Softwar

by Matthew Symonds

  The reaction in the room is silence. Surely it is too good to be true. The devil would be in the details. James McCann of 1-800-Flowers nervously asks whether this is “a new pricing paradigm.” “Totally,” says Ellison, and just before he has to start explaining what he meant, Art Weinbach, the tech-savvy CEO of ADP, comes to the rescue: “This concept works. It’s a strategy that would work over a long period of time. The pricing is fascinating.”

  Ellison relaxes. “It just kind of slowly dawned on me. I have complete confidence in the suite approach. I know it saves money. I know it provides better information. But how do I convince people? Simple: move the risk from the customer to us; guarantee the price, and guarantee the savings. IBM can’t do that. IBM’s telling them to go best-of-breed, because IBM makes a bundle gluing it all together. But the IBM approach costs a fortune. They can never compete with us on price. Everything they do is one-off custom. So with our new E-Business Suite outsourcing offer, we take the risk and we share in the reward. I’m very confident that we’ll make plenty of money, because we can lower IT costs at a much faster rate than five percent per year. We cut our own IT costs by more than half when we put in the E-Business Suite.”

  Martha Stewart, whose fast-growing company, Martha Stewart Living Omnimedia, is upgrading from a Microsoft to an Oracle technology platform, is impressed. She says to Ellison flirtatiously, “Do I call you? Can I have your home phone number?” The innuendo, however, seems to pass Ellison by. “Sure,” he says. “Find out how much you’re paying for IT, and then think about five percent savings each year and no capital costs. If twenty percent less after five years is interesting—let’s talk.”

  Elsewhere in New York, Mayor Giuliani is speaking to a packed news conference about the anthrax attacks on the television companies and calming panicky New Yorkers. Selling software doesn’t seem quite as important as it once did.

  * * *

  1. LE writes: I like Donald Rumsfeld, but not as much as Melanie does. She has a Donald Rumsfeld screen saver on her Macintosh laptop.

  2. LE writes: I submitted my oped piece to The New York Times, but they rejected it. I suppose they thought my position on ID cards was a bit too far to the right. However, shortly after my article appeared in the Journal, the Times ran a piece by the Harvard constitutional law professor Alan Dershowitz that called for a mandatory national ID card. All I wanted was a national standard for our existing voluntary government ID cards. So for the first time, I found myself standing slightly to the left of Alan Dershowitz on a civil liberties issue. Things had changed. September 11 convinced Alan and a lot of other people that we live in a very dangerous world.

  3. LE writes: During a long negotiation with HealthSouth, we became more and more uncomfortable with their senior management team. In the end, we decided that they were not people we wanted to work with, so we walked away from the deal. Eighteen months later, when Richard Scrushy was charged by the SEC with what it described as “massive accounting fraud,” John Wookey sent me an e-mail saying how thankful he was that we were watching this from the sidelines.

  4. LE writes: I don’t believe that American citizens should be required to carry digital ID cards; I do believe that all foreign visitors should be issued a digital ID card before they can get into our country.

  5. LE writes: Of course there is some information in certain intelligence databases that is so sensitive that there should be no automated sharing system. These databases are on computers located in underground vaults, with two marine guards stationed outside the vault door. If you’re trying to log on to one of these databases and you don’t “remember” your password, the marine guards shoot you.

  24

  THE GOLDEN NUGGET

  Amsterdam

  January 2002

  About fifteen thousand of the Oracle faithful have gathered at the RAI Center in Amsterdam for the European leg of AppsWorld, the annual jamboree to evangelize Oracle’s application products to its customers, its partners, and the technology media. Although Europe is SAP’s backyard, Oracle is the strong number two in most of the major European markets other than Germany, where the firm from Walldorf has things pretty well locked up.

  The big announcement in product terms at the RAI is something that Ellison calls Daily Business Intelligence. His big complaint with business applications is that although they have become quite good at automating key processes, it’s still extremely difficult to get any useful information about your business out of them. As a result, on top of the huge sums of money big companies spend on applications, they also have to shell out for separate data-mining tools and data warehousing. Unfortunately, these so-called business intelligence products have some serious flaws: not only are they expensive, but they’re difficult to use and the information they throw up is a snapshot of the past, not of the present. That’s okay if you want to know everything about your sales of, say, driveshafts in Brazil over the last three years. But it’s next to useless if you want important metrics at your fingertips about how every aspect of the business is operating in real time.

  Ellison has believed for a while that one of the overwhelming advantages of a complete and integrated suite of applications sharing one set of data schema and running on top of a single database is that it would make it possible to create an “executive dashboard” that constantly updates every key performance indicator. But what in theory should be perfectly doable, has turned out to be very hard. Nonetheless, Ellison still wants to demonstrate a working product here in Amsterdam that, he’s assured me, “will blow your mind.”

  There are a couple of other themes at Amsterdam of at least equal importance, both aimed at repairing some of the largely self-inflicted damage of the previous eighteen months. The first is to stress that although “completeness” is still a unique selling point of the E-Business Suite, what’s really vital is the underlying information architecture. Markedly changing its tune from the ill-tempered AppsWorld in Paris a year ago, Oracle is now making a point of saying that legacy applications can successfully be integrated with its suite as long as they can be hooked into the shared architecture. The same goes for “extension” applications made by specialist third-party software vendors, such as those in the health care business. By publishing its data schema and application program interfaces, Oracle is trying to show that it has created an open, flexible platform. Ellison says, “We’ve always been able to connect the E-Business Suite to your existing legacy applications or to other vendors’ applications. Now we use the very latest technologies to make systems integration as rapid and painless as possible. But understand, even though the tools have improved, integrating several separate software systems remains a complex and expensive process.”

  The second message—and maybe the most vital one—is that the suite, whatever users may have heard before, is now fully stable and delivering real-world efficiency gains and huge cost savings to an ever-increasing number of customers. The trouble is that for the best part of a year, Oracle has been saying with each new 11i release (and there’s been one just about every three months) that pretty much all the issues have been dealt with. Even if what’s being said is now completely true, there’s a big credibility problem, one that Oracle itself tacitly acknowledges. Since the summer, Ellison has been saying that the only way to sell the suite going forward will be with a wide range of customer references. But Oracle has been hit by the double whammy of still sluggish corporate technology spending and the “wait-and-see”1 approach within much of the installed base, which the press reports about the difficulties encountered by enthusiastic early adopters have contributed to.

  This means that those references that Oracle does have—and there are now well over a thousand mostly satisfied E-Business Suite users—are being milked in Amsterdam for all their worth. For several months, the reference poster child has been POSCO, the Korean steel firm that is also the world’s biggest and most efficient. Although the “go-live” on 11i was back in July, it’s still one of the big
gest “big-bang” installations to date. But what’s really making the Oracle execs purr is that just in time for Amsterdam, POSCO’s CIO has told them that the efficiency gains have been so spectacular that his team is doubling its already aggressive forecast of $4 billion savings over ten years—in other words, about $600 million in savings a year directly attributable to the E-Business Suite. You might never have heard of POSCO, but Oracle is doing its damnedest to rectify that.2

  Much of Ron Wohl’s presentation is devoted to the superiority of Oracle’s information architecture, and as the point man on the POSCO implementation, he also has plenty to say about that. It’s a confident and convincing performance from somebody who lacks a natural stage presence. Backstage after his presentation, I ask Wohl about the twelve-month struggle to bring 11i to maturity after it became clear that Oracle had launched a product that, though conceptually brilliant, just wasn’t ready for prime time. He sucks in his breath and begins, “Looking back at the release, you always have the wisdom of hindsight. Fundamentally, we tried to do something that I think was much larger in scope than even we realized and was much larger than the scope of any software we had released beforehand. In retrospect, we didn’t do a good enough job up front—our QA [quality assurance] procedures didn’t work because they had outgrown our scale. In retrospect, I would have liked to have done two things differently: we should have changed a number of processes within development to deal with the scale of the release, which we have now done; and two, knowing what we know now, we would have been wiser to withhold the release a little bit. We would have been better off as a company if we had deferred the release by three or four months.”

  Wohl was ready to concede that the heart of the problem was the gamble he and Ellison had taken with the order management system. “It was one of the last parts of the 11i software to be completed, and as a result we had more stability issues in order management than in other parts of the ERP code. But the stability improved at a very rapid pace. You could see it in the bug numbers, and you could see it in the customer satisfaction levels. Customers who had started on a later version—11.5.3 and above—were extremely happy. If you talked to customers who started on one of the early versions, they weren’t happy, and they had a right to be upset with us. But we made it up to them by really going out of our way to help them through their difficulties.

  “Take a look at where we are now, about a year and a half after the release of 11i. We have more than one thousand customers live, achieving billions of dollars of benefit. You have to compare the problems to the accomplishments, and the accomplishments, by any measure, have been profound. In this business, there’s always a three- to six-month lag between perception and reality. We benefited from that in the beginning, but now it’s hurting us. If we talk in three or four months, the reputation of our E-Business Suite will be much improved. Reputations change very quickly in our business. Right now, the reality is that our software is extraordinarily good, our references are extraordinarily good, and the demos are fabulous. Coming out of this, nobody will be able to stop us, I honestly believe that. No one can match our software. Before, there was a lot of doubt, and frankly some of the ongoing skepticism is because we didn’t do a good enough job initially. But now we’ve got it right, and we’ve proved it works. No one can argue with our live-customer stories; those are real.”

  Wohl’s other point was the same one that George Roberts, Oracle’s top salesman, had made to me some months earlier: the suite concept was winning acceptance in the market. Wohl says, “Some people thought we were crazy; they said that no one company can do all that. But look at what’s happened to all the best-of-breed vendors, except maybe Siebel, they’ve dived and hit the dust.3 They’ll stick around for a while, but they are just not going to grow, they are not going to evolve and succeed, because the integration costs are too high. Also, notice that SAP is now calling itself an integrated business suite company, and so is PeopleSoft. They’ve both copied our marketing message, but they can’t copy our software. The integrated application suite marketing message is what the applications vendors with a future have in common. But we’re the only suite with an information architecture based on this single-data model in one database, and our competitors know it. The value of the simplicity . . . can’t be overestimated. This is what will allow us to beat SAP and put PeopleSoft to bed.”

  Although in his quiet way, Wohl is as intellectually arrogant as anyone I’ve ever met, he is not normally given to making such bombastic statements. He has the air of somebody who has been through a great deal and now feels vindicated. It isn’t just the experience of the previous torrid twelve months that has created this battle-hardened certainty, but the last decade, when Wohl was held up for ridicule within Oracle (especially by Ray Lane) and time and time again depended on Ellison’s (to many people, quite baffling) faith in his ability, for his survival. He has the air of a man whose moment has come.4

  If Ron Wohl had the scent of ultimate victory in his nostrils, I expected more of the same from the naturally cocky Mark Barrenechea. But he turns out to be far from happy. He starts by saying that in the last year and a half “we’ve done some brand damage.” He puts the blame squarely on the order management system (and by implication on Ron Wohl): “We’ve been in order management for fifteen years, and we still didn’t get it right. We’ve had quality issues with many of the underlying ERP components [Wohl’s side of the suite], and in many cases we didn’t even do damage limitation well.” He even expresses some doubts as to whether the latest release to ship—11i 5.5—has dealt with all the problems. “Some of the early indicators suggest that many of the issues are behind us, but if history repeats itself it’s still going to take a lot to flush them out.”

  When I ask Barrenechea how the CRM modules, the part of the suite that he’s responsible for, are performing, he vents his frustration. He claims that because most of Oracle’s fourteen thousand existing applications customers inevitably begin their 11i implementation by upgrading their ERP systems, the quality problems they’ve hit have often meant that the money runs out before they even get to installing “his” CRM software. “By no means has our CRM execution faltered, not at all. But in many cases we don’t even get to the starting gate because all the implementation dollars have already been used up. It may take three years now before they get to try them. It makes me very impatient.”5

  Barrenechea accuses Wohl’s team not only of mishandling the major rewrite of the crucial order management module, but also of not writing all its software to conform to the new Trading Community Architecture (TCA), or “customer master,” that is meant to be the basis of the 11i platform. He says, “We’ve all been told to standardize on TCA, and yet we find the other organization somehow missed it along the way. So we had quite a bit of cobbling to do to get the new customer master and the new order management working across CRM. It kind of comes down to bad management in the other organization.”

  This is a pretty serious accusation from one of Oracle’s most senior executives. I point out to Barrenechea that Ellison argues that the most important single contribution he has made since becoming involved in apps development in 1998 is to get the various parts of the development organization working together cooperatively. Barrenechea laughs and nearly chokes on the cake he’s eating. “Oh, sure—it’s a lot better than it used to be. At least we’re not spitting at each other anymore. . . . Look, Larry has a brilliance, and he’s certainly a virtuoso when it comes to understanding organizational structures. But he likes seeing two organizations compete so facts arise. Part of the difficulty for me in understanding this is that it translates into nonconsolidation of responsibility. Most companies would consolidate responsibility onto one sales guy, one consulting guy, one chief product developer. Right? This is either one of the most wonderful characteristics of Oracle, or it’s one of the greatest flaws. And I can’t work it out yet. It’s so Larry.”

  I can see where this is leading, and sure enough Barrenech
ea says, “We need one applications organization.6 And we need one product development organization that, over time, does not consolidate to Larry. And we have to envisage an Oracle, someday, without Larry. And be working toward that.”

  As if realizing that he may be overdoing the criticism, Barrenechea returns to his usual optimism. Despite all the problems, which come from doing something that is intrinsically very hard, he is convinced that Oracle’s approach is “absolutely right” and that no one else can match what’s been achieved. He says that the release that’s about to ship—11i 5.6—is working really well within Oracle, smoothly integrating both the ERP and CRM sides of the suite, an indication that it’s now ready for very wide scale deployment.

  He takes particular heart from what looks like a major win over Siebel with AT&T’s business services division. The deal has had to withstand a raft of executive departures at the crisis-torn telco as well as the departure of Jay Nussbaum on the Oracle sales side. What has won the deal for Oracle, says Barrenechea, is its ability to automate an entire business flow (from lead to opportunity to configured quote to order to contract to billing) involving twenty thousand sales representatives, thus reducing the process cycle from five months to thirty days. Coming on top of the BellSouth deal, it will make Oracle’s CRM product very hard to beat in the telco market. He’s also excited about winning business from Dell Computer, which has decided to standardize all its applications—a mix of bespoke apps and Oracle’s products—on Oracle’s information architecture. Given that Dell has few peers as a business process innovator, it’s a major validation of Oracle’s approach.

  Mark Jarvis, the Englishman who spears Oracle’s always aggressive marketing effort, readily admitted that after “eighteen months of Hell” he hoped Amsterdam would mark a turning point. He said, “I’d never previously thought that I might be fired. But in the last six months I have thought about it several times. I’m used to getting occasional notes from Larry, but I’d never had one like the one I got from him a few months ago, saying that the marketing is the worst it has ever been and our press coverage is the worst it has been in the history of the company.” (I asked Ellison later whether Jarvis was in danger of getting the sack. He said, far from it. He’d “weep” if Jarvis ever left, but Jarvis needed better-quality managers at the level just below him, especially for dealing with the media.)7