Softwar

by Matthew Symonds

  Second, Ellison identified an easy-to-implement project that would provide “cheap thrills” and immediate proof that the new Internet systems actually worked: global e-mail. Instead of those ninety-seven separate e-mail systems, Oracle would run on a single Internet e-mail system running exclusively on two computers at its main data center in California (plus standby servers in the Colorado-based backup data center). Ellison says, “It was a stunning success. We used our global intranet and our database e-mail system, now called Collaboration Suite, to link everyone in the company together. The new global e-mail system cost one tenth as much as the ninety-seven local e-mail systems it replaced. And it was faster, more reliable, and more secure. Our users were happy because they could keep using the user interface they were familiar with: Microsoft Outlook.” One senior executive in Tokyo, betraying a limited faith in the Internet technologies that Oracle was selling, couldn’t see how an e-mail system in California could be faster than one in Japan. “Something to do with the speed of light,” muttered Ellison to him. “Please don’t act so surprised in front of customers.”

  Next up for globalization were marketing and sales. And in the case of marketing, it was not just its computer systems but the organization itself that was unified. Unlike with IT, marketing people were left in their countries, but from now on they all reported to Mark Jarvis rather than to their country managers. But it was the same deal as for IT: if you wanted to keep your own marketing organization, you paid for it. Ellison was trying to instill a new culture in which the status of senior managers was based not on the amount of money they spent but solely on how much profit they earned. But as Ellison had anticipated, there were other benefits from globalizing marketing: “The Internet made all our marketing programs visible in every country in which we did business. The programs that worked became global. The programs that didn’t were killed. Duplication of effort was minimized. Prices set by our pricing committee were published on our global Web site, where our customers could see them. That made it impossible for every country to set its own prices.3 We now have less than ten people working on product pricing. That’s down from two hundred. So duplication of effort ended, delays eliminated, and our costs greatly reduced. The pricing bureaucracy was gone. I could always make policy, but now, for the first time, I could enforce it.”

  By April 1999, Ellison felt sufficiently confident of the progress he was making to start talking about Oracle’s “e-business transformation,” claiming to the board that he believed it was possible to take out a billion dollars of annual recurring costs in just the next year. Two months later, at a headquarters press conference, he went public with the target. He was desperately hoping that before the end of the year, Release 11i, the E-Business Suite, the most ambitious software package in the history of enterprise computing, would be ready to ship, and he wanted the world to believe that it was thanks to using its own software that Oracle was saving unimaginable amounts of money. However, to reach the billion-dollar target and get beyond it, Ellison, Gary Bloom, Gary Roberts, and the other executives leading the e-business charge were now going to have to do something much more difficult than globalizing HR or accounting.

  Despite the degree of professionalism that Ray Lane had, at least initially, brought to U.S. sales after the near catastrophe of 1991, Ellison’s distrust of Oracle’s own sales force and the potential for trouble that stemmed from its wheeler-dealer ways was largely undiminished. Just as the Internet could help restore centralized control over Oracle’s Tower of Babel–like marketing operation, Ellison believed that an Internet-based sales system could not only make sales more efficient, but just as crucially, it could standardize much of it. What Ellison wanted was a uniform process that would minutely choreograph every single step of the sales process.

  Ideally, an increasing number of sales would be made through the Oracle Web store without any human intervention. When sales prospects visited Oracle’s Web site, the information collected would automatically steer them toward information about products or services that might interest them. Follow-up e-mails would be automatically generated to keep up interest and to discover whether a telephone call or a face-to-face visit from a salesperson was needed to close the sale. Where appropriate, some leads could be instantly shared with approved consulting and hardware distribution partners. At least in theory, the prices posted on the Web would be the standard prices that all but the very biggest customers would pay, thus avoiding the hugely time-consuming (for all concerned) ritual of protracted negotiation on every deal. A by-product of this transparent sales process would be far more accurate sales forecasting.

  But Ellison wanted to go much further than that; he wanted to use the technology both to help change the entire sales culture at Oracle and to make each salesperson vastly more effective. He says, “I wanted to get the ‘creativity’ out of the sales process. If you want to be creative, go write a novel. I want to ‘engineer’ the sales process. I don’t want the field people spending all their time wheeling and dealing on price. The primary function of our salespeople is to communicate and quantify the business benefits of our products. There should be a carefully engineered step-by-step process for most common sales situations—like selling our database against IBM’s. First, identify the decision maker and enter their e-mail address into our sales automation system. Second, send them a set of key customer references—case studies showing how much customers saved by converting from IBM DB2 to Oracle—showing how they got better performance and ran on lower-cost hardware. Third, send them a proposal quantifying their cost saving in hardware, software, and labor if they pick Oracle rather than DB2. Finally, send them a standard contract and a price quote. If we have an engineered, proven sales process, we will sell more software. The purpose of new sales automation systems is to allow standard sales processes to be defined and implemented across the sales force. That’s real sales automation. Most so-called sales automation systems simply automate opportunity management and sales forecasting. If you are going to spend a lot of money on sales automation, make certain you are automating selling—not just forecasting.”4

  Ellison adds, “An engineered sales process can also be effortlessly updated. If the sales pitch we want to use against IBM changes, as it does from time to time because of some cool new feature we put into Oracle or because we have a great new reference, I want every salesman to change the way they sell against IBM. Press a button, and everyone’s PowerPoint presentation becomes the latest version of the pitch. Press a button, and great new references are automatically sent to our sales prospects. Press a button, and the new demo is online.”

  Ellison recognizes that applications can’t be sold the same way as database technology. But instead of sending four or five people to, say, Slovakia to demonstrate those applications they know something about and think the customer might be interested in, why not use cobrowsing on the Internet to run the demos and presentations? That way, the best people are always conducting the demos, the customer is seeing the latest version of the software, and money is saved on travel expenses. It also happens to be a pretty good way of showing customers the difference between client/server applications and apps that run over the Internet and can be accessed with a browser regardless of where the computer they’re sitting on happens to be.

  Ellison is unrepentant about the charge by rivals, such as Tom Siebel and some of the analysts, that what was once the most feared sales force in the industry is being “dumbed down” and emasculated. Many of Oracle’s top sales guys, they say, have left to join other software firms because Ellison no longer allows them the flexibility they need to do their jobs. It’s typical Larry, they say; because he’s an engineer, he thinks that everything in the business can be engineered. He laughs. “Yeah, the Oracle sales force certainly was feared—mainly by me! Seriously, I think some of our competitors have been hiring our old cowboys—the kind of guys who will only be happy as long as they have horses, lariats, and guns. But I’ve never thought that ou
r very best salespeople were these crazy deal makers. Our best salespeople are very knowledgeable and service-oriented. I love and respect those guys, and I want more of them. I want a sales force made up of the kind of people you’d want your son or daughter to marry.”

  It’s understandable that Ellison should think the way he does, but I wondered whether George Roberts, the head of U.S. sales and a former protégé of Ray Lane, would feel the same way. Did he feel that Ellison was engineering the testosterone out of selling? “Oh God, I think that’s a ridiculous assumption. I think that what Larry is trying to do is find a way to make every sales rep behave like an ‘A’-grade sales rep. I have had this discussion with Larry many times, and what I tell him is that we will know we’ve arrived as an e-business when we release the product and the next day forty thousand Oracle employees can deliver the same-value proposition around the globe. Because that’s when size becomes an advantage. Historically, as companies get bigger they become more inefficient, and this whole e-business effort is all designed to leverage Oracle’s size and make it more effective all around the world. I have seventeen thousand people. Closing that $2 million deal isn’t going to solve my problem for me, but if I can make seventeen thousand people more effective in their job every day, that has a real impact top line and bottom line. That’s a huge win.”

  Doesn’t Roberts lament the passing of the “crazy deal makers”? “Absolutely not,” he says. “We want to make sure that people understand our key differentiators against the competition, what our value proposition is, and how we then match it against customer needs. That’s it. A salesman shouldn’t want to cut deals. Okay, at the high end, the top hundred accounts, there may be some relationships and arrangements. But as you move further downstream, the more standard the business should be. A large part of our e-business effort is aimed at optimizing our distribution model in a way that leverages all the different channels a customer can buy through. That’s tremendously important in helping us get market share. As for our competitors saying that we’re dumbing down, they’re nuts. And you know what? They all want to do what we’re doing. I’ve had a dozen companies through here in the last few months—one of them was BMC, a software company—and there isn’t one that doesn’t come away saying you guys are so far ahead of anything they’ve seen even if they’ve been doing telesales for years and years.”

  Chuck Phillips of Morgan Stanley Dean Witter, who is not only the doyen of Wall Street enterprise software analysts but, unusually, is held in high regard by Ellison, takes a balanced view of Oracle’s attempt to change the way software is sold. He believes that some “dumbing down” was necessary just to make things simpler for the customer. “Buying from Oracle was like trying to buy a building rather than a piece of software. I think Larry’s simplifying it is something customers like . . . but I don’t think he ever gets rid of those high-testosterone sales guys either. Those guys are still very important for the big deals. It’s not like people wake up and decide to spend thirty million on the Web site. You need hungry guys that can sell and sell high, and I think he knows that. . . . You still want those top five hundred guys who know how to move mountains and who won’t take ‘no’ for an answer. There is still a human element to selling software.”5

  Not everyone at Oracle was as enthusiastic as George Roberts about the changes, many of which were being driven through by Ellison’s “shadow management team” of Gary Bloom and Safra Catz. According to Ellison, one group that felt particularly threatened—perhaps not surprisingly, given the amount of independence they had enjoyed under Lane—were the country managers in Europe. That may have been true of some who equated their power and influence within Oracle with the size of their budgets and the autonomy of their systems. But the major obstacle to changing the way things were done in Europe was the man whom Ellison had himself chosen to run EMEA (Europe, Middle East, and Africa), a polished ex-Digital and ex-AT&T executive with a background in services named Pier Carlo Falotti.

  After joining Oracle in 1996, Falotti had built a substantial European headquarters in Geneva. Unfortunately, the infrastructure that Falotti had built around himself was extremely expensive to run, while contributing little to revenues. As Ellison demanded increasing cost savings, Falotti forced his managers to absorb the costs of the “services” that were being provided from Geneva. Philip Crawford, a tough salesman who ran the very profitable British business until being fired by Lane in 2000 for overaggressive customer license audits, says, “The problem with Falotti was that he said one thing and did another. I sat in meetings where blatant lies were told about costs. Blatant untruths about European and about EMEA cost structures. When Larry started to ask for lower costs, quite understandably the answer in EMEA was that the countries have to eat more of the cost of services. In the United Kingdom I got almost zero in the way of services that I actually wanted.”

  Safra Catz recalls, “It was not pleasant dealing with the senior European management team. I think some were well intentioned but didn’t know what to do. The guys had to push Ray’s line that they couldn’t change because they’d always done it in this way.” Ellison eventually fired Falotti on the last day of May 2000 by calling him on his mobile phone. The date of his dismissal meant that Falotti lost $10 million worth of options that were soon to vest. To avoid that fate, Falotti secured a note from his doctor stating that he was “ill and unable to work as of May 30 for an indefinite amount of time.” According to Swiss law, an individual can’t be sacked if he is too ill to work. Suspecting a scam, Oracle promptly slapped in a lawsuit against Falotti arguing that the doctor’s note “plainly misrepresented the facts.” So far, the legal decisions have gone against Falotti. His consolation is that he had already made nearly $30 million from the earlier sale of shares from an option grant.

  Another senior manager who left as a consequence of Ellison’s e-business crusade was Ray Lane’s old chum Randy Baker. Having already had education stripped from him in early 1999 because of the “dismal” 13 percent profit margin he was targeting, later in the year his leadership of Oracle’s six-thousand-customer support organization was passed to Gary Bloom. Two months later, in February 2000, the fifty-five-year-old Baker was fired. Like Falotti, Baker believes the timing of his sacking was linked to the date when options were due to vest. He sued Oracle for $18.5 million on grounds of wrongful dismissal and age discrimination. He had indicated that after seven years’ service he wanted to leave the company in August, by which time his options would have vested, but Ellison had other ideas.

  Ellison says, “Randy and Pier Carlo both felt they needed more people to do a good job. I didn’t agree. I wanted Randy to leave because his margins were dropping like a rock and he still wanted to hire more people. What’s more important to me, profitability or someone’s age? I’m actually slightly older than Randy. I wanted to automate, cut expenses, and improve profitability. That’s what I insisted on. It proved to be a very difficult time for many of our most senior managers, who were long accustomed to operating independently.” Ray Lane regards both men as “among the finest executives” he has ever worked with.

  By moving all of Oracle’s service information to the Web and everyone within customer support to a single Internet-based global system, Ellison maintains, customer service and satisfaction have increased without the need to hire any additional people. He’s especially proud of the fact that much of the support work is done out of Bangalore by Indian engineers, who earn a fraction of the wages expected in California or Europe. Problems can now be tracked around the clock, exploiting different time zones, and the best-qualified people can work on problems wherever in the world the customer may be based.

  Remarkably, within not much more than a year of Ellison’s beginning his campaign to cut costs by turning Oracle into an exemplary e-business, the targeted first billion dollars of savings—a number pretty much plucked from the air to help dramatize the project—had been secured and Oracle’s profit margin had moved from a little unde
r 20 percent to more than 30 percent. Ellison had been helped by the rapid growth in all parts of the business coming from the dot-com and telco-generated boom in technology spending, but Ellison was already talking about getting a second billion dollars in annual expense savings.

  The savings gave Ellison the marketing message he needed to sell Release 11i, which had finally been released in May 2000. Selling brand-new software is always difficult until a reference base of satisfied early customers exists. Ellison decided that Oracle’s own Internet-derived efficiencies would be the first reference. Soon, Oracle’s $300 million marketing budget was pumping out the insistent message: “By using our own E-Business Suite, Oracle saved $1 billion in one year.” The grammar might not have been precise, but the meaning could not have been clearer: Oracle’s new applications suite was a cost-cutting engine that could do the same for any business with the vision to use it. It was the kind of boast that was bound to attract both attention and skepticism in equal measure. Partly because of Ellison’s reputation for stretching the truth when it suited him, critics were keen to find explanations for the cost savings that didn’t depend on the performance of Oracle’s software.

  An article in the magazine InternetWeek by Mitch Wagner published in March 2001 and entitled “Oracle’s Savings Don’t Add Up” was fairly representative. It began, “Don’t buy into Oracle’s claims that it’s saving billions of dollars by implementing its own e-business software. Much of the savings actually have come from cost-cutting measures that could have been achieved without the Internet, according to experts and an analysis of Oracle’s financial statements.” One such expert, John Puricelli, an analyst with A. G. Edwards & Sons, opined that most of the savings had come from old-fashioned belt-tightening: “A lot of the cost savings that came were changes in employee behavior, and software doesn’t do that,” he said. Chris Shilakes, an analyst with Merrill Lynch, said that “revenue momentum” as much as cost saving had been responsible for the margin improvement. Spurred on by such pieces, a group of shareholders, disgruntled by the fall in Oracle’s stock price since the beginning of the recession, even filed suit in early 2001 alleging that the company’s savings were due to head count reductions (from forty-four thousand to forty-one thousand since 1998), as opposed to Internet-driven efficiencies. Stung by these criticisms, Oracle invited the Harvard Business School to carry out one of its famous case studies and an independent management consultancy to carry out a meticulous department-by-department investigation.