Startup: A Silicon Valley Adventure

by S. Jerrold Kaplan

  Gates agreed. “It’s only a matter of time until they can’t keep up,” he said ominously.

  Ballmer turned his attention to me. “I imagine you’re going through the same things we did with IBM.”

  “Probably.” I kept a poker face.

  “We were ultimately forced to make a hard decision,” Ballmer said. “Whether to build the right product or do what they wanted. It was very emotional for us.”

  After a brief silence, Gates spoke up. “I hear that your system is quite a bit bigger than you thought.”

  I was amazed that he would know this well-kept secret. To our horror, the latest version of Penpoint was far larger than we had anticipated. In fact, it wouldn’t even run in four megabytes of memory, so we had to load up the demo machines with eight. This would be a critical shortcoming if it couldn’t be fixed, as it would drive up the cost of the hardware significantly. Most of our programmers were working night and day to figure out what went wrong, and to squeeze it back down. But virtually no one outside our shop knew of the problem, except for a few of our most important IS Vs.

  “Not particularly” was all I said.

  The black-tie awards ceremony was even more glitzy than the one the night before. Lasers and confetti made it seem like a cross between a rock concert and a game show, and was clearly produced for the television audience. This time, Penpoint won the Standards and Systems Software Technical Excellence Award.

  Later in the week, I attended an enormous IBM party celebrating the release of the latest version of OS/2. The only problem was that the product had been delayed by several months, and so wasn’t being released at all. But that didn’t stop IBM from staging a festival of self-congratulation, during which a stream of executives toasted the development team from the podium.

  Afterward, I was cornered by a group of four or five Penpoint ISVs. They were plainly agitated. “Have you heard what IBM is telling us now?” one of them said. “They’re saying Penpoint is for vertical markets only—like the State Farm application—and we should target our work toward their new secret project, Pen OS/2.”

  I said, “This is a joke, right?” But they all shook their heads in unison.

  I immediately tracked down one of the senior engineers on the OS/2 team who had participated in a number of Penpoint evaluation meetings the previous year. “Rumor has it you’re developing pen extensions to OS/2. Is this true?”

  Fortunately, no one had bothered to tell him to deny it. “You mean our people haven’t told you? That’s why we’re so adamant about getting rights to your ‘look and feel’ for our other operating systems.”

  I quickly found a bank of pay phones in the hotel lobby, and was on a conference call with Bill and Randy within a minute to relay the news. The greasy metal phone cord made a sound like a machine gun as I pulled it around the last plastic courtesy panel to get some privacy.

  “Just stay cool,” Bill said. “We still haven’t closed on the loan of ten million. Then we can work on finding a real partner.” I slammed down the phone with such force that I nearly broke the little plastic tongue out of the handset cradle.

  As I walked away, I passed a hotel security guard lounging nearby. “What’s the penalty for murder in Nevada?” I asked.

  “Life—but figuring in good behavior, you’re eligible for parole in ten years,” the guard said. “It’s better to hire someone. Want me to look into it?”

  I laughed, because he seemed to take me seriously. It reminded me that this was only a game. A very big game, with serious consequences.

  Bill banged his fist on the table. “Financing has to be our number one priority. Let’s talk about options. And don’t mention layoffs on my watch—the next CEO can try that solution.”

  Robert, Kevin, and I knew the seriousness of our financial problems, and we had long since grown accustomed to living periodically on the edge of extinction. For the other executive staff, however, staring into the jaws of death was a new experience. They reacted like any sane person who suddenly faces his own mortality. Initially, they went into denial—continuing business as usual, planning off-site staff meetings, interviewing new hires, and visiting customers—figuring something would bail us out. Then came acceptance. Mike Homer had quietly started thinking about which of his people he could afford to lay off and which could easily find other jobs on their own. Eventually panic set in.

  “Our options are pretty limited,” Randy said. He was putting it mildly. We all knew that with almost two hundred people on the payroll, we were spending nearly $2 million a month, but it sure sounded different when you realized that AT&T’s $10 million would last less than six months. With Penpoint Japanese, Penpoint Hobbit, and Penpoint Intel—not to mention applications development, ISV recruiting, customer sales, hardware manufacturer support, marketing communications, public relations, technical support, documentation, quality assurance, product marketing, field sales offices, GO Japan, training, human resources, legal, accounting, and facilities—everyone was working around the clock and still screaming for help.

  Randy continued. “We can cave in to IBM on the loan terms, but this means handing them the right to rob us blind. We can ask our Penpoint licensees for advances, but they’re likely to see it as a sure sign we’re in trouble, and run the other way. Then there’s AT&T. I suppose we could see if they’re interested in making an additional investment.”

  Bill pursed his lips. “No way. Not while we’re fighting tooth and nail to win the Telepen RFP.” He was referring to a Request for Proposal issued by AT&T’s general business systems division, to design an operating system and a standard user interface to run on a wide range of computer equipment planned for the future, including all pen-based devices.

  Since the government had forced AT&T to divest itself of the regional Bell operating companies ten years earlier, the company’s management had wisely granted substantial autonomy to the remaining divisions and fiefdoms in an effort to sort out what was making money and what wasn’t. Not surprisingly, they found widespread duplication of functions and many competing projects. Now AT&T was paying the price, a virtual Babel of incompatible software and standards that operated the myriad switches, trunk lines, and phone systems sold and serviced by AT&T. When word of the EO investment had circulated throughout AT&T, related projects—some pure research and some more practical—surfaced like worms in a rainstorm. Determined to avoid chaos, the top AT&T managers agreed among themselves to establish one interface standard across all divisions.

  The stakes were enormous in this winner-take-all game—it was nothing less than a chance to create the interface for the next generation of telephones. Bill Campbell had assigned GO’s best people to working this opportunity. As far as we could tell, GO was bidding against Microsoft, which was offering a stripped-down version of Windows it was developing, and against an ultrasecret Apple spinoff called General Magic, which had a contract with AT&T’s data transmission division to build a new generation of data communications software.

  “The only other alternative is to try the financial community again,” Randy said. “But it’s likely to be tough during a recession.”

  “OK.” Bill was ready for his call to action. “Let’s get out there and scare up some new investors—somehow, somewhere.” He paused and looked around the room. “Anyone have any last suggestions?”

  “How about a bake sale?” Robert asked, biting into his muffin. “Let’s see, at ten thousand dollars a cake, we only need to sell fifty a week.”

  No one laughed.

  Armed with a fresh pile of awards and press clippings from COMDEX, Randy and I were ready to kick the financing into high gear. We put together a slide presentation for investors. “You know, this is pretty persuasive,” I said. “We’re technically the best. There’s a lot of hardware products being developed that will run Penpoint. We’re signing up licensees in Japan, IBM and AT&T are behind us, and the next version of Penpoint, the one for customers instead of just developers, is only months awa
y.” After working ourselves into feverish optimism, Randy and I hit the road like Hope and Crosby.

  The first stop was Bloomington, Illinois. We gave a quick update to Norm Vincent and his claims-estimating development team. He was as cordial as ever. I apologized for not visiting for so long. “Norm, I hope you don’t think we only come around when we need money.”

  “Jerry, don’t worry. I understand what you’re going through. I want you to know that we asked ADP to bid on building a Penpoint-based estimating system for us, but they came in and pitched us on working with Microsoft instead.” Randy and I held our breath: ADP was a large systems supplier for the auto claims market. “Do you believe that? I sent them packing, of course. As you know, State Farm just doesn’t like surprises.”

  He was as stalwart a supporter as ever. He took us to lunch with the investment folks. To our delight, they committed an additional $2 million, $4 million if we needed it. But this time they wanted someone new to set the price.

  Randy returned to the office to see if there was any last hope of wrestling reasonable terms from IBM, while I crisscrossed Manhattan like a sailboat tacking upwind, visiting old investors and fishing for new prospects. Having long since outgrown the venture firms, I worked my way up the food chain—past the corporate investors and pension funds, finally calling on the investment bankers who usually took companies public. I soon found the hook to interest them: if they put some money in now, they would get the inside track on our IPO. Hungry for “product” (as promising young private companies are known on Wall Street), they seemed quite interested in landing GO as a client. They all read the same market research studies and didn’t want to miss the boat in pen computing. Whereas most investors looked squeamish when they saw that we had already blown close to $40 million, the investment bankers seemed to take comfort from the larger than usual capitalization. As one wag put it, “You guys should be investment bankers—raising all that money while still retaining thirty percent for the common stockholders. Maybe it’s time for us to lend a hand before you put us out of business.” But no one was willing to set the price.

  By the time I got back to the office, we had about $12 million in sight. It looked as though we might pull through—if we could find a lead and close up the financing fast enough.

  Coincidentally, an old rugby teammate of Bill’s happened to call, and he asked Bill if we needed any money. “I can put you in touch with a buddy of mine who manages a pension fund,” he said. Within hours, I had a package off to a firm in Los Angeles whose partners had read numerous articles about pen computing and seemed grateful for an opportunity to invest. They soon committed $2 million, at $3.25 a share.

  “What does that make the pre-money valuation?” asked Scott Sperling of Aneas, one of our earliest investors, when I called to give him the news.

  “A hair over one hundred million,” I said. We both knew this was an aggressive valuation.

  There was a long pause, then he spoke hesitantly. “All right, we’ll take our pro rata share if absolutely necessary to make this happen, but only if everyone else will. But I want you to know that this is the last time. From here on out, you’re on your own.”

  “Scott, I understand. Thanks for your support on this one.” I shivered as I hung up the phone.

  Just after New Year’s of 1992, we were ready to close on this round of financing. Final papers had been sent to all the investors for review, which were due back with their funds on Monday, January 6. But the Friday before, around six P.M., I got a call from our attorney’s office.

  “We have a little problem,” said the paralegal assigned to the matter. “We can’t issue a clean Opinion.”

  One of the reasons lawyers are so highly paid is that they are on the hook for making sure that everything is legal and proper. Their blessing—essentially a written statement that puts them right out front in case of an irregularity that leads to a lawsuit—is a letter sent to the investors called an Opinion of Counsel. It attests that as far as they can determine, everything is aboveboard. Since for the most part they are simply stating that they did their job, I had never heard of a problem in issuing an Opinion, unless there was blatant evidence of fraud or misrepresentation.

  “Why the hell not?” I asked.

  “Sandra Shmunis.”

  “Sandra who?”

  “Shmunis. It appears that she may have been improperly issued options on ten thousand four hundred shares of GO common stock without a board resolution.”

  I was still stuck on the last point. “Who the hell is Sandra Shmunis?”

  “I don’t know,” the paralegal said. “But you better find out quick. If this is true, all the closing papers that we drafted, sent to investors, and filed with the Department of Corporations have incorrect capitalization tables. And our state authorization for the closing expires next Wednesday, the eighth.”

  “What happens then?”

  “We have to start again. New board resolutions, new closing papers, new filings . . .”

  I knew what that meant. Another $20,000 in legal fees at least, and several more weeks during which the investors could change their minds. “Can’t you go ahead and close, and we’ll sort this out later?”

  “Then the investors can sue you—and us.”

  “How about voiding the grant—if it ever happened—and reissuing it after the fact?”

  “Then she can sue you—and us.”

  I started to panic. I searched the halls for someone who might have a lead on Sandra Shmunis. No one had any idea who she was. In desperation, I got up on a chair, looked out over a sea of cubicles, and yelled out, “Has anyone ever heard of Sandra Shmunis?”

  A full minute went by, then someone wandered over and asked, “Isn’t she the new person in tech support?” In a flash, I was outside the tech support room. It was locked. I called the head of the department at home.

  “Sure, she started in October,” the supervisor said. “But I believe she just left on vacation.”

  I coaxed him into finding Sandra Shmunis’s home phone number, and I dialed it. I counted the rings. On the fourth, the answering machine clicked on. I was hosed.

  Then I heard someone struggling with the receiver, which promptly issued a loud squeal because the answering machine was turned up too loud. When it stopped, I heard children crying in the background. “Hello, is this Sandra?”

  “Yes?” said a timid voice.

  “Thank God. This is Jerry Kaplan. You know, chairman of GO? Sandra, have you been issued your stock options yet?”

  “I—I don’t know.”

  I had inadvertently scared her. A recent immigrant from Russia, she wasn’t prepared to get a call from the chairman questioning her about her stock.

  “Did you get any papers from human resources?”

  “Well, yes. Was I supposed to do something with them?”

  She didn’t realize she was supposed to sign the option grant form and turn it in. This would have triggered a procedure to add her name to a list of people whose options would be approved by resolution of the board at the next meeting.

  I took a deep breath. “Sandra, would you mind signing the form and dropping it by the office tomorrow morning?”

  “Of course. Is there a problem?”

  “No, I’m just trying to be sure you get your stock the way you’re supposed to.”

  She must have thought I was nuts.

  The next morning, the lawyers sent a blizzard of faxes. We tracked down the board members over the weekend and had them sign a “Resolution by Unanimous Consent,” thereby avoiding the need for a face-to-face meeting. Then new papers were transmitted to all the investors. By the morning of January 8, the last day we were legally allowed to sell the stock, we were ready to close. Or so I thought.

  “There’s one more problem,” said the paralegal. “The new Japanese investors wired their funds two days ago, and due to the delay, they want to invest the interest earned in the company.”

  “How much?”
<
br />   “It comes to a little over two hundred fifty dollars. We’d have to redraw all the papers, as this would increase the number of shares issued.”

  “Tell ’em no way. No, wait, I’ll tell them. I know—issue them a gushing apology and wire them the amount in dispute.”

  “You want to issue an international wire transfer for two hundred fifty dollars?”

  “You bet, if it will allow us to close. Just do it.”

  That afternoon, the attorneys deposited $12,935,503.70 in GO’s bank account—after subtracting the two days’ interest.

  While working on the financing, Kevin Doren accepted the job of securing licenses for certain key technologies that our hardware partners wanted us to use. One of these technologies, known as FFS2 (Flash File System 2), was a method for storing data in special memory chips that retain their contents when the computer’s power is turned off. Intel and Microsoft had recently announced this technique, which they had developed jointly, stating that they would make it public in order to establish an industrywide standard.

  Kevin started by contacting Intel, but learned that the technology was available only from Microsoft. After some effort, he succeeded in finding the person responsible for granting licenses to FFS2. Over the phone he was told that the written specifications would be made available to GO, but a license to the software that implemented it would cost $150,000.

  “I thought it was supposed to be freely available,” Kevin said.

  “It is freely available—it’s just not free,” said the Microsoft representative. Kevin politely objected and hung up.