Softwar

by Matthew Symonds

  It’s true that the tech downturn isn’t all bad news for Oracle. The likes of the once fashionable Ariba and Commerce One, which had been carried to extraordinary market caps by the Internet tsunami, were now in danger of being left high and dry. But Siebel Systems was a much tougher proposition. Founded and run with iron discipline by the former Oracle salesman Tom Siebel, the company was still outselling Oracle in CRM by a factor of four to one. Revenues in 2001 broke the $2 billion mark accompanied by net earnings of $255 million. Siebel also had a war chest of well over $1 billion—useful rainy-day money to ride out a prolonged tech slump. I first met Siebel in the summer of 1999 at his office, which cheekily stands just three miles north of Oracle’s on the opposite side of Highway 101. A small, intense man with a harsh, gravelly voice, he lost little time in launching a bitter attack on Ellison and what he called his “pathological lying.” When he heard I was having lunch with Ellison later that day, he immediately scrawled on a piece of paper “The six lies of Larry Ellison” and said, “Give him that.” Which, to Ellison’s vast amusement, I did. Siebel’s main point about Oracle was that however much it was prepared to throw at building a competitive CRM package, he would believe it when he saw it: “This isn’t simply something you can throw a thousand software programmers at. It’s something that somebody in the organization has to know something about.”

  In much the same way that Ellison uses Microsoft and IBM as the points to Oracle’s counterpoint, Siebel uses Ellison and Oracle to define what Siebel isn’t. He says, “It is critically important to me that Siebel Systems never develops yet another pathological [a favorite Siebel word] Silicon Valley corporate culture. Too many companies here have an arrogant self-image. Their attitude is ‘We are geniuses and visionaries who make insanely great technology.’ And there’s the cult of the software CEO, where it’s all about ‘me, me, me’ and my transitory fulfillments—my parking lot filled with Ferraris, my crushing the competitors, my money, articles with pictures of me in my fighter jet or being escorted around by PR assistants.”6 Who could he be talking about?

  Although Ellison patronizingly refers to him as “Tommy,” he has a grudging admiration for what Siebel has achieved in defining and dominating a market (much as Oracle did in its early days), but he doesn’t have much respect for Siebel’s technology. “Siebel Systems is all about sales automation. They defined the category and made the market. They were the first mover. They built a ruthless, big-deal-focused sales force. But they have a couple of problems. Despite what Tom says, their software’s still client/server and their Internet version isn’t due out until the fall. Two years ago the Siebel Web site claimed their software was architected for the Internet—but it wasn’t. Now they say that their Internet version is on its way. So I guess they moved to the Internet two years ago and it was so much fun, they’re going to do it again. So, who’s lying? Tom and Craig Conway were both key lieutenants in Gary Kennedy’s sales team. Were they hard chargers? Jesus, were they ever. They’re both killer salesmen and hard-driving sales managers. That’s what made them successful. But they’re not choirboys. And now that they’re both CEOs they’ll live and die as their quarterly results rise and fall. Tommy talked a lot about the ‘pathology’ of Oracle. And when Tommy talked people listened, because he was delivering the goods quarter after quarter. But that’s all about to end. He really struggled to make his last quarter. He pumped it up with software swaps and a big deal with IBM. He can’t do that two quarters in a row. Tommy’s going to crash.”7 Siebel was indeed finding the going harder, but Ellison’s predictions about Siebel hitting a wall had been wrong before. Two years earlier he had forecast that by the middle of 2000 Oracle would have knocked Siebel off its perch.

  Part of Oracle’s problem was that it had taken the company so long to build a product with enough features to match Siebel’s even remotely. And whereas Siebel had gone out and bought companies to fill feature gaps, Ellison had insisted on making everything at home to ensure the architectural integrity of the integrated suite. As usual, Ellison put a higher priority on the benefits that derive from engineering elegance than the crude business functionality that Siebel’s hotshot sales force expertly leveraged with customers. However, a new report from the Gartner Group had found that while Siebel remained well ahead in terms of customer wins, in the last year Oracle had emerged as its only significant challenger, putting distance between itself and the rest of the chasing pack.

  Ellison says, “We don’t have every feature that Siebel has with all their different CRM products. But our CRM front-office systems are tightly integrated with our ERP back-office systems. Our CRM systems and ERP system share a common database, so it’s easy to find the information you’re looking for. All our applications are architected for the Internet, so it’s quick and easy to roll them out globally. We have enough advantages to win our fair share of CRM deals. Siebel built their feature-rich portfolio of CRM products by acquiring lots of different CRM companies: Scopus and about a dozen others. Then they rebranded all these separate products as the ‘Siebel CRM Suite.’ But it’s not a true suite. The products are barely integrated, and they don’t share a common database. Writing checks is easy. Writing software is hard.

  “Could we have caught up with Siebel any faster? Sure. All we had to do was buy a bunch of companies like they did. But we wanted to stick to our product strategy of an integrated suite built on top of a common database. The only way to get one of those is to build it from scratch. You can’t go out and buy it because it doesn’t exist. We’re building software about as fast as we can. We have nearly four thousand applications developers working on the suite, about nine hundred of them in CRM. [According to Mark Barrenechea, the true CRM figure is nearer to fifteen hundred, up from ten when he arrived at Oracle four years ago—Ellison still doesn’t know how many people work for him!] They may have built the Empire State Building in nine months, but the E-Business Suite has already taken us several years, and we’re still not done. Complex business software takes a long, long time to build. Sometimes you have to build it two or three times before you get it right. Mark Barrenechea still has a lot of work left to do. Mark worked for Scopus when Siebel bought them but came to work at Oracle rather than going to Siebel—something to do with having to wear a tie while programming.” [Tom Siebel insists on suits and ties as a mark of ‘respect’ to customers and as another way to differentiate the company from the Silicon Valley norm.] As Barrenechea tells it, “I remember going over to Larry’s house in Atherton. This was my first experience of Larry—he shows up in shorts and a pink tank top, holding a glass of carrot juice. And I’m thinking, this is my kind of guy.”8

  I ask Ellison if he sees Siebel in the same category as niche players like the Internet procurement specialists Ariba and Commerce One. “Yes, Siebel is a best-of-breed specialty player,” he says, “I’d put Siebel and i2 into the same category. Why? Their stuff is too expensive to install, integrate, and operate. I’d guess that more than half the software these guys sell sits on the shelf and never gets used—shelfware. Companies still buy from Siebel because they’re the CRM leader and people think it’s safe to go with the leader. But i2 was the supply-chain leader, and look what’s happened to them. Oracle and SAP are rapidly getting i2 out of our EPR accounts. We’ll get Siebel out too. Every year Siebel and i2 will come under more and more pressure. I don’t see how they can survive.”

  The idea of speeding up that market consolidation through acquisition doesn’t much appeal to Ellison, even though the battered and reeling niche players would now go for peanuts. One possible exception was Clarify, the call center specialist acquired by telecom equipment maker Nortel a couple of years ago. He says, “We’re cheap, too cheap, probably. We’ve never paid more than $125 million for an acquisition. Clarify’s in that range these days now that SAP stopped reselling it. Selling Clarify and Commerce One was SAP’s brief experiment with best-of-breed integration. They weren’t any better at it than we were. At a recent co
nference in Florence, I was on the same speakers’ panel as Hasso Plattner [SAP’s co-CEO]. Hasso was explaining about how hard it is to integrate with the third-party products they’re reselling. Tell me about it! SAP’s in the process of declaring their integration experiment a failure and going back to building and selling their own suite of applications. So will we buy Clarify? We’ll look at it. We’ll analyze it. We’ll talk about it. But we won’t buy it, because we can’t integrate it.”

  One deal that Ellison wishes he had done was to buy PeopleSoft in 2000, when its stock was on the floor and Oracle’s was riding very high. “PeopleSoft was amazingly cheap. We could have bought them for two or three percent of our [outstanding common] stock. I wanted to do it at first, but I gradually became convinced the timing wasn’t right. Most of our executives thought doing our first big acquisition at that time was just too risky. We were in the midst of a massive reengineering of our own company that consumed a hundred and ten percent of our management capacity. No available management, no experience in big acquisitions—sounds pretty risky to me. But something else bothered me even more: PeopleSoft pretends to be a suite company, but they’re not. Sixty-five percent of their business is HR systems. They’re really a best-of-breed HR specialist, and you know the fate of best-of-breed specialists. I think their HR product will gradually get replaced by Oracle HR and SAP HR. Still, PeopleSoft was so damn cheap we probably should have bought them anyway. It was close, but we probably made the wrong call.” They certainly did. Although, in a predictable echo of Tom Siebel’s remark about Oracle, Ellison says that he never “sees” PeopleSoft competing for big accounts, there’s evidence in both the numbers and the analysts’ reports that the firm from Pleasanton is making something of a comeback with its version of the E-Business Suite, PeopleSoft 8. Although still in recovery, PeopleSoft is a much-loved company.

  Ellison accepts that one rival that is definitely not fading away yet is SAP. “SAP made a couple of exceptionally good decisions that resulted in them becoming the number one applications company in the world. First and foremost, they invented the applications suite business when they built the first integrated ERP system. Being first got them lots of big customers all over the world. Even when the going gets tough—it’s been difficult for them to get new customers in the U.S. over the last couple of years—they can always sell into their installed base and into the hyperloyal German market. Nobody can compete with SAP in the German market—they own it. Second, SAP has executed its strategy pretty well because it’s an engineering company at its core. In that respect Oracle and SAP have a lot in common: we’re both engineering-centric companies, and we both have software engineers as CEOs. Most other application software companies are run by salesmen. The Oracle and SAP engineering teams have great respect for each other and get along together quite well. Their engineers love our database. In fact, when we launch Oracle 9i in June, they’ll be there up on the platform with us9 [they weren’t, as it turned out], which is really pretty incredible considering the rivalry in applications. But their customers desperately need the speed and reliability of Oracle 9i Real Application Clusters [RAC]. For all the things SAP has done right, they still face one overwhelming challenge: their system is very, very inflexible, and that makes it extremely expensive to install and operate. The big consulting firms charge a fortune installing, integrating, and customizing the SAP system. The cost overruns on big SAP implementations are legendary. In this new climate of constrained IT budgets, it’s going to get harder and harder for them to sell to new customers. If people start asking for a price guarantee before buying or expanding an SAP system, then they’re in big trouble. SAP’s best hope is that people don’t ask.”

  • • •

  Hardly a month goes by without Ellison’s hosting at least one or two CEO roundtables. While he has frequent meetings with Oracle’s biggest customers and increasingly makes the vital calls that help to close deals, the roundtable format allows both efficient use of Ellison’s time and the opportunity to talk more widely about the issues that concern Oracle’s customers. Today’s roundtable is being held at his house in San Francisco. The Pacific Heights house couldn’t be more different from the restrained simplicity of his Japanese-style villa at Atherton. Four stories high, the house is famous for its extraordinary panorama of the bay, sweeping from the Golden Gate Bridge to Alcatraz. It’s also well known for being fitted out like a James Bond villain’s hideaway with touch screens directing the industrial-strength computers that control every function inside the house. It’s gorgeous but so minimalist that it’s not a cozy place to live in. It is, however, a spectacular venue for corporate entertaining. A few of the twelve expected guests are relatively local, such as Dick Kovacevich, CEO of Wells Fargo (and one of Oracle’s bankers), and David Lawrence, the chairman of Oakland-based Kaiser Foundation Health Plan, one of America’s biggest medical insurers. But others, such as Jim Courter of IDT, a telco in Newark, New Jersey, and Ron Orr, who runs the air force’s logistics operation from Washington, D.C., have flown across the country to attend. And while an invitation to lunch from Ellison is on its own probably enough to get the attention of most CEOs, the fact that the lunch is being held in a house that you will want to tell your friends about is an added lure.

  After the guests have had an opportunity to admire the view from the first-floor drawing room before the fog closes in, they are led down to the dining room and seated around the huge, rectangular white table. Oddly, there seems to be no seating plan. Other top Oracle executives, including Safra Catz, Sandy Sanderson (the head of Oracle Consulting), and Mark Jarvis, just spread themselves around, while a handful of the guests politely jockey for a seat near Ellison. The meal, served by a team of several black-uniformed waiters, is elegant but simple. Despite the opulently over-the-top surroundings, the atmosphere is serious and professional.

  At least until Ellison starts talking. With this group, he’s confident and ready to entertain. It’s a variation on the usual theme: what he’s learned from his own efforts to turn Oracle into an e-business and why it’s a process in which CEOs must become directly involved; the profound impact of the Internet; the failure of IT to deliver on past promises because it is an industry founded on a model that both celebrates and distributes complexity; and, finally, the extraordinary benefits to be had, in terms of efficiency and knowing what’s going on in your business, by adopting Oracle’s approach of implementing a fully integrated suite of applications on top of a single consolidated database. But the anecdotes of the chaos at Oracle before he saw the light are well honed and funny. What’s more, they strike an immediate chord with his guests, as he knows they will. The confessional tone is calculated to get others to tell blood-curdling horror stories of their own. And then for Ellison to point them in the direction of salvation.

  There is, however, a potential danger in being funny about the nightmares that the IT industry has visited on its wretched customers. In his own imagination, Ellison had long since moved on to a simpler and more rational kind of computing that banished armies of expensive consultants and the associated problems of tying together software packages from a multitude of vendors. Oracle, Ellison boasted, could now, uniquely, offer computing solutions that, for the first time in IT history were guaranteed—that’s right, guaranteed—to provide a spectacular and speedy return on the customer’s investment. However, even at Oracle, with its much-trumpeted $1 billion of savings and the boast of being able to extract as much again this year (recently slightly moderated because of “the economy”), Ellison was, according to his own CFO, Jeff Henley, running slightly ahead of reality. But for customers, including some of the ones around the table on Broadway, what might be galling was the fact that until very recently, Oracle had enthusiastically been selling the very approach, namely customized best-of-breed solutions, that Ellison was now describing as madness.

  By the time Ellison stops speaking (after a good forty-five minutes), one of the guests, George Buckley, an English
expat who runs the Brunswick Corporation, a medium-sized conglomerate that operates boating and leisure-focused businesses such as Mercury Marine, is seething to say something. He has written a number of points down on his pad. Looking intense and white-faced, he begins, “Larry, we’ve spent $250 million on systems in the last few years, and we’ve got jack shit. I feel like we were led down the primrose path by the IT guys. It was always jam tomorrow. We have all the best stuff in the candy store—PeopleSoft, i2, Oracle—it’s a real hodgepodge of systems.” Ellison has a “There you are!” kind of smirk on his face. He says, “That’s the absolute archetype of what not to do.” Buckley looks furiously across the table at him: “Well, I’ve got to tell you that it was Oracle consultants who, when pressed by my IT people on a perceived deficiency in your software, would say that maybe PeopleSoft or i2 might be better. Well, I’m not going to throw in another $200 million to fix it now.” IDT’s Jim Courter chips in, saying, “We were fed all kinds of promises that it will all work, and it doesn’t.” Buckley adds, “We’re in so deep now that spending money is not the way. We’re spending $100 million a year now. So how do we get out of this mess?”10

  It will be interesting to see how Ellison gets out of the hole he’s dug for himself. What he’d like to do is disown the advice that Oracle’s own consultants were peddling only a couple of years previously, when Oracle had some major gaps in its applications suite and couldn’t afford to be so dismissive about best of breed if it was to compete with the more complete product from SAP. But with Sandy Sanderson, who has run consulting since the departure of Robert Shaw in 1998, a few feet away from him, doing so might be embarrassing. Instead he tells Buckley that he should actually be cutting his IT budget. He suggests that Buckley stop the work on some of his existing systems integration projects and swap a portion of the budget out to install 11i in one part of the business. Once it has proved itself, Buckley can start rolling it out across the rest of his company. By firing some of the consultants wrestling with his existing software and starting again, Buckley will be able to save some of the money he’s currently spending. Says Ellison slightly smugly, “SI is the gift that keeps on giving.”11