Softwar

by Matthew Symonds

  Nor is it obvious that making Oracle as big as possible as quickly as possible and relentlessly pursuing market share was the wrong strategy. Ellison says, “Among our early competitors, Ingres and Informix are gone and Sybase is on life support. Silicon Valley is a killing field. Very few technology companies survive. We did.” It was this rather than the pursuit of a particular revenue number that fueled Ellison’s competitive drive. Ellison understood that once a customer committed to a database vendor, he was pretty well locked in for the next ten years or so. Releasing Oracle 6 when it wasn’t ready was extremely risky, but delaying it still further would only have allowed Sybase to become more entrenched in the market. It might have been a lousy way to treat customers and it might have done damage to Oracle’s reputation, but to Ellison, the alternative of losing those customers for good would have been worse.11

  Perhaps Ellison’s gravest fault was being instrumental in allowing the firm’s cash position to become so nearly terminal. He might not have been sufficiently aware of his duties to note the fragility of his own company’s balance sheet, but remarkably he felt qualified to ignore the warnings of others. That single sin of omission may well have had the most serious consequences of all. If, as Jeff Walker and Don Lucas had advised, Oracle had raised money when it hadn’t needed it, the consequences of the lost quarter and the revenue restatements would have been little more than an embarrassing stumble.

  Ellison’s claim that he was “incompetent” is only half true. He combined arrogance and recklessness to a degree that was very nearly lethal for Oracle. But against that he built something so inherently resilient and valuable that both he and Oracle would have the chance to reinvent themselves. As Jeff Walker puts it, “If you take a brilliant, compulsive personality and you subject them to a learning experience, they learn brilliantly and compulsively.”

  * * *

  1. LE writes: I admit to pushing our sales force to grow faster than the competition. Without increased sales we would have been unable to increase the investment in our engineering team. We had to increase sales to maintain our technical leadership.

  2. LE writes: The only thing that mattered at Oracle was results—not brashness or aggression. The European sales force was neither brash nor aggressive, yet their results were just as good as the U.S. sales team’s. The personality of our different sales forces reflected the personality of their sales leaders.

  3. LE writes: I am most comfortable working with engineers, so I wanted a company where most of the managers were engineers. But making Jeff Walker the CFO while he was still in charge of applications development was a big mistake. Accounting is a control function and should always be separate from sales and development. Unfortunately, I didn’t understand that at the time.

  4. LE writes: Legal, like accounting, is a control function and has to operate independently of the sales department.

  5. LE writes: Not true. The deal was proposed and passed to me for approval. I did not approve the deal, and it never happened.

  6. LE writes: Geoff Squire was nothing less than heroic during this crisis. Without his efforts, the company probably would not have survived.

  7. LE writes: Actually, I wanted to raise cash and was working with Safra Catz, who was then at DJL [an investment bank], to do just that. We just could not come to an agreement on structure. I wanted a bond deal. Others wanted a convert. Without an agreement on structure we did nothing.

  8. LE writes: Tom’s crisis of conscience was coincident with the drop in our stock price, which made his huge option grants worthless. Tom, like many other U.S. salespeople, left Oracle because he thought the big-money days were over. And I certainly never offered Tom the Oracle USA president’s job. In fact, I had already given that responsibility to Geoff Squire.

  9. LE writes: These days I believe that every process within an organization—marketing, sales, service, everything—should be carefully engineered. This can be done only if a company implements systems that automate and monitor its business processes. I’m still an engineer at heart, but now I apply engineering discipline to our entire business, not just product development.

  10. LE writes: I know that Microsoft seems to be getting away with breaking the antitrust rules, but it will probably catch up with them in the end. Or they may stay lucky and get away with more slaps on the wrist. Who knows? Still, by breaking the rules Microsoft took a huge and unnecessary risk. If not for a few careless comments by Judge Jackson, it would have paid dearly for that.

  11. LE writes: I did not release Version 6 of our database, or any other software product for that matter, knowing there were serious bugs in the product. When you rewrite a large portion of a complex product, as we did with Oracle Version 6, it is virtually impossible to test thoroughly enough to find all the bugs. Unfortunately, you do not discover some of the bugs until the product is in use by customers.

  6

  GROWING UP

  By the beginning of 1991, Ellison had recovered enough from his very public humiliations to be fully engaged in digging Oracle out of its hole. Jenny Overstreet says, “There were a lot of people saying, ‘Told you you’d get your comeuppance, you sonofabitch.’ He was badly frightened, but the Larry I knew responded to great fears by regrouping quickly. He’d decided, ‘I’d better go back to the only thing I know, which is to work really hard and focus all my best energies and all my best talents on fixing this because I’m not really ready to quit.’ ”

  Marc Benniof, who had taken over the telephone sales operation on the recommendation of Tom Siebel, sees it slightly differently: “I don’t see Larry as conventionally resilient. I see him as a driving force. Larry gets what he wants, and when he focuses on something he gets it.” Benniof, who became one of the few people working at Oracle in the 1990s to become a close friend of Ellison, is even reluctant to buy the idea that Ellison was all that surprised by what had happened: “Larry had his foot to the floor at a hundred fifty miles per hour, trying to leave all his competitors behind. He knew that eventually he would shoot the car over the cliff. But it was like in a James Bond movie—the car makes it across to the cliff on the other side even though it’s broken into ten thousand little pieces. It’s like he had to go back and say, ‘Sorry, Q, look what I did to the car!’ But he’d accomplished his goal.”

  Ellison had to do three things if he was to put Oracle back together. The first was to deal with the cash crisis. The second was to bring in seasoned managers who would bring some solidity to what was now a billion-dollar company and give confidence to investors that the mistakes of the past would not be repeated. Finally, he had to overcome the reliability problems of Version 6 and make sure that the next release would more than hold its own in the features race with Sybase.

  Of all these, the most urgent was Oracle’s dangerous cash crunch. Ellison knew that a conventional financing solution would not be available on anything less than the most punitive terms, terms that would undermine his position as Oracle’s dominant shareholder by destroying the value of the equity held by the company’s most important employees: the key developers. Ron Wohl was a young product manager who for several years had been a notably loyal staff officer. He remembers, “Not only did we not have cash, we had not made good banking or other arrangement for financing. We were in danger of violating our covenants [which could have resulted in the bank syndicate calling in the $170 million already borrowed], so we needed to raise some money. At that point, the Japanese economy was extremely strong and Japanese companies were very interested in participating in high tech. Larry had a very strong personal interest in Japan. So he began some initial discussions with Japanese companies which rapidly centered on Nippon Steel, a very, very capable company. Larry asked me to take a lead in the negotiations because our CFO at the time, Jeff Walker, had lost credibility and Larry was in the process of looking for a new CFO.”

  Ellison made it Wohl’s mission to get $200 million while giving as little Oracle equity away as possible. Although initially th
e Japanese were bewildered to be dealing with a rather junior thirty-year-old, Wohl turned out to be an inspired choice. Wohl has an almost oriental inscrutability that makes him an extremely cagey negotiator and, at times, an infuriating colleague. Even in normal conversation his wariness and reluctance to commit himself can be unsettling. Ellison says, “Ron is extremely smart and very cautious. He hates errors of commission, which means he can be slow to make decisions. That’s very Japanese.” Not surprisingly, the negotiations proceeded at a glacial pace. Wohl says, “Given the complexity, it was a long-drawn-out negotiation involving a lot of money, a lot of complicated financial arrangements, some partial ownership of Oracle, and some board arrangements in Japan. During that time, Oracle’s financial position was getting better by the day and we needed much less money.” So much better, in fact, that it dawned on Wohl that his boss didn’t really want him to do the deal at all.1

  With an agreement imminent with the cash-rich Japanese, Ellison was able to hold off Oracle’s bankers and buy time to renegotiate a different and much more advantageous deal with Nippon Steel than originally conceived. By the end of 1991, with expenses under reasonable control and the beginnings of recovery in the economy, Oracle had returned to robust, if unspectacular, profitability and the stock was climbing. Instead of surrendering equity in Oracle to pay off the banks, Ellison had secured a deal that gave Oracle an $80 million loan on highly favorable terms, in exchange for which Nippon Steel had warrants to buy up to 25 percent of Oracle Japan at an exercise price that gave a value to the subsidiary of $400 million. It wasn’t quite a case of “and with one bound, Jack was free,” but it was getting close.

  Another reason for the restored order of Oracle’s finances was the arrival in early 1991 of a new CFO. Ellison accepted that Oracle needed a high-class, conventional, professional CFO rather than someone else who, in Geoff Squire’s words, was smart but couldn’t do the job. About the only thing that Jeff Henley had in common with his predecessor was the same first name. According to Don Lucas, Henley was “a straight arrow,” a perfect choice. “We were extremely fortunate in finding Jeff. He was very highly considered by his peers. He’d been there, done that. He’d worked for Memorex and one of the semiconductor companies. You don’t have to understand software per se, you just have to know how to do things right.”

  At the time Lucas approached him, Henley was CFO at the real estate firm Pacific Holdings Company, controlled by the controversial entrepreneur David Murdock. Henley had worked for Murdock for four and a half years and “couldn’t stand the guy.” He feared that on the basis of what he’d read, Ellison sounded very similar to Murdock. “Murdock was a self-made guy, much like Larry, very opinionated, very volatile temper, and wouldn’t listen to anybody. And here’s this Ellison, that they say is an egomaniac and all this other stuff, so why would I go and work there?”

  Henley was won over in December 1990 at a dinner with Ellison at the Sofitel hotel across the road from Oracle’s new “Emerald Towers” headquarters. Henley says, “He had clearly been in deep shock, but now he had gotten over that and was trying to get things going. He was very concerned and, as I would later discover, worried for his own job. I found him very resolute and committed to turning the company around and that he had taken a very active role after a number of people in the leadership had been fired. He was even signing all the checks in the United States as a way of really trying to get his arms around what was going on, but he clearly acknowledged that he needed more experienced management.”

  Far from Ellison’s personality being an obstacle to him, Henley wanted a guarantee that Ellison would not be forced out. “I didn’t want to join the company and see Larry leave. I found him very different from Murdock—Murdock thought he knew everything and was in a thousand different businesses, but Larry was very focused on the business of technology. Murdock had street savvy, but nothing like the intellect of Larry. Murdock had no sense of humor, while Larry had a great sense of humor, so I saw a guy who I could also have a lot of fun working with. I’d been ten years in the tech industry, and I came to believe that most tech companies lose their way because they don’t have leaders who really understand the technology, but Larry did.”

  Ellison also liked Henley a lot more than he expected to, finding in him much more than the kind of super bean counter he feared being saddled with. What impressed him was Henley’s calm rationality. Although he was only a year or so older than Ellison, in his late forties, with gray hair, craggy features, and the quiet confidence that only the truly well grounded have, he had a somewhat senatorial air that was far removed from the Oracle of the 1980s. Henley also appreciated Ellison’s willingness to learn from his mistakes. “He acknowledged that he had created too much competition inside the company. He hadn’t pushed teamwork enough, so he had a group of people on his executive team who just hated each other. He also realized that hiring really smart people, very smart young engineers or whatever, had to be balanced by bringing in people who had some business experience. The blowup made him modify some views he’d held for a long time, it was like ‘I had a couple of blind spots, and now I need to go work on them.’ ”2

  Henley’s arrival made an immediate difference. Noosheen Hashemi had been banging her head against a wall trying to get Jeff Walker to agree to bring in fifty accountants from Peat Marwick to go through the boxfuls of weird contracts and deals that had to be reconciled to get the numbers into shape. “Walker said I could spend just twenty thousand dollars after letting go two thirds of the people who worked for me. I said, ‘We can’t. We have to restate. This is serious.’ He had his head in the sand. The day Henley walked in, he just said, ‘Go clean it up, go do it.’ He signed up just like that.”

  Other things also began to change with Henley’s arrival. He began to modify Ellison’s policy of hiring only exceptionally clever young people. Hashemi says, “We had honor students delivering the mail. That was superb for the company when it was growing so fast because people could take on jobs that evolved in value very quickly. When Henley came in, he said, ‘I don’t see growth like that. I’m going to hire a payroll clerk for fifteen thousand dollars that has twenty years of experience and no college degree, not some twenty-four-year-old prima donna who doesn’t expect to be doing the job for more than six months.’ From that moment on, I felt the company lost the incredible culture it had.”

  Henley also started chipping away at expenses such as lavish parties and massive bonuses. Hashemi, a zealot who thought nothing of working eighteen or twenty hours a day for her beloved Oracle, says that in the past it would have been unusual to find an empty conference room at nine in the evening. “Oracle became a job, when it had been a mission.” Most shockingly of all: “People started having children.” In other words, Oracle was doing what it had to: evolving into a more stable, grown-up company. Something may have been lost in the process of dialing down the manic intensity that had helped get Oracle into trouble, but not much of it was worth keeping.

  With Henley in place bringing reassurance to Oracle’s investors and order to the company’s finances, Ellison’s next task was to find someone to run U.S. sales and consulting. After the departure of Gary Kennedy, Ellison had appointed his polar opposite to run U.S. sales, a genial, almost cuddly man named Mike Fields. Hashemi says, “He was sort of like a teddy bear, he was what we needed for a short period of time. The worst thing that could have happened was to have put in someone arrogant like Tom Siebel. What was needed for a year was for Mike Fields to go around shaking hands with customers, saying, ‘Sorry, we’re a different, kinder, gentler company now.’ ”

  Although Ellison eventually fired Fields, he is still quite fond in his recollection: “Mike was an exceptionally nice man—he was the Bob Miner of sales. He was a father figure. He was effective at sales because people trusted him. But Mike had never managed a sales organization as large as Oracle’s. I wanted someone experienced managing a large organization. And I wanted someone who was servicecentric�
��not salescentric.”

  Ellison knew what kind of manager he was looking for to change the cowboy culture of the U.S. sales organization. “I had never heard a customer say, ‘Gee, what I need are a few more salespeople calling on me.’ Customers always want better products and better service. With Oracle Version 7 we were about to introduce a greatly improved product. Now we had to improve our service. The era of the cowboy was over. We needed someone who could build and run a railroad.”

  For practically the first time, Ellison used the services of a headhunter, Heidrick & Struggles, to identify candidates for Mike Fields’s job as head of Oracle USA. The two men who were shortlisted were a world away from the old Oracle. One was George Conrades, who was running IBM’s sales in America and later went on to be the CEO of Akamai, the successful pioneer of Web-caching technology. The other was Ray Lane, a senior partner of the big consultancy firm Booz Allen Hamilton. Lane had been at Booz Allen for about twelve years and had built its IT practice—the Information Systems Group—into one of the four major lines of business within the firm. With the handsome features and firm voice of a film director’s idea of a cliché presidential candidate, and a good opinion of himself to match, Ray Lane certainly looked the part.

  However, Lane had to be talked into seeing Ellison by Heidrick & Struggles’ John Thomson. Oddly, Lane had never heard of Ellison and knew next to nothing about Oracle. Thomson suggested that even if nothing came of meeting Ellison, Lane might be able to sell himself some consulting work down the road. When Lane agreed to fly to California from his base in Dallas, he decided to do some research on Oracle. “It was all really bad. I remember one article in Upside magazine which had Larry portrayed as Don Quixote [tilting at windmills], and there were others about the demise of Oracle and its near bankruptcy. Frankly, I didn’t know why I was going to meet him. But I did, and I spent three hours with him, talking more at each other than having a conversation. About eighty percent of the discussion was my views of how companies don’t understand how to use computer technology and what they needed to be doing. He kept on talking about the technology. So at the end we were really firing each other up in different dimensions—talking about the same things, but looking at both sides of the elephant. He wanted somebody who understood what the technology did for customers and who could sell solutions. Finally, it got to the point after two hours when he says, ‘You’re the guy. You’re exactly what we’re looking for. Will you do this?’ ”3