“But the night before we were scheduled to make our big Java announcement, I asked Richard Sarwal, the guy in charge of our server performance group, to download the performance test on IBM’s Web site and test IBM’s Java. Richard looked extremely depressed: it took him twelve hours to download everything he needed. But he finally got the test running. When I came in the following day, I found out that IBM was not twice as fast as BEA; BEA was twice as fast as IBM. IBM just confused the subject with the predicate. That’s a pretty serious typo. After we completed all our testing, our Java ended up running twice as fast as BEA and four times as fast as IBM. I don’t know how IBM gets away with making claims that are so easily disprovable. I guess because they’re IBM.”14
The ads that Oracle had started running against IBM were as aggressive as anything it had done during the 1980s, when it was just a scrappy upstart. I wondered whether a self-confident, mature company should feel it necessary to define itself by trashing its competitors. Wasn’t there something undignified about it? Ellison grins. “I know it’s bad manners to publicly point to your competitors and say, ‘Nyah, nyah, nyah, we’re better than you.’ I know this style makes people think I’m a rude jerk. So be it. But I don’t know any other way to prove the superiority of our technology without directly comparing it to our competitors’ technology. We urgently need to make our points. There’s no polite way to point out that what IBM’s saying is not true. We present verifiable facts as clearly as we can, and then we ask the market to decide. Here’s us; here’s them; here’s the proof. You decide.”
Another part of the logic in going for IBM, as Ellison sees it, is that many of Oracle’s other rivals march under Big Blue’s banner, companies Ellison describes as IBM’s vassals. He says, “A lot of Siebel’s success is due to the fact that IBM recommends Siebel. It costs you $2 million for Siebel and another $20 million for IBM to put it in for you. IBM says, ‘We’ve looked at Oracle, we’ve looked at Siebel, and we really think you should buy Siebel.’ IBM is the supermarket for all of our competitors. They do two things. First they provide the skilled labor to install the stuff. Fair enough. The problem comes with IBM’s pretense of objectivity when they recommend application software: ‘We’re IBM. We’re an independent agent. We’ve done a careful and detailed analysis, and we honestly think SAP or Siebel or PeopleSoft or i2 or anyone is always a much better choice than Oracle.’ Of course they’re going to recommend SAP or Siebel because it costs five or ten times more to install and run that stuff. That’s how IBM makes their money: installing and running the software. IBM brings Siebel, i2, Ariba, Commerce One, and other economically fragile software companies brand credibility. Without IBM they’d all die. Even with IBM they’ll all die, just more slowly.
“We compete with IBM directly in the database and application server business and indirectly in the applications business. We’re beating IBM in the database business and the application server business. The only place they do well is on mainframes, and mainframes matter less and less every year. Time is on our side. In the applications business, it’s not IBM’s to win or lose. That battle is between best of breed and the integrated suite. If the integrated suite wins, then IBM loses. We’ll win the database and applications server war. I’m confident of that. And the integrated suite will kill best of breed. I’m confident of that too. But we need to beat SAP in the integrated applications suite war. That’s the hard one, but it’s doable. If we win all three battles—and that’s a big if—then we’ll replace Microsoft as the most important company on earth. But I don’t dream of an Oracle that’s like Bill’s Microsoft or Lou Gerstner’s IBM version 2. Our role model is Tom Watson’s magnificent creation, the technically brilliant and innovative IBM of old. That’s who we want to be like when we grow up.”
* * *
1. LE writes: If you’re a fighter, the only way up is through the top fighters in your division. So we picked fights with IBM and Microsoft because they’re the ones we had to beat to reach the top. By constantly measuring ourselves against the two top heavyweights, we constantly improve the competitiveness of our products and services.
2. LE writes: What bothers me most about Microsoft is the fact that they’ve been found guilty of repeatedly violating our antitrust laws but they’ve escaped all punishment—so far, anyway. Those guys are lucky as well as smart. I hate that.
3. LE writes: Recently, in a Wall Street Journal article, I predicted “the end of Silicon Valley as we know it.” I believe that a thousand Silicon Valley companies are in the process of going out of business and that the computer industry will consolidate around a few giant technology survivors. These large companies will dominate the industry and be the source of innovation in the years to come. Marc Andreessen, Netscape’s founder, responded in both horror and disbelief. He said that large companies are incapable of innovation; that innovation is done in small companies by nineteen-year-olds. Of course that’s what Marc believes. Marc is in his thirties. All he’s seen is the complete absence of innovation at Microsoft. He never witnessed the cornucopia of inventions that poured forth from IBM in the 1960s and ’70s.
4. LE writes: During Netscape’s heyday, John Doerr, Silicon Valley’s top venture capitalist and a big investor in Netscape, told me that Marc Andreessen liked to say, “Microsoft just didn’t get it.” That’s a saying in Silicon Valley. It means you’re technically behind the times, you can’t feel the change of seasons, you’re a dinosaur just waiting to die. But in Microsoft’s case, they were a dinosaur waiting to eat. Netscape was just getting fat before they were served up for lunch.
5. LE writes: They had reason to be afraid. They’re all dead now. There is no PC software industry anymore. There’s just Microsoft.
6. LE writes: Bill Gates, like most of the other very smart programmers I know, loves to prove his brilliance during arguments with other smart programmers. But if Bill has to choose, he’d rather win the war than win the argument. He doesn’t really care where the ideas come from as long as he’s the one who gets paid for them. The very last time I ever spoke with Bill was in 1999, when he called and invited me to visit Microsoft and debate the future directions of technology with their top technical people. When I declined, he asked me to reconsider, arguing that I would find it interesting “because there are a lot of very smart people at Microsoft.” I told him that was exactly why I didn’t want to go. He had expected me to show up and show off by explaining how clever Oracle’s ideas and plans were. Many of Microsoft’s competitors fell into that intellectual vanity trap. Not us. We want to get paid for our ideas.
7. LE writes: Actually, I do care a little. It sure would be nice if desktop Linux took off. I’d love to see Microsoft’s desktop monopoly broken and their revenue stream reduced.
8. LE writes: We created our Collaboration Suite [about which more later] to make it easy for users to replace Microsoft’s Exchange e-mail servers and Microsoft’s file servers with an Oracle database. Data, all data, belong in a database.
9. LE writes: Andy’s the only guy whom both Steve Jobs and I agree we’d be willing to work for. Andy’s absolutely brilliant and brutally honest. I’ve always loved him, but I fear my love has gone unrequited. I’m afraid Andy thinks I’m a bit of a flake. That’s what he told me, anyway.
10. LE writes: Joseph Heller’s novel Catch-22 is set in Europe during World War II. One of Heller’s characters is Milo Minderbinder, a budding entrepreneur serving in the U.S. Army Air Corps. To supplement his army pay, Milo forms M & M Enterprises with the grandiose marketing tag line “Whatever is good for M & M Enterprises is good for the world.” Milo gets into a bit of trouble when M & M Enterprises signs a contract with the Germans to bomb American forces. But Milo strongly defends his actions by explaining that the war was being fought to “preserve the free enterprise system.” Any restrictions on the behavior of M & M Enterprises would conflict with the goals of the war and be bad for the world.
11. LE writes: I suppose they could build televisions t
hat receive only MSNBC and give them away for free. I’m sure they’re thinking about it.
12. LE writes: More likely Bill simply calculated the risk-reward ratio in breaking the law to break Netscape and acted accordingly. Getting rid of Netscape benefited Microsoft approximately $25 billion to $50 billion in market cap. The punishment for breaking Netscape and the law seems to have been a few harshly worded editorials in The New York Times and The Economist. So far Bill’s calculations seem to be correct; for Microsoft the benefits of breaking the law have been great and the financial cost near zero.
13. LE writes: Perhaps people shouldn’t care, but they do. So we had to adapt our operating system strategy from “Don’t ask, don’t tell’ to the very latest, hippest fashion—namely, Linux. The Linux movement was gaining tremendous momentum, so we decided that we’d better jump on the train, not in front of it. But we had to add a number of critical features to Linux to enable it to run our fault-tolerant database clusters. We had to assign a number of our engineers to work with Red Hat [a Linux distributor] and the open-source community to build the features we needed into the Linux operating system. We even took our distributed file system and donated it to the open-source community so that it could become a standard part of Linux. After a lot of time and effort, Linux has now developed into the prime-time, mission-critical operating system that we need to compete head-to-head with Microsoft Windows. Today, Linux on Intel beats Windows on Intel in every important area: performance, reliability, security, and cost.
14. LE writes: IBM continues to make numerous incredible claims about its software. Perhaps the strangest was the claim that DB2 was the only database that ran on UNIX, Linux, and Windows. After we called them up and pointed out that we ran on UNIX, Linux, and Windows, they quickly took the claim off their Web site. I don’t think IBM is lying when it makes these claims; I just don’t think it knows the facts. This is not the IBM I so admired when I first started working in the computer industry. This is not your father’s IBM.
16
CHAINED TO THE JOB
Early on I had wondered just how committed Ellison really was to making Oracle as great as IBM in its heyday. I thought that given his age, the range of his interests, his unimaginable wealth, and his own self-confessed dislike of much of the routine drudgery that’s involved in running any business, it was almost inconceivable that he would have the necessary stamina and the hunger. Two things made me realize that I had underestimated him. The first was Ellison’s evident willingness to put in fourteen-hour days while on the road. The second was sitting in on the meetings in Oracle’s eleventh-floor boardroom, from which he drives and directs the company. This was not a “big-picture” chief executive who passively absorbed sanitized briefings at ten thousand feet from nominal subordinates who were actually running the business.
Whether he was suggesting a solution to some obscure engineering issue concerning the database, deconstructing some back-office process at Oracle to understand what was needed from applications to automate a business flow seamlessly, tweaking a sales compensation plan to eliminate perverse incentives, designing a “portlet” for Oracle’s executive dashboard, penning the latest ad, or sharpening an online sales presentation, Ellison was everywhere. Nor did it strike me that he was interfering with the work of others for the sake of it. Although most of the people who work at Oracle are formidably intelligent, Ellison’s constant probing and refining seemed both to put them on their mettle and genuinely to add value. Ellison also constantly monitored every shift in his competitors’ business tactics or product strategies, ensuring that Oracle could rapidly exploit any emerging weakness or block off any new threat.
Ellison had told me that he was unable to strike a balance between an extreme form of laissez-faire delegation and an equally extreme obsession with influencing every aspect of the business. He could either be semidetached from the day-to-day management of Oracle, as he had been during the mid-1990s, or furiously involved, as he was now. Although Ellison’s energy and commitment were hugely impressive and there was hardly anyone at Oracle who preferred the hands-off Ellison to the hands-on version, I could not help remembering Jeff Henley’s anxiety that perhaps Ellison was now being too controlling and that the creativity of some very smart people might fail to flower as a consequence.
Overdependence on Ellison carried other risks: according to some people who knew him, there had been a cycle in the past in which Ellison’s appetite for running Oracle would wax and wane depending on the challenges facing the business. Henley actually believed that in some ways the pounding that Oracle was taking in 2001 was a stroke of luck because it meant that Ellison had no option but to tie himself to the job. If success with 11i came too quickly, there was a danger that Ellison might relax. Henley and other Oracle executives were also worried about the commitment Ellison was making to winning the America’s Cup. When I inadvertently mentioned to Henley, in July 2001, that a successful America’s Cup campaign might mean Oracle’s losing its CEO for the best part of five months, he had been horrified. At the very least, he said, it would mean that some important deals, which only Ellison could close, would be lost, maybe forever. He was so upset that he called Ellison, who was cruising the Mediterranean on board Katana, and threatened to resign if it was true. Ed Screven, Oracle’s chief software architect and the man whose job it is to make sure that the various development groups are working synergistically, had an additional concern: “Larry casts a pretty long shadow, so he can be gone for an extended period of time and his shadow is still there. But what will happen is that some disputes that only Larry can settle will remain unresolved.” I asked Ellison how seriously he took their misgivings. No matter how much he wanted to win the cup, he said, if Oracle needed him, Oracle would come first.
• • •
A typical day for Ellison starts when he gets up at about seven o’clock and checks his overnight e-mails. Unlike many CEOs, Ellison has only one e-mail address. He gets more than a hundred e-mails a day, most of which he answers himself or forwards to the appropriate person. He prefers short e-mails that deal with one specific issue rather than essays because it makes collaborative working easier. After dealing with e-mails he heads for the spacious deck that overlooks the exquisite Japanese garden. There he usually breakfasts on scrambled eggs and bacon (Ellison’s Judaism is of the lapsed variety) or fish, rice, and miso soup while reading the newspapers (the Financial Times and The Wall Street Journal; The Economist on weekends). Even if it’s raining and cool, he likes to sit out of doors as much as possible.
At about nine, he heads back inside to continue work on e-mails, writing memos, checking through ads, and working on larger “think pieces” both to clear his mind and to outline strategic issues and forthcoming challenges for colleagues. Some of the papers that Ellison writes will go straight onto Oracle’s intranet, especially those aimed at sharpening the message of the sales force. Also part of the routine is a visit to Oracle Sales Online, which might prompt him to leave a message questioning the status of a particular sales opportunity. During the morning, he also takes telephone calls, patched through to him by either Carolyn Balkenhol at Redwood Shores or by Joyce Higashi, the highly efficient Japanese American who works from his home and meshes together the personal and professional sides of Ellison’s life. Some of these are regular conference calls with major customers in the middle of implementations. Every fortnight for a year, for example, Ellison talked to the senior management at GE Power during its rollout of the E-Business Suite.1 Another part of the routine is to check out the Web sites of rivals: software companies tend to tell the world about what they are intending to do long before they actually have the products ready to ship. Thanks to the Internet, there’s no excuse for ever being taken by surprise.
At about eleven o’clock, Ellison usually jumps into one of his cars—his current favorite for everyday use is a Mercedes SL55—and drives a couple of miles to his gym for a workout until lunchtime. Although he has a well-equip
ped gym in his house, he prefers to exercise with other people rather than at home with a personal trainer. He says, “I can’t work unless I work out. I can’t push myself mentally unless I push myself physically.”
Softwar Page 39