-fives all around.

  We immediately vacated the building, in case of aftershocks. When I got in my car, I had my first indication of just how serious the damage might be. I turned on the radio and found that most stations had been knocked off the air. As I inched my way back to the city, through streets with no working stoplights, I suddenly felt ashamed of myself. I had been secretly thrilled by the solitary peril of that final, magic moment.

  Within a few days, power was restored to the building and we returned to work. There was a message from John Kalb waiting on my voice mail, asking me to call him back right away. I immediately returned the call.

  “Cannavino wants to see you,” he said. His tone was grave and deliberate, as though delivering the word of God. Clearly, David Liddle had gotten through. I was pleased, not only because he had pierced the IBM armor, but also because I hadn’t heard from him since our aborted phone call, and it meant he was uninjured.

  “Cool. When can I see him?”

  “There will be two meetings. The first of these will be for one hour, in the IBM executive trailers next to the main conference hall, the Tuesday of COMDEX, and the second one will be two days later, at IBM worldwide headquarters in Armonk, New York.” His formal phrasing made COMDEX sound like a week-long religious holiday. Perhaps this was his way of expressing disdain for my casual response to this rare opportunity, or maybe he was just annoyed at being cut out of the loop.

  “What does he want to talk about?” I asked. “Should I do any special preparation?”

  Kalb paused as though deciding whether it was safe to discuss this over the telephone. His voice took on a mysterious tone. “Come alone. Make no special preparation. Just be ready to talk,” he said. “And there’s one other thing.”

  “What’s that?”

  “Don’t tell Sue King. In fact, don’t tell anyone at IBM that this is happening.”

  I caught an early flight to Las Vegas that Tuesday. Much of the road into town was nothing but barren, empty desert, dotted with billboards tempting pious visitors with forbidden delights. Even by midmorning, a relentless dry heat caused these lures to shiver like devilish mirages.

  I arrived at the convention center thirty minutes early, only to consume the extra time walking from one end of the building to the other. To my amusement, I passed through the same door that John Doerr and I had used to escape the previous year, and followed the signs to IBM Conference Trailer B. I was ushered into a small white room with orange-beige shag carpeting. A folding conference table and six stackable chairs filled the room. In one of the two small windows, a large air conditioner slaved loudly in a futile attempt to stave off the desert heat, making conversation difficult.

  Jim Cannavino entered and greeted me. He was a compact, broadshouldered man with neat, jet-black hair who looked and talked like an earnest plumber. We proceeded to use up about forty-five minutes of our hour getting to know each other. I talked about my vision for GO and pen computing; he told disarming anecdotes about himself, his career, and the company. “I started by repairing keypunches,” he said, “and got to where I am by bitching about what a lousy job my boss was doing, until they gave me his position. I bitched about the PC division, and here I am, heading it.”

  When he noticed that my time was running out, he got to the point. “I want you to know that I really like your stuff. Now IBM”—he gestured expansively, like a king surveying his land of plenty—“does not have a good history with small companies as partners. We tend to bog them down in bureaucracy. But the partnership with GO is one that I want to see succeed. I’ve had a look at the plan our people are putting together, and I’d like to know what you think of it.”

  He stopped and looked at me. I assumed that he was stating this rhetorically, and that after a dramatic pause, he would continue. I couldn’t imagine that he actually wanted my opinion of this plan, when his own staff had at best taken no steps to share it with me, and at worst were actively trying to keep it from me. Since they needed a great deal of our help to make the project credible, of course, I had a pretty good idea of what they had in mind, and thought little of it. Unconvinced of the idea that pen computers were really a new class of machine, they were planning to add our technology to OS/2 and shoehorn it into modified laptops, which would have a keyboard that could be detached when you wanted to use the optional pen. But this was fine with me, because if they followed through, it would leave the field wide open for GO’s lightweight tablets, which customers would vastly prefer.

  “Its OK, you can level with me. It’ll stay between us,” he said, encouraging me to speak. Then it dawned on me that this was why he had insisted on secrecy, so that his people wouldn’t be able to lobby me into supporting their plan. Maybe Cannavino wasn’t such a bad guy.

  “Frankly, Jim, I think your team doesn’t quite get the point of pen computers yet,” I said. “They want to build hybrid laptops with pens, and perhaps that will work. But personally, I think they may not be taking the best path.”

  He leaned toward me and lowered his voice. “So you believe they’re wrong, then?”

  “I suppose so,” I said hesitantly. I wasn’t sure how he was going to react if I disagreed with the recommendation of his staff.

  He sat back and exhaled loudly. “Good. That’s why I asked you here, to get my organization off your back. All they’re going to do is throw rocks in your backpack, weigh you down until you sink. But I don’t just want your technology, I want you to call the shots.”

  I felt like a man stranded in the desert who was promised an ice-cold drink. I was so drained after enduring months of IBM negotiations that I could almost taste the freedom of pursuing the project unfettered by their demands. He could see this reaction in my face and immediately went for the close. “I’m ready to take a risk on you. If I have my way, I’d buy you right now, and then let you do what you want. How is that?”

  I was totally unprepared for this, and really needed some time to think. Working for IBM wasn’t the cruise that I had signed up for, but that probably wasn’t the point. The question was what’s best for the employees, the investors, and ultimately the customers. I was speechless.

  “Why don’t you think about it and we’ll talk again on Thursday in Armonk,” he said.

  “That sounds good to me. I’ll see you then.”

  As I walked from the trailer back to the conference center, I was uncertain about the right course of action. I thought about the team back at the shop, wondering whether they would think of an IBM acquisition as selling them out or setting us on the path to success. It was a very tempting offer, but it appeared that the price of freedom, financing, and IBM’s support was to sell my soul to Jim Cannavino.

  My next
move was to talk to John Doerr. I found him in a corridor of the Las Vegas Hilton, and we looked around for a place to sit. There weren’t any chairs, so John dropped to the floor and crossed his legs. I hunkered down and replayed my conversation with Cannavino.

  John looked exasperated. “Jerry, do you want to sell out?” he asked.

  “No, not particularly. I don’t think I have the patience to work for IBM.”

  “Right, then don’t give in to Cannavino. That’s not what we’re in this for.” John’s voice grew louder. “We’ve got work to do, and we can do it without them if we have to.” He was working himself into a frenzy.

  A hotel guard patrolling the hall stopped and eyed us suspiciously. There we were, two guys in suits having an animated discussion while sitting on the floor. He leaned over and said, “Are you the ones waiting for the girls?”

  John didn’t seem to understand what he meant.

  “Nope, sorry,” I said.

  As the guard walked on, John said, “Just tell Cannavino it’s not for sale.”

  Sometimes I wasn’t sure what motivated John. It certainly wasn’t money. To most investors, the prospect of a quick sellout to IBM would be irresistible. But for John, investing in high technology was a calling, and monetary reward was the byproduct of a successful mission. The higher goal was to create something enduring, a growing enterprise that delivers products, employs people, and enhances the wealth of its stockholders.

  Since I would be arriving late in the evening at New York’s Kennedy Airport after a full day of meetings at COMDEX, I arranged in advance for a car and driver to take me to the Holiday Inn in White Plains.

  I scanned the row of limo drivers waiting in the baggage claim area. They were all husky, tough-looking men with greasy hair, standing shoulder to shoulder and holding up little signs bearing the names of their charges, like an accidental police lineup. I pointed to a fellow with leathery skin and a frayed polyester tie. “You’re the one.”