Startup: A Silicon Valley Adventure

by S. Jerrold Kaplan

  “This is it, Mother,” I said. “Call everyone together.” Soon my family gathered at the house. We sat quietly in the library, taking turns monitoring my father. The periods when his breathing would cease grew longer. Each time, those present would wonder whether he would start again. Eventually he stopped altogether, but there seemed to be no way to know for sure. My two brothers-in-law and I were the only ones in the room at that moment.

  “Do you think he’s dead?” Hal asked. None of us had ever been posed this question before—to determine that intangible, illusory boundary between life and death. The gruesome possibility of declaring it crossed too soon would be more than we could bear. This dreadful conflict hung in the air like a damp chill. Then I realized they were waiting for me—as Murray’s son, I was the only one with the moral authority to declare an end to the nightmare. I looked at my father’s motionless body one more time, then nodded my head once, as though ending a prayer. They both nodded in response, making it official: he was gone.

  I called the removal service, as I had arranged with Dr. Mash, while the others returned to the library to deliver the news. Again, my mother surprised me. She didn’t cry or break down. Instead, she seemed relieved, as were the rest of us. The strain ended at last, the uncertainty over. But she had one last request, to go in and say goodbye. I accompanied her.

  She entered the room, as if visiting a sacred shrine, and straightened his sheet, saying nothing. I could see only love in her eyes, no hint of grief or any other emotion. She was completely at peace. At that moment, I understood what marriage really meant—and realized that I might now be ready for it.

  Then she was distracted by his hand, which was extended over his head, gnarled and pointing. She held it lightly. “I always loved his hands,” she said. After a few seconds, she looked up. “I’m done,” she said abruptly. We returned to the library.

  While we sat waiting for his body to be picked up, the phone rang and I answered. Everyone looked at me to see who it was—a ritual that occurred every few minutes during the day. It was GO’s acting VP of marketing, calling for approval to hire a new director of corporate communications we had been recruiting.

  “Ah, maybe this isn’t the best time to discuss it,” I said in a low voice. “My father just died. How about if I call you back in the morning?” I hung up the phone.

  The scene seemed so ridiculous that mother started to giggle softly. Then my sisters joined in. “Could have been worse,” whispered my sister Amy. “At least it wasn’t for him.”

  “Why are we so quiet? It’s not like we’re going to wake him up,” my other sister added.

  Now we were all laughing out loud. Then my mother spoke. “You know, it’s kind of a shame that you die when you’re old. This would be a lot easier to handle when you’re young,” she said. With this, everyone was in hysterics—just as the doorbell rang.

  Standing on the front step were two men wearing the colorless, drawn expressions of undertakers. One was tall and stooped with a long, thin face; the other was short and bald. Both were as pale as wax figures, as though they came out only at night. “We’re here for the deceased,” the taller one said. Everyone was watching, trying to regain composure. But from our flushed faces it must have been obvious that we had been laughing, even if the men hadn’t heard us from outside. They looked at us as though we were crazy.

  The custom among Jews is to observe a period of mourning for seven days after the funeral. During this time, friends come to the house to pay their respects. Although my father was in no way religious, it seemed appropriate to honor this custom, which was expected by the Palm Beach community. This was the first time I got a chance to meet and talk with his circle of friends. What emerged was a picture of a man I had never known, one obsessed with breaking free of his ethnic roots to find acceptance in the broader society. He avoided synagogues, shunned Jewish fundraisers, and never spoke a word of Yiddish in the house, although I learned that this was his mother’s native tongue. Despite his efforts, he was always confined by his background, restricted to joining the Jewish clubs, the Jewish businesses.

  I came to understand what he had tried to do for me—why we had moved from the Jewish suburbs of Westchester to the polyglot world of Manhattan; why he had sent me to something called Ethical Culture school on weekends rather than to Hebrew school; why he rarely spoke about our family history. He had molded me not in his own image, but in the image of what he wanted to be. He understood that much of his personal pain was caused by his own awareness of being different, and he believed that the less ethnic baggage I carried around in my heart, the better I would be able to break through these barriers. To achieve his dream for me, he had to let me go.

  “Jim Cannavino has asked me to step in and advise him on this situation.”

  I was relieved to find someone who was at last willing to pay attention to making a deal. The endless technical requests had continued unabated, but Sue King steadfastly refused to handle contractual matters—unless she wasn’t getting what she wanted. It was now mid-January 1990.

  The caller was Mike Quinlan, a long-time IBM marketing executive who had recently returned from a tour of duty overseas. He was apparently between assignments, and Cannavino turned to him to pick up where Kalb had left off. There was only one problem: Quinlan had little idea of what had transpired to date, and knew even less about GO. He had simply been instructed to work something out.

  To hear Quinlan speak, you would think that IBM was an official arm of the Department of Defense. He used the military lingo of a Pentagon general, asking for background briefings, alluding to backchannel communications, and taking orders from the commander in chief. He had the broad smile and tanned appearance of a politician who got elected on the strength of his image and war record. I could envision him charging up a contested hill in North Korea, proudly carrying the Stars and Stripes aloft at his own peril.

  Sue King assured me that I was very lucky to have him assigned, because he was a “corporate vice president.” Like a bank, it seemed that everyone at IBM was a vice president, but she explained that this special title was bestowed on only sixty-four meritorious individuals, who were thereby afforded wide latitude to marshal company resources as they pleased.

  Quinlan was quick to tell me that he was empowered to make a decision on the GO partnership and that he was a close friend of Cannavino, whom he referred to as Jimmy. He made it clear that all communications with Cannavino were to go through him: he would decide when and how Jimmy was to be consulted, if necessary.

  Quinlan liked to conduct business in person, so I flew to the IBM entry systems division’s headquarters in Boca Raton. This conveniently allowed me to stay with my mother, as Palm Beach was only thirty minutes away. The night before our first meeting, an IBM courier arrived to deliver Quinlan’s opening proposal. It was frightening.

  IBM was to get 50 percent of all revenues we received from parties other than IBM and licenses to all of our patents and copyrights. Equally outrageous, IBM’s royalties to GO would be capped, meaning that after a period of time, IBM would have free use of our products. The promised $5 million investment was replaced with a $2 million prepayment of royalties. Gone was any hint of killing IBM’s own project and backing ours, as Cannavino had suggested.

  No sensible business person would consider these terms, because they were virtually guaranteed to put GO out of business. No one can afford to pay a sales tax of 50 percent and survive. I called Vinod Khosla, one of our outside directors, who had a well-deserved reputation as a shrewd negotiator, and read him the IBM proposal. I held the phone for a long moment while he considered what to do.

  At last he said, “Come home.”

  “What?”

  “They’re not serious. There’s no point in continuing the discussions.”

  “But that’s suicide!”

  “We can cut back to a small core of staff, which will stretch the cash for a few more months while we figure out some other strategy. Mayb
e we’ll go into some other business . . .”

  I figured I’d better call John Doerr.

  “Sounds pretty bad, pretty bad indeed,” John said. “You’re going to have to do something to turn them around. One way or another, you have to make this deal. GO’s survival depends on it.”

  I also sought out David Liddle. “The problem is that the cupboard is bare,” he said. “They aren’t going to tell you this, but someone discovered that the cookie jar is empty. IBM’s last quarter was a disaster—they’re going to post a major loss for the first time in decades and write off over two billion dollars in bad investments. The word inside is they might even have layoffs for the first time in history. If you want to make a deal, do it fast and make it cheap. We’ll have to find operating cash elsewhere.”

  I was stunned. The great IBM, whose international influence rivaled that of a world power, was headed for a fall? I chuckled at the thought of Silicon Valley being flooded with the resumes of IBM lifers with obsolete mainframe skills, as though the British royal family now had to find jobs waiting tables. But if IBM wasn’t going to be a sugar daddy for my company, despite all the big talk, I had to close some sort of deal fast and seek other investors if GO was to survive.

  The following day, I was led to a conference room with Mike Quinlan and a dozen supporting staff. We spent the first half hour on introductions, explaining who was who and why we were all there. It was clear that Quinlan was the only person present who was free to express an independent opinion.

  I began with a briefing on the nature of our business and the market. This was the first time he’d gotten the complete story, as far as I could determine. Then I started in on the terms, attempting to sweep them aside in a conciliatory manner. “Mike, let’s talk about the problem you’re trying to solve with this proposal. I’m sure you’re not looking to put us out of business. What’s the real issue here?”

  “We want to accomplish three things,” he said. “First, if we’re going to make you a success, we want a piece of the action. Second, we have to be sure we can meet our commitments to customers if you disappear. And third, if this thing is big, we don’t want to be paying you off forever.”

  I noticed that his points corresponded to the three main terms of his proposal: the sales tax, the intellectual-property rights, and the royalty cap. It soon became apparent that Quinlan wasn’t wedded to the actual proposal, so I surmised that he had just given the same directive to an assistant and instructed him to draw something up. Since he’d made no personal investment in the terms, I decided to try to win him over with something that sounded even more dramatic.

  “Mike, I think I can meet your goals in a much simpler way. I’m willing to give you an option, which you can exercise at any time, to take our technology and do whatever you want with it. If you exercise this option, you’ll have to pay royalties on a declining schedule for five years, then you can do whatever you want without paying us a cent.”

  Quinlan’s eyes widened. “Why would you want to do this?” he asked.

  “I want to build a long-term relationship, and I’m willing to bet that we can continue to earn your respect and support every day. If this will help build the trust that’s necessary for IBM to get behind us, then so be it.”

  Now his eyes narrowed. “And what do you want in return?”

  “I ask only three things. First, I want to convert the five-million-dollar loan we already have into advances on royalties, in addition to the two million you are proposing. Second, at our option GO can walk away from the relationship as well. You’ll get the source code, the technology you need to meet your commitments, but no rights to future versions. And third, you have to agree to continue to support our APIs.”

  From my perspective, a deal like this would have a number of advantages. It would get us out from under the intellectual-property security agreement. I wasn’t worried about their taking the source code and competing with us, because I doubted their engineers’ ability to make substantial improvements, much less to keep up with us. And most important, as long as we controlled the APIs, it didn’t matter what IBM did with the code, since its efforts would only reinforce our own in recruiting applications developers.

  Quinlan asked me to leave the room for a few minutes. I was hoping that no one in his entourage understood that the APIs, not the code itself, were the crown jewels.

  “I’ll take it,” he said when I returned. “Except for one thing. I’ll only convert three of the five million dollars in loans into advances. Two million stands.” He wasn’t going to be satisfied unless I agreed to his terms. I desperately wanted to kill the security interest, but I decided it was better to fight that battle another day.

  “OK, let’s do it,” I responded. He smiled approvingly.

  I spent the next two months in a mad scramble to sign the deal and collect the $2 million before our cash ran out. Whenever it came time to discuss details, Quinlan himself was often absent. He’d fly out to our offices, give a pep talk about the value of having Jim Cannavino stand up at the press announcement, and then depart to deal with other matters. He’d leave behind his negotiating entourage, who’d wear me down on the licensing terms, which were as onerous as those of the loan. By March, with only a few weeks of cash to spare, we signed the deal.

  As I dropped Quinlan off at the airport to return to Armonk, we talked about the “joint walk-away” provision—our mutual right to terminate the agreement and go our own ways. But he didn’t want to leave on this sour note. Standing on the curb at the departure level, his carry-on bag slung over his shoulder like a soldier’s rifle, he smiled and shook my hand as though he were presenting me with a medal. “Anyway, you and I both know that it will never come to that. Can you imagine the damage to IBM’s reputation if it tried to destroy a small, innovative company like GO? We simply wouldn’t do that.”

  8

  The Announcement

  IT HAS BEEN said that a man lives two lives, one his own and one on the lips of others. The image makers of Hollywood have refined this concept by eliminating the original, leaving only the glittering reflection. But the engineers of Silicon Valley have once again built a better mousetrap: they have found a way to take appearance and transform it into reality. The trick to establishing a new computer product is not to make it great, but to make everyone think it is going to be great before it’s actually delivered. Here’s why.

  All computer products are hopelessly interdependent, stacked like faces carved in an endless technological totem pole: applications designers exploit operating systems tailored to computers built around chips that implement protocols based on specifications developed in committees staffed by applications designers. Everyone in this evolving loop must guess which new technologies are most likely to provide a solid and enduring platform on which to build their own piece. The problem is that these commitments must be made long before the surrounding pieces are ever delivered, because the computer business moves so fast that waiting is suicide. So resources are allocated on the basis of promises and reputations, not observable facts—like oceanfront condos sold on the strength of a developer’s rendering. But once a foundation is chosen, it is always easier to work around flaws than to start anew somewhere else. So the successful players in this elaborate guessing game are those who are able to persuade the most people to bet on them, because this justifies enough continuing investment to shore up their goods before their supporters defect to another platform. Insiders know the sorry truth: PR is as important as R&D.

  In the computer business, getting your message heard is mainly a tug of war with a pack of skeptical yet gullible news hounds who make their living by tracking down stories before others do. Their job would be a simple matter of snooping and writing as quickly as possible, except for one nasty problem: different types of publications have different lead times—the interval between when a story is submitted and when it actually gets published. Some monthly magazines may require up to four months; a newspaper needs only
a few hours.

  This is the same problem that afflicts jungle animals, both predators and prey, who must find a way to share a common resource, such as a watering hole, without wiping each other out. As in the wild, a set of explicit but unwritten rules has evolved that makes it possible for everyone to survive: you don’t attack a gazelle while it’s drinking; you don’t break a story on which another publication has arranged an exclusive.

  None of this would matter to the people who make new-product announcements, except that no publication will print your story unless it thinks your product is news. As a result, you have to engage in an orchestrated plan of background briefings, lunches, press mailings, and conference calls, culminating in a newsworthy “event,” so that everything can hit the street at the same time. You do this by notifying the press that an announcement of interest is forthcoming, scheduled for release on a particular date. By convention, all publications agree to honor such dates as long as they can get the story directly from you in advance. This is called a news embargo, and it affords you the breathing room to explain your product to all interested parties without blowing its news value.

  It is the job of the PR firm of a high-tech company not only to know and obey these rules, but also to keep track of which reporters are hungry enough to break its client’s embargoes. Since reporters seem to change jobs as frequently as migrant farm workers, maintaining this blacklist is a thankless and time-consuming task. The PR firm is your guide through this jungle, knowing who will feed you and who will eat you alive. What makes life hard for PR people is that in a pack, human reporters may not always be as trustworthy as animals.