He didn’t have any better idea how his March 27 promise was going to be achieved either. But a commitment was a commitment, and it was Gary Kennedy’s job to deliver the number. Astonishingly, given a slowing economy and Oracle’s continuing product problems, he just about made it. Revenue was up 63 percent, to $334 million, and earnings per share came in at a respectable 39 cents. A great deal has been made of Kennedy’s idea of incentivizing the sales force by giving them the option in that crucial quarter of having their commissions paid in gold coins. Kennedy’s “Go for Gold” campaign is a nice metaphor for the hysteria that accompanied selling Oracle software during that period, but it was relatively harmless compared with the other things that were going on.
The real damage was being done by the way future sales were being mortgaged against the present. Customers were being encouraged by outrageous discounts to buy software licenses they might not actually need for years. By the time the next quarter arrived, instead of the “planned” 50 percent increase in sales, the figure was closer to 30 percent. With expenses still geared to doubling revenues, Oracle was heading towards its first-ever loss. Geoff Squire, who was by this time running European sales from London, says, “Every customer had been squeezed as much as possible to help the May quarter, with the result that the cupboard was really bare. I got a call from Larry in late August, and he said, the quarter’s gone and Gary’s going. He said he was going to focus personally on fixing America and he wanted me to take over International.” Squire argued successfully that International could look after itself and that he would come over to America to help Ellison restructure the company. He started by giving Ellison a list of all the things that needed doing. Ellison interrupted: “Okay, what time does your plane land?”6
When Squire arrived, he found Walker telling Ellison that with the company about to miss its numbers by a mile, he had to make a statement. A warning issued two and a half weeks before the scheduled earnings call suggested a loss of 20 cents a share. Within days Walker was saying that he had gotten his sums wrong. Squire says, “By this time Larry was getting a little upset about this—that was two hits. He didn’t scream and shout, but you could tell he was angry. I was in the room when he took the decision that Walker should go. He didn’t say anything. It was just a realization—he simply wrote a note to himself and carried on with the executive committee meeting.”
When Walker announced the real losses during International Oracle User Week in Anaheim, California, on September 24, 1991, they were 27 cents a share and $36 million. Ellison, having fired Kennedy a couple of weeks earlier (Kennedy claims he resigned because he had been offered the chance to lead a Mormon mission and wasn’t prepared to commit another two years of his life to Oracle), was eager to put as much of the blame for the “fiasco” on a botched reorganization of U.S. sales. He was also able to point to the fact that he was at last pruning costs, having laid off four hundred people a week earlier. Squire says, “We basically had to find a revenue number we couldn’t miss and then build a cost plan underneath, which meant getting rid of an awful lot of people.” By way of painful irony, the day the layoffs were announced was also the day that the magnificent pool and gym complex opened at Oracle’s spectacular new campus at Redwood Shores.
Whatever Ellison had said, it wouldn’t have cut much ice. On November 1, less than eight months after Oracle’s stock had reached an all-time high of $28.38, the shares were trading for $4.88. To make matters worse, Oracle also had a serious cash crisis. It was this, as so often happens with ultra-fast-growing companies that suddenly hit a wall, that turned a painful reverse into something almost terminal. Jeff Walker stills likes to play down the seriousness of what had happened to Oracle: “It was one lost quarter coupled with some accounting irregularities as a consequence of rapid growth, lack of central control, and a far-flung empire of somewhat autonomous organizations. If the company had held adequate cash at the time, it would have been nothing more than a little blip. What made it a near-death experience was the lack of cash. As the CFO of Oracle, I and the board worked hard to encourage Larry to raise cash in a variety of ways throughout the late 1980s. I wanted to do additional equity rounds. We came within days of doing convertible debentures to raise large amounts of cash. Larry at the last minute decided not to do it. The board could have done something about it had it chosen to do so.”7
Ellison was determined to avoid the dilution of raising new equity and by his own account was too uninterested to read the new numbers. A prudent insurance policy against a downturn in the economy or Oracle’s fortunes was simply too boring, and the board did not press him on it. It was a recklessness that came very close to losing him the company.
In August, Oracle had entered into a three-year revolving loan agreement of up to $250 million from a thirteen-bank syndicate, but by December the banks were so shocked by the deterioration in both Oracle’s business and its stock that it forced a renegotiation of the terms. The size of the credit line was scaled back to $170 million, the maturity date was brought forward to November 1992, and additional covenants relating to Oracle’s future financial performance were imposed. Unfortunately, most of the $170 million had already been borrowed. Squire, who by now was in charge of worldwide operations and effectively number two at Oracle, was doggedly trying to bring in cash by running the receivables down: “I’d ring the customer up and ask them why they hadn’t paid. It was always the same thing: the product didn’t work; I never ordered the product; the salesman said I didn’t have to pay until whenever . . .” Squire, an accountant by training, remembers that the cash position was so bad that at one point Oracle was within $3 million of its overdraft limit.
The danger of a hostile takeover was small because without its key people to develop the product, Oracle had little value to a rival. A much bigger danger was that those people would walk because their options were so far under water that a deep-sea diving bell would be needed to find them and because many thought that the company was finished. Squire says, “Every time the phone rang, you feared it might be a headhunter. They were all over the place like leeches, trying to take all our top people away.” One of the brightest talents to leave was Tom Siebel. Siebel was not only one of Oracle’s most successful sales managers, he had designed and built from scratch a highly effective telephone sales operation running on an early version of the customer relationship management software that he was later to pioneer. He said later, “Oracle looked to me to be an organizational disaster at that point. Larry offered me the position as president of Oracle USA to stay, an offer which I declined. I was tired. I was vested. I was looking for a higher-quality work environment in an organization committed to higher ethical standards and committed to fulfill customer obligations. I left.”8
For Ellison, a personal as well as professional disaster was unfolding. He had borrowed heavily against shares that had slumped to a point where he faced the equivalent of a margin call from his creditors. And a great many people were only too happy to see him getting his comeuppance. His marriage to Barbara had broken up, and Bob Miner, who should have been his closest friend at Oracle, wanted out. Ellison says, “Even after our stock dropped, Oracle was worth much more than Bob ever imagined. He thought the prudent thing to do was to sell the company.” It was personal in another way for Ellison. His adoptive father, Lou Ellison, had always told him that he would never amount to anything. “My father said I would never succeed. It seemed he might be right after all.”
Although he was clearly shaken and depressed, it didn’t occur to him to give up. Geoff Squire says that he never thought that Ellison would walk away. “He had some low points. I remember, the day my father died, finding my mother clinging to the bedpost not knowing what to do because the rock on which her life was built was no longer there. I saw Larry almost like that a couple of times. There was desperation. But what he clung to was his belief in the company.” Jeff Walker says, “My observation is that he dealt with it in a forthright and manly way. No one
was happy about it, but there were things that he had to do and decisions that he had to take. I never saw petulant or childish behavior. I never saw withdrawal or a fear to face facts. Pressure can make us react in perverse ways, but I never saw any of that. He dealt with it.” Ellison says, “I couldn’t run away. I had to save Oracle to save myself. I had no choice.”
In one sense Ellison was very lucky indeed—he could easily have lost his job before he had a chance to fix the mess he had helped create. Don Lucas says, “I can say this now, and Larry knows this, that not all the board was in favor of keeping Larry on. I was chairman, and I was one hundred percent convinced that this was not a good time to make a change. Larry was inspirational to the company, he understood the technology, and despite the problems with Oracle 6 the technology was great. The problem was in not having proper accounting and the need to clean up and push that out to the field, to sales, and say, ‘Thou wilt not do this, and this is the way you will do things.’ This was a simple problem. What we really needed to do was get a good financial officer.” Ellison says, “I’ll always be grateful to Don for trusting me and giving me another chance, even though I probably didn’t deserve it.”
How much of the blame for what happened should stick to Ellison and how much to Kennedy and Walker? As CEO, Ellison clearly has to take overall responsibility. The extreme form of delegation he pursued—what he describes as his abdication model of management—was a form of self-indulgence rather than a brilliant insight into how to get the best from talented managers. “I was interested in the technology. I wasn’t interested in sales or accounting or legal. If I wasn’t interested in something, I simply ignored it. The managers who wanted to run their part of the business with minimum interference were quite happy that nobody was looking over their shoulder. Unfortunately, if people ran amok, the problems they created would not become visible until they had grown so large that they were difficult to fix. All I had to do was keep legal and accounting organized as separate control functions, and a lot of the problems we had in ’90–’91 could have been avoided. You cannot run a company without strong checks and balances.”
Ellison’s defense today is still pretty much what it was at the time—that he was naive and inexperienced. But he knows that it’s not enough to absolve him. He adds, “I just wasn’t paying proper attention to my job. I was doing only the things that interested me. It was the same problem I had in school. But this happened in my forties. I wasn’t a kid anymore.” Exactly. The truth is that Ellison found almost everything about running a company, other than creating a vision of what Oracle’s technology could do and driving the product development strategy to achieve it, boring and, in some sense, beneath him. But later, after he realized he had a duty to know about the supposedly tedious details of how a sales force should be compensated and revenue recognized, he discovered it wasn’t so dull and trivial after all.9
Ellison also had a genuine respect for Walker’s intellect, which blinded him to his other faults. “I like very smart people, and Walker’s very smart. In those days, whenever I was defending somebody, my defense would be to point out how smart they were. Geoff Squire wasn’t impressed by this argument. He said, ‘Yes, Larry, he’s very smart, but can he do his fucking job?’ I just stared at Geoff and said nothing, but I was thinking, ‘Oh, God. He’s right. Brilliance is not enough.’ ”
As far as Kennedy was concerned, Ellison seemed to think that his religious fervor would prevent him from doing anything morally dubious. “I never thought Gary was capable of being unethical. But I also knew he hated losing more than anything.” Ellison used to play basketball with Kennedy. On one occasion, the Oracle team was down by twenty points in the fourth quarter to a team from Stanford. “Late in the game I passed the ball to Gary and jokingly said, ‘Hey, we can still beat these guys.’ Gary gave me a look of total disgust. How could I give up before the game was over? He hated losing. The feeling was so strong, it could only have come from his childhood. Gary came from a very modest background [his father worked as a miner in Wyoming, forty miles from the family home in Randolph, Utah]. In his family, you were taught never to give up. Gary would fight to the limit not to lose.”
But surely that refusal to accept defeat was precisely the quality that convinced Ellison to put Kennedy in charge of U.S. sales? When Ellison committed Oracle to doubling its revenues each year, when he presented the “Five in Five” plan, didn’t he realize he was creating the conditions in which somebody as driven as Kennedy would be unable to stop himself from going too far?
I think Ellison is disingenuous in arguing he was not responsible for a good deal of the pressure that turned the Oracle sales force into an out-of-control monster. If a demand from the CEO to keep doubling revenues each year, even with a shaky product and weakening economy, is not pressure, it’s hard to know what is. The frequently declared determination to wipe out database rivals, such as Ingres and Ashton-Tate, and the testosterone-charged advertising that Oracle indulged in to convey its message of dominance—one series of ads depicted Oracle as a sleek F-15 shooting down Ashton-Tate’s World War I biplane—bred a win-at-all-costs culture that came at least as much from Ellison as from Kennedy. Given Ellison’s abdication model of management, it was also surely his responsibility to have a qualified, tough CFO to establish and police business rules. Making Walker CFO just because he was “smart” and was designing financial applications was a decision of extraordinary frivolity, while integrating the legal department with U.S. sales was equally irresponsible.
In partial mitigation, however, it has to be remembered that the software business more than a decade ago was extremely immature. Until the early 1980s, nearly all software had been sold with hardware according to the vertically integrated model established by the likes of IBM and Hewlett-Packard. Enterprise software, because of the peculiar nature of the product and the time it usually takes to install, lends itself, more than anything else, to companies that are likely to spend large sums of money on exotic and innovative deal making.
Don Lucas says, “Software by its very nature is an intangible. The same tape can be worth $100,000 or $200,000 and so is susceptible to a vast number and variety of business relationships. And of course at that time we didn’t have standardized contracts, and the accounting profession was just coming up to speed and, you know, we were called cowboys, but there were many different ways to enter into a contract, with the provisions and the details and so on. Some salesmen gave customers the right to return the software. Of course that’s lunacy, that’s not competent. But in between that, which is zero, and a clean booking, cash down, there’s a vast array of considerations.” Ironically, given Tom Siebel’s strictures about the ethical standards at Oracle when he was there, Ellison always cites Siebel as a salesman whose “creativity got him and the company into trouble.” “Tom was and is a super salesman. He is the king of the deal. But not every deal was good for the company.”
Ellison hasn’t seen Kennedy since he left Oracle. (To almost everyone’s consternation, when Jeff Walker founded Tenfold in 1993, he asked Gary Kennedy to be its CEO, even if it meant basing the business in Salt Lake City.) “Gary and I didn’t have a lot in common other than work and the occasional basketball game. We never spent time together.” Although he won’t quite say it, it’s clear that he still blames Kennedy in large part for Oracle’s crash and for the humiliation of having to restate earnings for three quarters and submit to a brutal SEC investigation that resulted in a $100,000 fine. It has left Ellison with a profound distrust, not to say distaste, for the things that salespeople will do to get a deal.
In one particular, however, Ellison does recognize something of himself in Gary Kennedy. “Like Gary, I’ve always been more motivated by fear of failure than by greed. And I hate losing. But bending the rules is a bad strategy. It’s more likely to bring on defeat than victory.”10
There is something slightly disturbing in Ellison’s treatment of Kennedy. Ellison gives the impression of having
been almost Kennedy’s victim. Yet Kennedy recruited the most talented and feared (by both competitors and customers, it has to be admitted) sales force the technology business has ever seen. Some of them may have gone off the rails and gotten Ellison into trouble, but without them it’s doubtful that Oracle would be the powerhouse it is today. Ellison likes to believe that, by and large, great products sell themselves. But Oracle’s were not always unambiguously better than those of its rivals, whereas its sales force was invariably smarter, tougher, and more relentless.
The question of how much blame attaches to Ellison remains difficult to gauge. One thing he accepts without equivocation is that he practiced management by abdication. He does not deny negligence. There is even some justification for his abdication model. It wasn’t just that he enjoyed driving product development more than the other stuff, he was convinced that it was in that area that he could make the greatest contribution to Oracle’s success. While Bob Miner was supposed to be running development, he was never prepared to lead his programmers in the kind of “death marches” that winning software companies have to be ready for when products are late or failing.
Ellison says with fond exasperation: “Bob loved and took care of the people that worked for him. He would never ask them to work weekends because he thought they should spend time with their families.” The fact that Oracle was having such a terrible time getting Version 6 of the database to work properly and the advances that were being made by Sybase made that focus on the product all the more urgent. Ellison says, “To beat Sybase, we had to deliver Oracle Version 7. That’s what I spent my time on. I ignored everything else.”
Softwar Page 12