Sometimes, usually on a Friday, Ellison will host a business lunch at home in Atherton. For instance, in July 2001, Ellison’s entertaining the top management team at The Gap to celebrate a deal to install the E-Business Suite to run the retailer’s financial and purchasing systems and various other modules to cut costs in its core business functions, including a nifty piece of Internet software called Oracle Product Development Exchange. On hand to greet the khaki-clad Gap executives as they roll up in their black Lincoln limos are Jeff Henley and Sandy Sanderson, the head of consulting. As usual, Ellison waits to make his appearance until all the guests have arrived (he can be heard playing classical guitar in an adjacent room) and are being seated at the immaculately laid table outside on the deck—he is conscious of his celebrity CEO status and knows how to nurture it. With its new clothing lines failing to find favor with its traditional customers, The Gap is having, if anything, an even more bruising time in the market than Oracle. So, while much of Ellison’s conversation is aimed at encouraging his guests to buy into Oracle’s philosophy of redesigning processes to get the best out of the software, he and The Gap’s CEO, Mickey Drexler, also want to compare notes about the economy. It’s an interesting contrast: Drexler, who along with Ellison serves on the Apple board, sees everything through a consumer lens, whereas Ellison is more concerned about the confidence of business customers to invest in an uncertain climate. Naturally, both are far too polite to suggest that any internal factors could be contributing to their current problems. When the guests eventually tear themselves away from their beautiful surroundings, Ellison expresses his frustration that despite the Oracle sales pitch, The Gap seems determined to go down a partial best-of-breed route, mixing custom software with the E-Business Suite. “They’re going to have a hard time with Retek [a retail software package]. It’s not an easy package to implement or integrate. They need to understand that,” he tells Sanderson.
When he doesn’t have lunch at home, Ellison usually heads to the office at around 1 P.M., driving himself the ten miles along the congested Highway 101 to Oracle’s headquarters at Redwood Shores. When she knows he’s on his way, Balkenhol will phone around, telling everyone who’s scheduled to attend Ellison’s meeting to gather. She then uses his travel time to update him on anything he needs to know, run through the afternoon’s schedule, get approvals on spending and hiring decisions, and find out whether he wants his lunch of chicken nuggets and ice cream.
Until Ellison’s arrival, Safra Catz will push things along in the eleventh-floor boardroom. But it’s slightly awkward for her, a fact reflected by her unwillingness to sit center stage. An important part of the way she sells herself within Oracle is that she doesn’t give the impression of running the show in his absence, in the way that Ray Lane and, to some extent, Gary Bloom might have done. When Ellison does stride in, the atmosphere becomes highly charged. It’s not that Ellison’s presence exactly makes anyone nervous (although a relatively junior executive waiting to make some careful PowerPoint presentation might disagree), nor do people become overtly respectful. There’s still plenty of tinkering with PalmPilots and frequent visits to the fridge in the hall to get drinks. If Ed Screven is there, he’ll be sprawled in his chair, his feet in decaying running shoes planted on the expensive table top. It’s more that there’s an air of expectation. Ellison is always in performance mode, cracking jokes, telling stories, and generally dominating every part of the discussion. It’s almost as if he feels the need to relegitimize his leadership each day by demonstrating the completeness of his grasp.
What’s noticeable in these meetings, which typically continue until six or seven in the evening, is how differently Ellison relates to the programmers compared with other workers at Oracle. They know that they are the elite, and a good few of the old hands are confident enough of their standing with Ellison to argue with him if they think he’s wrong and they have the facts on their side.2 From that point of view, the concern that an excessively controlling Ellison may be squeezing the creativity out of colleagues seems overdone. Although I have been present in a few meetings when there’s been an air of tension, usually when it turns out that somebody has fallen down on the job, Ellison no longer goes in for public executions. By his own admission, it’s quite a change from the way he used to operate, and, as always with Ellison, it’s the reputation he earned in a different era that stays with him now.
• • •
When Ellison was fourteen, he had an argument with his sister Doris, then in her early thirties: “She asked me the question ‘What’s more important to you? To be loved or admired?’ I said I’d like to be admired. She just said, ‘No,’ gave me her ‘You’re just a stupid adolescent’ look, and left the room. It took me a long time to understand that we all want to be loved. Even me. Imagine that. Anyway, I used to have a management style that earned neither love nor admiration. Rather than using conventional MBO [management-by-objectives] techniques, I invented my own style of management called MBR. MBR stands for ‘management by ridicule.’ MBR is perfectly suited to a young smart-ass programmer who just started his own company. A popular book at the time, The One Minute Manager, recommended that managers spend a minute a day praising their employees. MBR takes more time. You need to spend hours every day in meetings with your key senior people where you point out at length exactly why this person or that person’s ideas are utterly ridiculous. Not everyone can do it. You’ve got to be good at intellectual intimidation and rhetorical bullying. According to MBR theory, your brilliant arguments establish a clear intellectual dominance that gives people the confidence to accept you and follow you—their leader. Interesting theory, isn’t it? I was pretty brutal in attacking ideas and embarrassing people. I’d excuse my behavior by telling myself I was just having ‘an open and honest debate.’ The fact is, I just didn’t know any better. All my experience in business had been as a programmer. Programmers routinely play a game called ‘Who’s the smartest person in the room?’ It’s still a popular game at Oracle. I see it played all the time. When playing WTSPINR—pronounced ‘wet spinner’—you get points by showing how irrational or suboptimal other people’s ideas are. It happens in all disciplines to some extent, but programming is similar to mathematics in that there are clear right and wrong answers, so there are clear winners and losers. It’s classic primate behavior, figuring out where you are in the monkey troupe.”
I thought that maybe Ellison’s use of “MBR” was fairly typical of an insecure CEO trying to establish his leadership; he disagrees: “No.3 In fact, CEO is a title that’s looked down upon by most programmers. Remember, the name of the game is who’s the smartest in the room, not who’s got the most stars on their shoulder. There’s very little respect for rank among programmers. Programmers have tremendous intellectual integrity, and you have to be able to clearly explain why something is right or wrong if you expect to win an argument. What is unnecessary and unacceptable is to humiliate people in public while you’re explaining something. People didn’t sign up for that. You know: ‘Oh great, between ten and twelve I get to go to a meeting and be publicly humiliated. I can’t wait.’ ”
What was it that convinced Ellison that beating people up was not always the way of getting the best out of them? “I was in a big meeting between Oracle and another company. The other company’s CEO, an absolutely brilliant guy, started attacking ideas and people in a frighteningly brutal and destructive way. I was an MBR amateur compared to this guy. The overall dynamic in the room instantly changed. Everybody became afraid to talk because they didn’t want to be intellectually embarrassed. But it was worse than that. People started to line up against this guy, not because he was wrong—he wasn’t—but because he was inhumane. They grudgingly accepted that he was smart, but that didn’t keep them from wanting him to fail. From then on I was very careful about using MBR. Maybe it’s okay for programmers, but it’s certainly not okay for CEOs. It’s hard enough to run a company; you don’t want everyone rooting against you. Ni
nety-nine percent of the time MBR is degrading to the practitioner and damaging to the business.”
Although Ellison has learned from past mistakes, there are still aspects of the way he runs the company through these quite large meetings that can have some negative consequences. Ellison feels he has to perform to demonstrate his leadership and his prowess. Is it possible to be in both performance mode and listening mode? It makes picking up the nuances of a meeting more difficult and makes him all the more dependent on a few trusted people, Safra Catz above all, to tell him what he may be missing. And some very bright people, who might have important contributions to make, simply freeze in that kind of environment. And although Ellison no longer practices “MBR” and now tries hard to encourage debate, there probably still isn’t enough of it at Oracle.
Gary Bloom says, “The way Larry uses these meetings, they’re a fine environment for communicating strategy and decisions he’s already made, but it’s not a fine environment for debating those decisions once he’s made them. It’s an issue I raised with Larry, and I think he’s been working on trying to get people to be more open if they disagree with him. I once said to him, not long after I became a member of the PDMC (Product Development Management Committee), that I didn’t get what the meetings were for. Is it for you to articulate what you want or is it for people to tell you what you need to hear? He said, ‘It’s for people to discuss and tell me what I need to hear.’ I told him, ‘That’s not what happens.’ I said, ‘Larry, we have separate discussions from you, and we might sometimes all be in agreement that you’re wrong about something. . . . But as soon as you violently disagree or get emotional, everybody folds their cards. . . . The debate’s over—they’re unwilling to take you on.’ The very next day in the meeting, he took up a deliberately controversial position on an issue against me. I kept debating him, and eventually he turned to everybody and said, ‘You must all agree with my position because none of you are jumping in on Gary’s side.’ And they all went, ‘Oh, actually, we do agree with Gary.’ But even if he’s got better about this, a group meeting is still not a substitute for a one-to-one interaction between a CEO and key leader within the company.”4
This is an issue about which Bloom feels especially strongly. From the end of 1997, for at least two years, during the period when Ellison was slowly wresting power back from Ray Lane, Bloom was Ellison’s right hand. Whenever Bloom wanted to speak to Ellison in private, his boss made the time, especially if Bloom had uncovered yet more evidence of what they both at the time regarded as Lane’s incompetence.5 When the Oracle board insisted in 1999 that there should be a formal plan of succession in place, the only internal candidate whom Ellison was ready to endorse was Bloom (the top choice outside Oracle was Sun Microsystems’ Scott McNealy).
However, within twelve months Bloom was finding it difficult to get the kind of access to Ellison that he had come to expect. In part, it was because Ellison was beginning to think he might have promoted Bloom a little too fast and as a consequence, like Lane before him, his ego was starting to outstrip his ability (although Ellison denies it, I think he was riled by Bloom’s claim in Fortune magazine’s “Why Oracle’s Cool Again” cover story that he was perfectly capable of running Oracle). But it was also because of Ellison’s growing reliance on and trust in Safra Catz. Although Bloom and Catz got on well, respected each other, and were united in their contempt for Lane, it became increasingly clear to Bloom that he was losing out to Catz in terms of influence with Ellison. Bloom says, “This isn’t a slam on Safra. I like her. She’s very smart, and she really does understand what needs to be done. But what happened is that Larry, out of convenience, started using her as a path of communication. He felt it was easier to tell something to Safra and then have her go and communicate it to everyone else, including me. She became the filter and the funnel through which everything from Larry came. If I had to communicate something to my organization, which was about sixteen thousand people, I needed to get that little bit of background or color directly from Larry to provide the context for a decision. I never wanted to be in the position of a leader saying, ‘We’re doing this because Larry says so.’ It was so damaging to our culture. All I would need was thirty seconds of discussion direct with Larry. But when things started coming through Safra and I couldn’t get that, I couldn’t effectively do my job. I’m not criticizing Safra. I told Larry this and he said, ‘Okay, I see your point; I’ll try to change.’ But there was never any change. It got to a point where I was responsible for the vast majority of the company, yet I had no contact with the guy who actually ran the company.”
Carolyn Balkenhol confirms that Ellison became increasingly reluctant to see Bloom. As well as being frustrated and personally wounded, Bloom also calculated that Ellison’s new appetite for running the company meant that his own role, let alone his chances of one day becoming CEO of Oracle, would diminish. When, in November 2000, Veritas, then the fifth biggest software company in the world (thanks in no small measure to the efforts of Geoff Squire, the man Ellison most regrets firing), offered forty-year-old Bloom the top job, it wasn’t a difficult decision.
Although Ellison was annoyed about Bloom’s departure, partly because of the bad impression it gave to investors, coming so soon after Ray Lane’s “resignation,” that he couldn’t keep his senior executives happy, he has some sympathy with him. “I understand Gary’s frustration. Safra and I share a high-bandwidth communications link. We finish each other’s sentences. We come to the same conclusion in the same amount of time. We rarely disagree, but when we do, she’s not shy about expressing her opinions. If she thinks I’m wrong, she freezes me with one of her piercing stares and tells me I’m wrong. She’s not afraid of me or any other human. Anyway, I can count on her telling me the truth as she sees it. I respect her and rely on her as my chief confidante and counselor. In that respect she did replace Gary. That was very hard for Gary to take. Actually, come to think of it, he didn’t take it. He left.”
Although Ellison accepts that his key “directs” should have a right to one-on-one time with him, it’s still something that he feels ambivalent about. “It’s true. I don’t like doing a lot of it. If it’s something personal, then you have to do it one-on-one, but if you’re trying to decide a business policy issue like pricing, then I’d much rather have the whole team there so that everyone can weigh in. That way everyone will understand how we arrived at a particular decision rather than having to guess. Meeting with the senior management team as a group forces us to focus on the goals of the company as a whole rather than one department at a time. Looking at the gestalt of the business has resulted in less infighting between departments and a more open and collegial work environment.”
• • •
What’s very clear is how much these days Ellison leans on the slight, black-haired, black-clad figure of Safra Catz. He says, “She makes up for one of my biggest areas of weakness. She’s disciplined and thorough. I’m not. I’m pretty good at separating the good ideas from the bad ideas, and I’m pretty good at drilling into detail and solving problems. But once a problem is understood, once a plan is in place, I usually move on to the next thing rather than following up and making sure that the agreed-upon plan is actually implemented. It’s called execution, and Safra is brilliant at it. Safra is so exceptionally bright that she keeps all the whys and wherefores of all our policies and plans in her head. That enables her to make interpretations, modifications, and improvements during the execution phase of the plans without any intervention from me. Now when we decide on something, it actually gets done.”
Catz has always downplayed the extent of her power and influence, insisting that she has no agenda of her own (unlike, by implication, Ray Lane) and is there only to make sure that the things that Ellison wants to happen get done. She says, “I’m not interested in building power and I don’t have any individual power here. People will send me things for my approval and my response will always be okay, if it’s within the scop
e of a decision I already know Larry has approved. I say that as a reminder that I don’t have any power of my own.”
It’s an impressive (and genuine) demonstration of loyalty as well as being clever—it puts her out of the firing line when she’s doing something that ruffles feathers or bruises the egos of senior executives, especially on the sales side, who resent the degree of oversight and supervision she represents. But one of the criticisms that I sometimes hear about Catz both within Oracle and outside is that, unlike other managers, she is not really accountable—an impression that’s strengthened by her reluctance to become a public face in the way that Ray Lane and, to a lesser extent, Gary Bloom did. Ellison says, “She’s accountable to me—and to the board. Together with Jeff Henley, we’re collectively accountable for the performance of the company. If we’re not doing a good job, it will show up in the numbers. I understand her desire to stay in the background, but that’s going to be very hard to do over the long term. She’s gradually becoming more visible. Joining the board of directors was a big step in that direction.” Given that Ellison had sometimes referred to Catz as a possible successor, how did he want her role to evolve over the next few years? “I’m not sure it needs to evolve much. We work extremely well together. Sometimes we’re too aggressive about pushing new systems and procedures into the company, but we keep on improving our margins, so it must be working. We’re trying to define how a modern business operates. We’re continuously improving Oracle and our applications software suite at the same time.”
From Ellison’s point of view, Safra Catz is the answer to his prayers. She makes him much more effective within Oracle than he could be without her, he can trust her completely, and he can communicate what he wants through her almost intuitively. But I think there are potential dangers in the relationship for both of them. Ellison must be careful not to do what he did in the past with favorite subordinates, which is to load more and more responsibility on them until something snaps. In Catz’s case, there’s little chance that it will go to her head, as happened with Ray Lane; she is a remarkably grounded person. I once asked how she felt when Ellison said in interviews that she could run Oracle. She said, “I don’t want it, and why would I? My parents used to say to me that I could do anything I wanted to and that gave me confidence in life. But equally, I never felt I had to do something, I didn’t feel that I had to prove anything. I don’t have an individual agenda, so don’t make me one.”
Softwar Page 40
