“So this stuff about Internet was not even a debate. I have a feeling that Larry is trying to position himself as this Renaissance man who saw the light and saved everything. He doesn’t have to. He did. That doesn’t take away from the fact that our application business was a disaster. But by then Larry had come up with a strategy and everyone knew that it would take time before we had applications that did it. . . . A great strategy, Larry. We love it, we’ll do it, but when we stick this stuff in our quotas, it has to be something real. We have to ship a real product that works. Otherwise we don’t get paid.”6
As usual, the truth seems to lie somewhere in the middle. Ellison exaggerates the degree to which Lane was orchestrating a campaign against the strategy of getting to the Internet ahead of the competition. As a salesman, Lane was as keen as anyone to have a new story to tell. As Ellison himself frequently says, technology is a fashion business, and by 1998, the Internet was nothing if not fashionable. Lane’s overriding concern was that the transition to Internet computing would not happen as quickly as Ellison hoped, leaving his sales force and consultants to carry on the struggle with a suite of client/server applications that had been effectively orphaned by the development organization. He was worried that without any prospect of Oracle’s client/server software getting better, customers would desert in droves, unwilling to be guinea pigs for Oracle’s first Web-based applications.
That said, Ellison was almost certainly right in thinking that Oracle couldn’t afford the luxury or manage the complexity of developing radically different versions of all its application software. It’s a measure of Lane’s understanding of the technology that he either couldn’t or wouldn’t see that continuing with client/server would make it far less likely that the Internet strategy would succeed. Ellison’s determination to cut off client/server entirely might have looked reckless, but in fact, it was the only option open to Oracle that gave it any chance of shaping a future in applications that would be better than its troubled past. By not only questioning that decision but making it clear to the field that he disagreed with it, Lane was straying perilously close to challenging Ellison’s judgment on a major technology issue, something that even he would agree he was ill equipped to do.
He was also ensuring that Oracle would become an even more divided and factionalized company. His long crusade against Ron Wohl, whether justified or not, had become morally corrosive and commercially destructive. Furthermore, while Lane constantly demanded personal loyalty of a high order from the people who worked for him, increasingly he did not appear to think it necessary for them to extend that loyalty to Ellison or the people who built Oracle’s products. There is little doubt that Lane believed that he was acting with the best of intentions, but his years of frustration over applications, the flattering media coverage that he had encouraged, and the boost to his already sizable ego from the Novell episode had eroded his judgment, certainly regarding his relationship with Ellison.
As had happened with Gary Kennedy and Jeff Walker, somebody who had once been bathed in the beneficent glow of Ellison’s approval and admiration was about to experience the chill of his displeasure. Ellison says that the thing that disturbed him most was not so much the feeling that Lane was invading his territory but that he had lost touch with any realistic idea of his own capacity. “When Ray first came to Oracle, he spent most of his time with customers. He built the kind of personal relationships that he could rely on when it came time to close a big deal. That made him a fantastic salesman. But now, all of a sudden, he thinks he can come up with a technology strategy that would enable Novell to compete with Microsoft in PC server operating systems. Oh my God, fight with Microsoft in the PC operating system market! Not my first choice if I’m looking for a fight. So Ray’s a technology strategist now. That’s why he thought he could turn around Novell. That’s why he fought with me over client/server versus the Internet. Ray decided that he’s good at things I didn’t think he was good at. Scary. Thinking you know things you don’t know leads to a lot of mistakes. But I was also vulnerable to him leaving because I wasn’t yet competent to take over from him all the things he did. First I had to move his departments to other managers. And that would take time.”
I put it to Ellison that there was something ruthlessly calculating about leaving Lane in post while submitting him to a kind of corporate death by a thousand cuts. Ellison replied, “I had to start repairing things piece by piece and transferring responsibility from Ray’s team to the ‘shadow management team’ without overwhelming it. I’m not confrontational, but I am pretty calculating. This was carefully calculated.”
Ironically, by pushing Ellison into deciding that he had no choice other than to take over applications, Lane had triggered the process by which his own authority would unravel. Having removed all application development from Lane’s part of the company and put Ron Wohl in charge of sorting out the mess that had been created by best-of-breed and the industry vertical suites, Ellison stripped two other functions from Lane: first marketing and then education. Marketing was run by a bright, attractive, but somewhat controversial woman named Karen White who was something of a Lane protégée. Ellison says, “I had hired Karen into the company a while back. I always felt that she was exceptionally bright and one of the most determined people I had ever met. But I also felt that she was miscast in marketing, and I suspect that she felt the same.” White was pushed sideways, and Mark Jarvis, a personable Englishman with a background in engineering, was put in charge and reported not to Lane but to Gary Bloom.7
The next to go, in May 1999, was education. For most big software companies, educational programs fulfill two important functions. The first is to spread expertise to IT professionals about the use of a company’s products. The second is to make money: companies and individuals are prepared to pay substantial amounts to gain useful and highly marketable skills. However, as Ellison discovered during budget review in March 1999—the first he had been fully involved in for years—Oracle’s education business was becoming less and less profitable.
The executive responsible for running education was Randy Baker. Lane hired Baker from Tandem Computers to run Oracle’s product support organization. Later Baker was given the additional responsibility of running education. Ellison says, “Randy proposed an education budget for the following year where margins were planned to drop from seventeen percent to thirteen percent. I thought margins should go up—not down. In fact, I thought margin should be much higher—at least thirty and as much as fifty percent. Nobody could give me a satisfactory explanation why our education margins were so low, so I moved the education business away from Ray and Randy and to Gary Bloom and John Hall.”8 Within a year, Baker had left and support was also reporting to Bloom.
Lane professes that he wasn’t worried about losing either marketing or education: “I really didn’t care. I wasn’t interested in keeping marketing; marketing was the most hated function at Oracle. With education, I guess it probably was the first signal that Larry was either doing something for Gary or taking something from me. But I didn’t mind all that much; it was a small business that wasn’t important to me.”
Even now, Lane seems to have only a relatively hazy idea about what was actually going on. Ellison’s immersion in the applications business and his determination to reduce his dependence on Lane had a number of consequences. Most obviously, Ellison was carefully and deliberately building up Gary Bloom at Lane’s expense. Ellison wanted people both inside and outside Oracle to see Bloom as a manager who had been given “tremendous stature.” Bloom says, “I became the guy who got all the broken stuff and went and fixed it. But I was also building up a following based simply on respect for delivery.” What Lane didn’t realize was that Bloom, whom he had generously praised and encouraged during his rise up the hierarchy, was an enemy, contemptuous of his competence and eager to bring back to Ellison any information that would cast Lane in a bad light.
Wherever Ellison looked, he seemed to find
glaring examples of inefficiency, wastefulness, and intellectual sloppiness. A prime example was something called the Energy Center of Excellence. This was intended as a place to showcase Oracle’s best-of-breed suite for the energy industry. In the course of the 1999 budget sessions, Ellison says that he uncovered plans to spend millions of dollars to run the center. “I asked how much energy upstream and downstream software we were planning to sell next year. The answer was about $10 million. So I said, ‘Does it bother anybody here that we are forecasting sales of $10 million and we’re spending $5 million to run an Energy Center of Excellence? Does that strike anyone as strange?’ What struck me as strange was the fact that Ray had approved all of this.9 I asked Ray to explain why we were spending so much money on the Energy Center of Excellence. He says, ‘Larry, that’s not what you’re paying me for [managing the Energy Center of Excellence]. You pay me to bring in big deals.’ I guess that’s true. But while he was out selling, somebody needed to look at our sales costs and say no to things like a multimillion-dollar Energy Center of Excellence. But nobody was.”
In part because of the work he had been doing in applications, Ellison was beginning to realize how chaotic many of Oracle’s own processes were. At the same time, he was discovering that in many of the areas that Lane was responsible for, getting hold of the basic information required for the budgeting process was disturbingly difficult. “The most basic budget-planning module we have in applications asks a few basic questions: How much did you sell last year? How much are you going to sell this year? How much did you spend last year? and How much are you going to spend this year? Simple, right? Wrong. It took many hours and multiple meetings to pry answers to these four basic questions from some of Ray’s senior managers. One of Ray’s people, George Kadifa [now CEO of the leading application service provider (ASP), Corio], came in with two hundred slides. I’d said, ‘George, I’ve got four questions. Once you’ve answered these four questions, I’ll look at however many slides you want me to.’ He was quite offended. He thought it was very important for me to look at his slides first because they explained and justified why he would sell less [than last year] while spending more [than last year]. After a while I realized these weren’t really planning sessions at all. They were compensation negotiations disguised as budgeting and planning. Everyone wanted me to approve as low a profit target for their group as possible so they could easily exceed it and get a big bonus. I learned that a key to making money in sales at Oracle was negotiating a low profit target or quota. I decided to fix that problem.”
Ellison was determined to get involved in another area that had been entirely Lane’s responsibility: the design of sales force compensation. Patronage is power, and, after hire and fire, there is no greater power within a company than setting the remuneration of the sales staff. By deciding that he, rather than Lane, would in future determine how the sales force was paid, Ellison was signaling in the clearest possible way that Lane’s power was in terminal decline.
In any sales-driven organization, the way in which the “comp plan” is structured is a vital ingredient in determining success or failure. The objective of a good comp plan is simply stated: it should be transparent and fair and encourage the achievement of carefully worked out “stretch” targets; the salespeople who do most for the bottom line should be able to get rich themselves, while those who are hoping for a free ride should be penalized and cleared out. In other words, the best comp plans are simple to understand and work with the grain of human nature. When comp plans start delivering the opposite of what’s intended, it’s usually because they have become too clever by half. That was precisely what Ellison thought had happened at Oracle.
Jay Nussbaum recalls, “The 1999 budgeting sessions were a real watershed. I was telling Larry about how there was so much fucking abuse and waste in the system. And Larry’s saying, ‘Tell me more, tell me more, tell me more.’ So I called Ray and told him that Larry was asking me a lot about compensation and do you want me to brief you about it. So I started to tell him, and he said, ‘You have no fucking right to talk to him about that.’ I said, ‘What am I to do? He’s the CEO. If he calls me and says, “Talk me through this compensation,” am I supposed to tell him I’m under instruction from you not to talk to him? Let me help you and then you can tell him.’ And from that moment, it was like we were on different sides of the world.”
Ellison says, “Ray would experiment with different compensation ideas and sales force organizations every year. One of the worst ideas I can remember was when Ray decided we didn’t do enough selling through partners. The sales force convinced him that the way to fix this was to pay more money to the sales force if the deal went through a partner than if the deal came directly to Oracle. For example, if you sold a million-dollar deal directly, Oracle would get a million dollars and you would get a $100,000 commission. But if you sold a million-dollar deal through a partner, Oracle would get $600,000 and you would get a $120,000 commission. Needless to say, our sales force pushed as many deals as they could through partners that year, so the partners were happy. The sales force got higher commission payments for going through partners, so they were happy. The only loser was Oracle Corporation. We ended up paying huge partner bonuses and larger sales commissions, so our sales margins went down. The only possible justification for this extra pay for partner deals was if the partners started bringing in large numbers of deals on their own. But they didn’t.
“Our sales managers converted Ray to the fundamental belief that the best way to change the behavior of our sales force was to pay for it. So when Ray wanted to encourage better cooperation among our different sales groups, he decided to pay multiple groups on the same deal. We had industry sales teams, product sales teams, geographic sales teams, global sales managers, etc. All could get paid on the same deal. They wouldn’t split a commission—no, no, no—instead we’d pay double commissions, triple commissions, and so on. But Ray told me not to worry because we could afford to pay everybody on the same deal so long as we raised everybody’s quota. I said, ‘Ray, paying people for things they don’t sell is not a good idea. Raising quotas is not a good idea.’ So one of the very first things I did when I took over the sales comp plans was eliminate most double compensation on deals. It allowed us to bring quotas down dramatically. And it smoked out the people who were making a living getting paid on other people’s deals. But something else happened when I took control of the comp plans. When Ray lost the power to pay the salespeople, he lost most of his influence in the organizations that still reported to him.”
The final nail in Lane’s coffin was the unheralded arrival in April 1999 of Safra Catz, a former investment banker from Donaldson, Lufkin and Jenrette (acquired in 2000 by Credit Suisse First Boston) who had covered the software industry since 1986, the year Oracle had gone public, and who had become a close friend and admirer of Ellison. Catz, a petite, black-haired Israeli American, had come to take Ron Wohl’s old job as Ellison’s chief of staff. Ellison had been keen to hire her for some time but had initially thought she might run one of the companies he had recently bought in Israel. But Catz made it clear that if she was going to work for Ellison in any capacity, it would be at Oracle itself. Disillusioned by the investment hysteria over dot-com firms that had little or no proprietary technology, she was ready to quit banking. She suggested to Ellison that she could come and do corporate development—a role that was vacant following the departure of the able David Roux to run the interactive television company Liberate Technologies.
Catz had been an Oracle partisan for years: “I was already on the Oracle bandwagon. I said that if Pets.com or Ariba trade at those kinds of numbers, then Oracle’s worth a hundred or a thousand times that. This is kind of the center of the universe, right, not Wall Street. You want to be in the sun, this is the sun. I thought, you know, shame on me if I don’t try this. And of course I thought Larry was great, awesome, with this ability to think without walls.”
Catz didn’t have
any operational experience, but Ellison quickly realized that she had precisely the combination of qualities that he needed to help him run Oracle. As a lawyer, she had an almost forensic approach to digging out information, while her thirteen years as an investment banker had left her with finely honed analytical skills. She says, “I came in with absolutely no agenda other than to help Larry. That actually made my job incredibly easy. If Larry wants something done, now it happens because I’m going to check that it has. That was the thing that was really missing. Larry would make a decision, and nobody would check that it had been carried out. People would hope that because he had so many things to do, he would forget.”
The first thing she did on her arrival was analyze previous years’ budgets in the light of those submitted for the next financial year. She says, “It was clear we were spending a lot of money for no good reason, that there was an enormous amount of duplicative effort and overlapping objectives.” Much of the inefficiency that Catz found, she laid at Lane’s door. “We had any number of initiatives in place that when I got up the actual numbers it was obvious that they would lose us money. Ray would say, ‘This is what the customer wants to buy. He wants lunch. You want to sell him lunch?’ I was looking at this stuff and saying to Larry, ‘Did you know this. As far as I can tell so far, I can find at least $100 million we’ve lost.’ ”
Ellison says, “Safra is a slave to facts. She is unwilling or at the very least extremely uncomfortable about making decisions without the facts. Not all managers are like that. Some managers rely on their ‘gut’ to help them with difficult decisions—especially if the facts are hard to find. The fundamental clash between Ray and Safra was that Safra dug out the correct facts while Ray relied on what people told him. In an argument when nobody has any facts—and I’ve seen a lot of those—the person with the strongest personality wins. But when one person has the facts and the other doesn’t, the one with the facts always wins. When both people have the facts, there’s no argument. Safra never tried to embarrass Ray in a meeting. He just kept saying ‘This is true’ or ‘That is true,’ and she would say, ‘No, I don’t think that’s true and here’s why.’ It was painful to watch.”
Softwar Page 21
