Softwar

by Matthew Symonds

  Ellison says, “Ray’s solution was to put Robert Shaw in charge of everything. But Robert had never developed a software product in his life. Finally, out of extreme frustration and great fear of putting Robert in charge of development, I said, ‘Okay, I’ll be Mr. Applications. I’ll do it. I can do it. I don’t want to do it. I really don’t want to do it.’ Order management, tax tables, accounting—are you kidding me? I’m working on parallel server, man. I didn’t want to go from computer science to accounting. I thought that understanding and automating all these business processes would be dull and duller. It just wasn’t a book I wanted to read.”14

  As for Lane, he says, “I had no choice . . . I sent an e-mail out to the entire field saying Larry is taking over applications as head of development. I showed it to Larry before it went out. He did a few edits to protect Ron. But it was very direct.” What Lane didn’t realize was the speed at which he was losing Ellison’s confidence and support. It wasn’t long before Ellison was able to convince himself that after the thwarting of his power play to replace Wohl with Robert Shaw, Lane was now hoping that, having taken responsibility for applications, Ellison himself would fail.

  Jeff Henley thinks that Ellison may well have felt threatened, but he has no time for the idea that Lane was trying to take Larry’s job. George Roberts, who succeeded Barry Ariko as head of U.S. sales, is convinced that Lane never wanted Oracle to be anything other than successful and, while he might have sounded off against Ellison at times of great frustration, never thought he could replace him. Jay Nussbaum, no ally of Lane, says, “Let me make something perfectly clear. At this point, Ray had such a huge financial upside that if Larry failed, Ray also failed. There was no rooting for Larry’s failure. None whatsoever.” Gary Bloom, who was shocked by what he saw as Lane’s undermining of Wohl in public, agrees: “I don’t think Ray wanted Larry to fail. I would not put that burden on him. I think they had different views on how to fix applications and how to make it better.”

  The truth, I think, is that Lane was determined to make Ellison take his responsibilities for fixing applications seriously and that the best way of achieving that was to put his boss under as much public pressure to deliver as possible. When I put the question to Lane, he looked shocked and sad. “I’m going to give you a very strange response to this. Had Larry read the situation right and had he worked on the real problem, . . . then maybe we’d still have a relationship and there’s no question Oracle was much better off with both Larry and me there as a team. He totally misread it. I was tired of not being competitive. I was tired of Ron producing sucky applications that we couldn’t sell, and I desperately needed somebody to replace him. When Larry said, ‘How about me?’ I totally supported that. I went out and sold that strategy to everybody. I said, ‘Larry’s going to do it,’ and three months later he comes up with a strategy. At first I doubted he would spend the time, but he did. I didn’t want to run applications. It wasn’t a power play at all.”

  In which case, why did Ellison, who so recently had seemed to think that Lane was close to being the perfectly complementary number two, someone he was prepared to give virtually equal billing in what was now the world’s second biggest software company, turn so dramatically against his deputy?

  * * *

  1. LE writes: It was a big mistake, and it was my mistake. I didn’t think that Microsoft Windows would crush IBM OS/2 and all the other desktop systems—but it did.

  2. LE writes: Ray and I both agreed that our database sales force was not very good at selling applications. But we disagreed as to why and what to do about it. Ray thought the problem was they didn’t understand how to sell applications due to lack of training and relevant industry experience. I agreed with those points but thought the problem was compounded by the fact that our salespeople would be foolish to spend their time competing for applications deals when they could make much more money by concentrating on selling the database. No amount of training and experience could ever solve that problem, so I thought we should create a separate, dedicated applications sales force. But I could never convince Ray. He wanted Oracle to have one sales force because he insisted that customers wanted to deal with one and only one Oracle sales force. This grew to become a huge point of contention between Ray and me, and it wasn’t until after Ray left that we began splitting and specializing our sales force.

  3. LE writes: Robert’s analysis is right on. Building our own consulting organization was a two-edged sword. There were as many pluses as minuses.

  4. LE writes: The biggest problem in applications development was the mediocrity of the management team—people who were promoted because of their unquestioning loyalty to Jeff Walker. Ron’s job was to analyze the organization and figure out who was good and who was smart, and put them in charge. Ron has no tolerance for muddy thinking, so I knew he would quickly find and weed out the mediocre managers. Ron also had to prioritize what features we were going to put into the applications to become more competitive. We needed someone extremely detail-oriented and very, very analytical. Ron is one of the best thinkers in the company so I gave him the job.

  5. LE writes: If consulting claimed that a product bug had cost them extra un-billable time at a customer, Ray would give the consulting team “budget relief” and the team’s bonuses would not be affected. It sounds reasonable, but it created a perverse incentive to blame everything on product bugs. If consultants admitted they had made a mistake, their bonuses would go down. If consultants blamed it on development, their bonuses would be safe. Not surprisingly, just about every delayed and failed consulting project was blamed on the product. This perverse process of paying for the placement of blame poisoned the relationship between applications development and consulting.

  6. LE writes: Verticalizing an applications sales force by industry made sense. Verticalizing our database sales force did not. We needed to segment into separate sales forces. But Ray refused. He verticalized everything. It was a costly mistake.

  7. LE writes: Frankly, I was getting tired of hearing Ray tell me over and over again how he could do a better job than Ron by having consulting build vertical industry application suites by integrating best-of-breed third-party products. I decided to let him try. If he succeeded, great; if he failed, he would better understand the complexities of application development. Expensive mistake. I never should have let Ray build and sell best-of-breed suites.

  8. LE writes: Joe had joined Ray in demanding that I fire Ron. But I don’t think that Joe wanted to stop there. I believe he wanted me fired as well.

  9. LE writes: One of the most expensive mistakes I have ever made at Oracle. I never should have let Ray do software product development of any kind.

  10. LE writes: In fact, we did win more deals. But they weren’t profitable.

  11. LE writes: A little tougher? Nobody has found a rapid, low-cost way to integrate a lot of applications from different vendors. Not yet, anyway. But I have to give Ray credit. He didn’t give up on the idea after he left Oracle. He immediately helped start up a couple of companies to solve the application integration problem, this time using Web services. So far these companies have been unsuccessful. It’s a hard problem.

  12. LE writes: Nonsense. IBM does software product integration with a lot of labor for a lot of money. Low-cost software integration is still beyond the state of the art.

  13. LE writes: Ray’s managers constantly lobbied him for expense relief, and over time he gave in to it. Once that happened, managers who wanted to increase their profit bonuses could do it one of two ways: cut real expenses or lobby Ray for more expense relief. Thus the company’s profits went down while managers’ profit bonuses went up. Oracle now has a very simple way to hold managers accountable for expenses: If a manager makes the decision to spend the money, then that manager pays for it out of his budget. If somebody else decides to buy something and you use it—like our global network, for example—you don’t get charged back for it. I believe managers can be measured a
nd held accountable only for expenses they control.

  14. LE writes: After I got started, I found engineering our business processes (it wasn’t reengineering because they had never been engineered in the first place) was just as interesting as engineering the database.

  8

  FALLING OUT

  On June 30, 2000, Ray Lane was enjoying the last day of a family vacation when a call came through from Larry Ellison. Lane remembers, “Larry says that he didn’t know I was on vacation and that he’s sorry about calling me and telling me this, but he wants to take the president’s title back. I said, ‘Really? Why?’ He said, ‘What we’re doing here at Oracle, the reengineering of the company, is so important that authority has to be driven from my office. The whole company needs to understand that there is one single centralized point of authority, and it will be with the CEO.’ I said, ‘Well, Larry, I have never really opposed that. I’m on your team.’ He said, ‘It’s kind of like kids playing off Mom against Dad. If they get a no from Mom, they go to Dad.’ He was very uncomfortable because I was asking direct questions. I said, ‘Larry, let’s just step back. What went wrong? Everybody in the industry would look at this team and say that this is the reason I want to do business with Oracle. You on the technology and the vision; me on the business. What went wrong?’ He said, ‘I don’t know. It’s like a marriage that went bad. I don’t know.’ ”

  Characteristically, Ellison did what he could to avoid an embarrassing confrontation. The main reason he gave to Lane for wanting him to leave was his concern about how divided the company had become with two leaders. Ellison says, “I explained that . . . there was too much ambiguity of authority at a time when we were trying to reinvent ourselves and restructure the company to make it more centralized, more automated, more of an e-business. It’s very difficult to make changes when a lot of the existing managers didn’t believe in the changes I wanted to make. Most of our senior people liked being autonomous general managers. They wanted to be given a budget and left alone to run the business as they saw fit.”

  Lane insists that Ellison didn’t have the stomach to sack him. “Larry was incapable of firing me. I said, ‘I guess I have got to leave if you’ve taken back the president’s title.’ He said, ‘Well, that’s your choice.’ I now wish I’d said, ‘Okay, you’ve got the title. What do I call myself now?’ He couldn’t fire me. He doesn’t know how to do it.”1 The two men then briefly discussed a press release, and Lane suggested that they should talk about the terms of his departure. He had $70 million worth of options due to vest in three weeks, and he proposed a termination agreement “so that I say good things about Oracle and we both have the same story about why I left—Larry said, of course, ‘Let me look into it and I’ll call you back.’ He never called back.”

  The terse press release that Oracle released that evening offered no explanation for Lane’s sudden departure. Attempts had been made by the PR department to dress it up with a tribute from Ellison to Lane’s years of service. But Ellison admits that in the end, he couldn’t bring himself to say what he didn’t feel; Lane had already been more than adequately rewarded for his time at Oracle. Lane says, “The press release was in the papers the next day. My mother, my sister, my friends were all with me in Oregon, and they got the paper the next day. What’s this? So I had to explain it to the family. It was totally below the belt.”

  Nor was there any termination agreement. Lane claims that the board tried to put together something that was “fair”; a payoff of $13 million was suggested. Ellison rejected it, offering instead $1 million in cash. Ellison says, “I can’t understand why a receptionist who gets laid off gets two weeks’ salary and has to worry about paying the rent, when some rich senior manager who made millions gets additional millions when they get laid off.”

  Lane refused to sign the deal. He really didn’t need the money anyway. In subsequent interviews, his version of events was that he had resigned over changes within the company that Ellison was proposing to make and with which he disagreed. For his part, Ellison insisted that Lane was leaving because, as a result of the way Oracle had recently centralized its operations around the Internet, he no longer had a big enough job to do. Only later, when he could no longer control the urge to trash his former business partner, did he admit that he had fired Lane.2

  What happened was that the man who many people outside Oracle believed to have been responsible for the firm’s reinvention and much of its success in the 1990s had been slowly frozen out and systematically stripped of authority. When Ellison said that he had no need to replace Lane, it was because there really wasn’t much left of Lane’s job by the time he left.

  Within three years of his arrival at Oracle, Lane’s reputation in the industry was so high that he was soon receiving a stream of job offers. With most, he never went further than the first step. However, there were a few that Lane was willing to explore further. One came from John Doerr, a senior partner at Silicon Valley’s most exclusive venture capital firm, Kleiner, Perkins, Caufield & Byers, and Jim Clark, the former boss of Silicon Graphics and the founder of Netscape, the software firm that triggered the mania for investing in Internet with its pioneering Web browser. Doerr and Clark wanted Lane to come and run their latest big idea, a company called Healthscape (later Healtheon and later still WebMD) that was going to bring the Internet to medical practice. After taking a close look at what was involved, Lane said no. “I didn’t see myself as a start-up guy.”

  Much later, Lane would be on the short list of candidates to become CEO of Hewlett-Packard, the stately computer hardware company that launched the Silicon Valley phenomenon in the 1930s. But the job he came closest to taking was with the once high flying networking firm Novell. At first he wasn’t interested. Novell, once the leader in network operating system software, was being badly beaten up by the onward march of Microsoft’s Windows NT, while its GroupWise collaboration software was being squeezed on both sides by Lotus Notes and Microsoft Exchange. Novell still had plenty of cash, but what it lacked was any kind of strategy to get it out of its hole.

  What changed his mind was an unannounced visit by David Burn, one of the toughest and most persuasive headhunters in the business, and John Young, who was not only a Novell board member but a highly respected former chairman of Hewlett-Packard. “They came into my living room and put this offer in front of me that included $15 million of cash that would be put straight into my bank account and $4.5 million in options at the current price, which was an all-time low for Novell. And I’m just going ‘holy mackerel’ and said, ‘Give me an hour and I’ll just write down some thoughts about a strategy for a turnaround, because that will help tell me whether it’s possible.’ So I wrote four pages that I shared with them and said, ‘Here’s what I’d do,’ and they said, ‘That’s why we want you—you’ve just written a strategy on the back of an envelope. . . .’

  “I was nervous. I was an Oracle devotee, there for life, and I was very, very nervous about this. But I was saying, CEO, turnaround, lots of money, and I agreed that I’d go and see Larry and tell him I was resigning. So I went to see Larry that afternoon in his house and I said, ‘I’ve got an offer from Novell.’ He said, ‘Novell!’—just like that. Novell, that brain-dead company. I said, ‘Larry, look at it my way, it’s a turnaround opportunity and they’re going to pay me all this money.’ He said, ‘You’re not going. Their server isn’t programmable. You have no APIs to build applications. It’s dead meat, and Microsoft will take them out.’

  “So he basically had scared me out of taking the job anyway, but then he said, ‘How about two million options, would you stay for two million options?’ I said, ‘Yes, it’s more than adequate—I’m not doing this because I think Novell’s great, it’s for the money.’ I went to Young and Burn, and said, ‘Sorry guys, I’m staying at Oracle.’ That night I went to the fiftieth birthday party of Robert Shaw’s wife and Larry called me on my cell phone. He said he’d talked to the board and he thought $2.5 million i
n options was the right number. You deserve it. I thought he’d gone way overboard, so of course I stayed. I didn’t find out until I left Oracle that the board was pissed off about this. No one ever told me, and I certainly wasn’t holding Oracle up for money.”3

  Ellison’s recollection of what happened is a little different. He says that the figure of 2 million was Lane’s price for staying at Oracle and that the main reason he increased the number to 2.5 million was that, perversely, by volunteering additional options he felt a little more in control of the situation, a little less responding to blackmail. He also wanted to ensure that Lane really would give Novell the brush-off. That said, the most convincing explanation for Ellison’s excessive generosity to Lane is that he panicked.

  His first reaction was sheer amazement that Lane actually thought that he might be able to fix Novell. “I said, Ray, Novell doesn’t have a service problem, and it doesn’t have a sales problem—they distribute through third parties. Those are the things you’re best at. What Novell has is a very serious product problem. I don’t know what I’d do if I went to Novell. I think it’s too late.” However, his second reaction was fear that despite his demolition job on Novell’s prospects (correct, as it turned out: even Eric Schmidt, a first-class technologist, proved unable to save Novell) Lane might still go through with his resignation. “At this point in Ray’s career, he thought he could do just about anything—even turn around Novell, which I thought was impossible. But I couldn’t afford for him to take the Novell job. I felt terribly vulnerable because he was running such a large portion of the company. With the problems we were having in applications development, which was where I thought I should focus my time, I didn’t have the bandwidth to take on anything more. Besides, I really didn’t know how marketing ran or how sales ran. I just didn’t feel I had the time or competence to take over those things if he left . . . so I paid him off.”