A review of both investigations by the Economist Intelligence Unit concluded, “These studies provide impressive evidence of a company in the midst of dramatic changes inspired by the Internet—although as Oracle’s senior executives freely concede, many of the efforts at transformation derive as much from old-fashioned sound business practices as they do from Internet-driven innovation.” Pointing out the huge gains in Oracle’s own IT operation—spending nearly halved since 1999 and average help desk response times down from three minutes to less than ten seconds with 44 percent fewer IT staff—the EIU summed up, “The company undertook simultaneously to transform every aspect of its operations in accordance with three key principles: standardisation, centralisation and web-enabled automation. Driven with near-fanatical zeal by the CEO, Larry Ellison, this trinity has shaken up virtually every job function within the company. There is little doubt that an Internet-based business transformation has taken place.”
What the EIU and most others fail to note, however, is that while standardization and centralization may sound like “old-fashioned sound business practices,” it was the Internet that made it possible to achieve the Ellisonian ideal of “one network, one database” so rapidly and so completely. The suggestion that most of the savings were simply the result of cutting head count is absurd; it was only successful automation that meant Oracle could do more with fewer people. As for the accusation that much of the margin improvement came from the revenue momentum fueled by the dot-com boom, that too appears simplistic.
The second quarter of fiscal 2002 was hard pounding for Oracle—its earnings per share of 10 cents were a cent lower than the previous year, while new software license sales were down 27 percent and services revenue was flat. Announcing the numbers, Jeff Henley said, “For Oracle, this quarter included the September eleventh tragedy and its impact on an already weak economy. In a difficult economy, we are gratified that our business automation has allowed us to continuously take cost out of our business while expanding our engineering capacity.” Ellison put it differently: “It was our toughest quarter in a decade, but we still made over $800 million in operating profit and a thirty-five percent operating margin. When the economy improves, we will earn a lot more.” Ellison was pointing out two things. The first was that in extremely difficult selling conditions its margin gains had not deteriorated despite the fact that, almost alone among technology firms, Oracle had not been forced into large-scale layoffs in 2001. The second was that when business picked up there was likely to be a nearly one-for-one relationship between increasing revenue dollars and profit—Ellison believed that Oracle had effectively uncoupled expenses from revenue growth and that in an “up” economy, Oracle was capable of getting near a 50 percent margin that would beat even Microsoft’s phenomenal profitability into a cocked hat.
Software was as much a catalyst as a cause of the revolution in Oracle’s business practices. Ellison frequently tells prospective customers that unless they are prepared to review all their business processes in the light of the Internet and unless their top management is willing to make available the best people in the company to work out what should be done, they shouldn’t buy Oracle’s software because they will only be disappointed by the return on their investment. His constant message is: Don’t change the software to fit the way you have always done business; use it to figure out how you want to do business for the next twenty years. Gary Bloom, who left Oracle to run the software firm VERITAS in November 2000, agrees. “I was the driving force behind the idea that the way we were spending money and making business decisions at Oracle was lunacy. But the software forces you to think about exactly how you want to do business, and that makes you go do it. It gets your attention, and it’s very powerful.”
Ellison believes that one of the greatest benefits of Oracle’s e-business transformation has been to the way people within the company now behave: “Oracle had been a company made up of many independent business groups, managed by these self-reliant generalists who valued their autonomy. Now it’s a company of interdependent business groups managed by specialists who value their knowledge and excel at teamwork. As a result, it’s more fun to work at Oracle these days and there’s less management conflict because decisions are based on up-to-date, shared information.”
As Oracle’s own information systems improved, Ellison was increasingly struck by how executives who had accurate data at their fingertips were able to overwhelm managers who were nominally senior to them who didn’t. He had seen Safra Catz relentlessly drive home arguments with Ray Lane by virtue of better information. “She always had the facts. I was watching how people manage, how they make decisions and how they debate during meetings. When you have someone who has facts versus someone who has none, facts, not personality, rule the day. The more we know, the more rational our decisions and the less we argue. Because once you knew that certain facts were true, the argument is over. It’s really quite extraordinary how different an organization becomes when decisions are based on facts rather than force of personality, rank, and opinion.”
Despite his enjoyment of controversy, Ellison doesn’t have a great appetite for argument, at least not in a professional context. While he’s always prepared to listen to the views of developers when they differ from his and is willing to be persuaded if he’s wrong, he doesn’t enjoy debates about business issues that depend on hunch and feel. Ellison is a self-confessed “hyperquant” who feels fully in control only when he’s processing data. Even when he’s flying a plane or racing Sayonara, he prefers to be guided by the numbers coming out of the computer systems rather than to put all his trust in what he’s feeling through the seat of his pants.
Another of the cultural changes that Ellison feels proud of is the attitude toward budgeting and expenses that he insists has now spread through the company—an approach, he says, he learned from close observation of one of Oracle’s biggest customers, GE. He says, “A few years ago when we had a budgeting session, it used to go something like this: A bunch of people walked into the room. There was a big pot of gold sitting on the table. The team that came in felt it was their job to use all their rhetorical ability and cleverness to leave the room with as much of that gold as possible in their budget. My job, on the other hand, was to hold on to as much of the gold as possible. The fight—excuse me, budgeting sessions—would go on for days and days. It took a long time for them to explain why every department needed to spend more money than it had spent the previous year.
“I said, ‘Wait a second. We all work for the same company; we’re all shareholders here. Why am I the only person in the room trying to hold on to as much of the gold as possible? If we spend less gold, we’ll make more money. That’s a good thing.’ Something I learned from working with and watching GE over the years was that GE expected its managers to come into planning sessions with budgets showing profit margins going up. Good idea. Now every manager in Oracle is required to come to planning sessions with a first slide that says ‘How we’re going to do more while spending less.’ The rest of the presentation goes into detail as to what you’re going to do this year that you didn’t do last year and how you’re able to spend less while doing it. During hard times, like the last few years, this approach to planning is essential if you want to have any hope of preserving profitability. If people are looking for an outlet for their creativity, this is the place for it. We learn from each other. We no longer have a culture in which people think that the more money they spend, the more important they are. We have a culture that finally understands: expenses are bad.”
The changes that Ellison has driven through at Oracle since 1999 are extraordinary—all the more so because almost nobody who knew him before would have dreamed that he could ever get so fascinated by every mundane detail concerning how Oracle operated or that he would have the determination to drive such profound cultural changes through a large and far-flung organization once the first quick wins had been realized. Nor is there any doubt that he has done
it by turning Oracle into an e-business laboratory, creating a virtuous circle that greatly benefits the development of Oracle’s application software. It’s a measure of his success that during the last technology industry downturn at the beginning of the 1990s Oracle was nearly sunk by poor cost control and the ensuing cash crisis, whereas in the tougher business conditions of 2001, an admittedly much bigger Oracle remained formidably profitable. Nonetheless, there are some important qualifications.
The first is that Oracle was unusually ripe for an e-business transformation. By any normal standards, it had grown extremely rapidly over a twenty-year period and, as in many high-growth companies, controlling expenses was not a priority. It had also adopted a highly decentralized model, partly to enable it to establish a global presence quickly. Because it was a technology company, spending large amounts of money on the latest IT systems was lodged deep in its DNA. However, in the era of client/server, as Ellison is fond of saying today, the more you spent, the worse it got. Thus when Ellison began his campaign to standardize, centralize, and automate everything that moved at Oracle, he quickly realized that the returns had the potential to be spectacular.
The second reservation is the risk that having created systems that have turned Oracle into a disciplined machine capable of responding very quickly to inputs from the top, Ellison may be in danger of isolating himself from the influence of colleagues and depriving the company of some of their creativity. It’s one thing to curb the “creativity” of deal-crazy salesmen; it’s quite another for a company full of very smart people to become obsessively rule-based. Ellison is clearly fascinated to see how far he can go in bringing the principles of engineering to the running of a complex business. There is a danger, however, that he may go too far.
Jeff Henley articulates some of those anxieties: “Larry has played a very healthy role in coming back in and cleaning everything up. But my worry would be that he has become very powerful and is not delegating very much at all. There’s a danger that he’s trying to control everything so tightly himself that he’ll stifle the growth of the company. It’s nothing immediate, because he’s done much more good than harm. I’m just saying that in the longer term, he’ll have to figure out a way to be a little bit more inclusive. He’ll disagree with me that we can’t grow and flourish, but the fact of the matter is that there will be limits because of the approach he’s taking right now.
“I think a lot of it goes back to our software. To create the E-Business Suite, he’s been figuring out the entire enterprise and how it can be programmed. He’s obsessed with this idea of programming every aspect of the way we all work. The idea that everything can be engineered is as intellectually exciting as hell to him, so he wants to get right down to the lowest levels of detail and work through the processes. So we’ve been making our product much better because he’s been getting so deep into it, and I’m all for that. It’s just that there’s a risk [that] Larry, who’s a very able guy, will choke off free thought and expression in the company because he’s trying to make too many of the decisions.”
Jeff Henley may be overstating the dangers. Although dominating, Ellison is usually open to ideas from people who are not so overawed that they forget to express them. But as Ellison candidly puts it, he now uses Oracle as a laboratory for product development. And in any laboratory some experiments are bound to fail. He needs colleagues who are not afraid to tell him when that happens.
The final qualification about Oracle’s story of transforming its efficiency by using its own E-Business Suite is that the claim is only partly true, because the first billion dollars of savings was pretty much in the bag before 11i was finished. The software that Oracle used to turn itself into an e-business was, in fact, a combination of modules from Release 11.0, which shipped in late 1998, and a number of customized prototype applications that were global, Internet-based systems. A key element of the E-Business Suite, however, was missing: the underlying information architecture, including the vital shared data schema that were the basis for Oracle’s claims about the tight integration of all the applications. Ellison says, “That’s absolutely right. The software that saved Oracle a billion dollars in the first year was a combination of our standard applications products plus prototype applications that had not yet been integrated into the E-Business Suite. But all the applications, products, and prototypes were global Internet applications systems. It was these systems that became the driving force in Oracle’s transformation. Once the prototype applications proved their worth, they were rewritten and integrated into what would become the E-Business Suite. As we automated every part of Oracle, just about everything went into the E-Business Suite.”
So was it a stretch for Oracle to brag in its ads that it had saved a billion dollars by using its own E-Business Suite? Not from Ellison’s point of view. For him, the process by which the suite had evolved and Oracle had been transformed into an e-business was utterly symbiotic. But in one sense, the marketing claims were misleading. The reassuring message to prospective customers was that Oracle had been successfully running the same E-Business Suite that they would buy for at least a year. That was very far from the case. Not for the first time in Oracle’s history, early guinea-pig customers would pay a price for taking too much on trust. Not, however, as great as Oracle and Ellison would pay.
* * *
1. LE writes: The term “clerks” was an intentional exaggeration in a number of internal speeches at Oracle. I was trying to make the point that ERP systems delivered very little value to senior managers because they were process automation systems, not information systems. ERP systems were pretty good at automating a business process like issuing purchase orders, but they were terrible at providing information, such as telling you if this purchase order put you over your capital budget.
2. LE writes: I thought that a major test of our new applications should be the lowering of IT expenses within months and consistently thereafter. I never bought into the notion of “investing” in Internet systems or any other kind of computer system. Internet systems should save you money very quickly, or you shouldn’t put them in.
3. LE writes: At least a country could not set its own prices without justifying the difference to the global pricing committee. We approved different prices in India and China. Everywhere else in the world our product prices are the same.
4. LE writes: For the same reason we’re trying to persuade our customers not to customize our software, because the lowest-cost way to get quality is with a standard product—a customized sales process is expensive and generally of low quality. Very few people can sell well. An engineered process is almost always—not always but almost always—better than an ad hoc process.
5. LE writes: Chuck, as usual, is right. The big deals will always be driven by our best and most senior people. I believe that means our best managers, best salespeople, best consultants, and best engineers. In a great company everybody sells—not just the salespeople.
10
READY OR NOT . . .
For Larry Ellison and Oracle, it probably didn’t get any better than this. The May 8, 2000, issue of BusinessWeek magazine carried a portrait of Ellison on its cover wearing shades and what had become his trademark black crewneck pullover. He looked good—a little too much like a Mafia heavy, but pretty good all the same. Even more satisfactory was the unequivocal headline “Oracle Is Cool Again.” The story inside by Steve Hamm wasn’t bad either. After noting the quadrupling of Oracle’s stock price in less than a year that had taken it past IBM in market cap and the widely reported fact that three weeks earlier the value of Ellison’s stake had surpassed Bill Gates’s holding in Microsoft ($52.1 billion to $51.5 billion), Hamm argued that with the shipping of the E-Business Suite, Oracle was set to be perhaps as dominant a force in the new era of Internet computing as Microsoft had been in the heyday of the PC. The piece was balanced by the attacks of rivals and the skepticism of some analysts who questioned whether Oracle could pull off its grand designs
, but the implication was clear: the future was Ellison’s to lose. Over the next twelve months, he came pretty close to doing exactly that.
In 1999, it had become accepted as fact that Oracle was one of four companies whose products were the foundation stones of the Internet and e-business—the others being the powerful servers made by Sun Microsystems, the massive information storage systems that came from EMC, and Cisco Systems’ ubiquitous routers. No serious dot-com company would look further than these four for the essential IT infrastructure it needed to do business. Amazon, eBay, E*TRADE, and Yahoo! were all Oracle customers. And the dot-coms weren’t alone: the deregulating impact of the 1996 Telecommunications Act meant a massive surge in spending by spectacularly well funded “bandwidth barons” and new “challenger” telcos, such as Qwest, and an equally determined response from incumbents determined to match their technology and conquer new territories to offset the effects of the intensifying competition in their traditional markets. Fueled by the prevailing Internet fever, never before in history had there been such an appetite by businesses to buy the latest in computing technology, and Oracle was one of the leading beneficiaries. Instead of the evangelical “The Internet Changes Everything,” the company’s ubiquitously displayed promotional message was now the much more assertive “Oracle Software Powers the Internet.”
Softwar Page 25