Larry Ellison's One-Man Show

Critics say Oracle is faltering--and the CEO needs help

Oracle Corp. (ORCL ) CEO Lawrence J. Ellison isn't afraid to go it alone, on land or at sea. Last month, because of what he considered "prima donna" behavior, Ellison demoted one of four veteran skippers of his elite yachts, which will compete for the America's Cup this fall in New Zealand. It was the second such move in less than a year. Ellison, a respected skipper himself, says he'll pick up the slack at the helm.

Sailing enthusiasts should hope that Ellison does better without his skipper than Oracle Corp. has done since losing a handful of senior execs over the past two years. The latest piece of bad news came on Mar. 1, when Oracle warned it would fall short of its earnings forecast for the third fiscal quarter, ended Feb. 28. Oracle is expected to be just a penny shy per share, but this is the second time in a year that the company has disappointed. Analysts expect quarterly revenues of $2.4 billion, down 10% from a year earlier. The stock, trading at $14, has plummeted 70% from its peak 18 months ago and is down 12% since Oracle warned of the shortfall.

Ellison can't blame the crummy economy for this one. Once-fleet Oracle is in danger of becoming a slow-growth company even after the economy picks up. Its core database market is maturing and is expected to grow just 1.6% this year and 3.5% next year, according to Gartner Dataquest. And the rest of Oracle's products aren't humming either. Ellison put together a suite of applications a company needs--from financials to supply-chain management--all designed to run on the Internet. But the software was buggy, and sales have been dismal. They shrank 26% in the fiscal second quarter.

Making matters worse, while Oracle is stuck in the doldrums, other big software rivals that were late to the Net are now sailing faster than the company. PeopleSoft's revenues climbed 19% last year, and SAP and Siebel Systems saw revenues grow 17% and 14%, respectively. This year, Oracle's sales are expected to drop 9.7%, to $9.8 billion, and to increase only 2.5% in fiscal year 2003, says US Bancorp Piper Jaffray. By contrast, PeopleSoft Inc.'s revenues are pegged to rise more than 15% over the same period. "Oracle? They're dead in the water," crows PeopleSoft CEO Craig A. Conway.

Not at all. Oracle is still one of the most profitable software companies in the world. Net income is expected to come in at about $540 million in the fiscal third quarter, down $42 million from the year-ago quarter. Its 35% operating margin trails only Microsoft Corp. (MSFT ) and Check Point Software Technologies Ltd. (CHKP ), which makes security software. And its efforts to fix the buggy software are starting to pay off. A study in January by the Oracle Applications Users Group and Morgan Stanley Dean Witter & Co. found Oracle customers are getting comfortable with the software.

And on Mar. 21, at the Palace Hotel in New York, Ellison plans on revealing his strategy for boosting revenues. His pitch: Customers should forgo the hassle and expense of installing new applications on their own computers and instead pay Oracle monthly fees to handle all that at its data centers. "Guaranteed, you will never have to pay for another upgrade. You will never pay for another piece of software. You will never pay for another piece of hardware," says Ellison. He says customers will cut their bills by half, and, within a few years, the business could add $1 billion to Oracle's annual revenues.

This isn't a new push for Oracle, but Ellison says he has a fresh take on how to gin up business. Oracle began hosting corporations' software three years ago. Ellison says the difference now is that Web hosting has become a priority for his company. Every Oracle sales prospect will be presented with the hosting alternative before they buy new software. And Ellison plans on owning computer centers to ensure quality.

No matter how he rejiggers things, hosting may not reenergize Oracle. The company logged just $50 million in hosting sales last fiscal year. While analysts believe the $1 billion goal is possible, they say it will likely take up to five years to get there. "They need to sell software," says analyst Robert Schwartz of Thomas Weisel Partners.

What the company also needs is management reinforcements. Three years ago, rather than playing his normal big-think role, Ellison gradually took over daily management and transformed the company's internal operations using Internet technology. That ranged from centralizing data-processing operations to automating sales. Some results were positive: Oracle's operating margins doubled in two years.

In other areas, Ellison has been his own enemy. His insistence on running operations led to the departure of a handful of key executives, starting with former President Raymond J. Lane some 21 months ago. Lane had helped lead the company to nine straight years of healthy growth. Next to go was Gary L. Bloom, who was being groomed for a top slot but left to become CEO of Veritas Software Corp. (VRTS ) after Ellison made it clear that he would continue his tight control of Oracle. In the past year, Oracle has lost two top sales execs, one because of health problems and another was forced out, according to former executives. "Larry has done a marvelous job of setting direction, but he needs somebody to execute on his vision. It takes two people to run these companies," says analyst Rick Sherlund of Goldman, Sachs & Co.

Even Ellison's defenders admit he has his hands full. Oracle Chief Financial Officer Jeffrey O. Henley says concerns that Ellison has taken on too much are fair--although he insists Ellison's management style hasn't harmed Oracle's performance. Others disagree. A former executive says Ellison does well when he focuses on one thing at a time: "If he fixes one part of the business, another one is going to have to slip."

Oracle denies this, and Ellison says he has a strong right hand in Executive Vice-President Safra A. Catz. An investment banker before joining Oracle three years ago, Catz has been acting as his chief operating officer since Lane left. Ellison says he offered the COO title to Catz, but she turned him down because she wanted to stay out of the limelight. Catz declined to comment. Ellison says Catz probably wants to avoid the divisiveness of Lane's final days at Oracle--when the pair feuded over Lane's diminished role. "There was a public war inside the company," says Ellison. "I think she is afraid of creating a fracture" by becoming COO.

There's no denying that Ellison's management stumbles can be traced back to the spring of 2000, when he and Lane had their falling out. The delivery of the Oracle 11i business suite, the package of Net software, had been delayed for months and Ellison lost patience. He insisted on shipping it that May--ready or not. But 11i was riddled with glitches. Industry analysts estimate the suite required 5,000 to 7,000 "patches" to fix faulty programming.

Once customers loaded the software on their computers, they were in for unpleasant surprises. Execs at Dallas-based shipper Dynamex Inc. are still peeved about the human resources software they bought from Oracle last year. Because of quality problems, the software took longer than expected to install, and Dynamex felt Oracle should pick up the tab in extra consulting costs--some $25,000. Dynamex felt it couldn't abandon the $3.5 million project, so it hung in. Still, "I walked away with a bad taste in my mouth," says James H. Wicker III, vice-president of information systems at Dynamex.

Ellison doesn't apologize for shipping buggy software. He says it was unavoidable because the product was so complex. But, a year ago, he made matters worse at an Oracle conference in New Orleans when he blamed customers for their frustrations with his software. It would work fine, he said, if they stopped tinkering with it to fit their business processes. "The view of customers is `screw 'em,"' says Lane, now a venture capitalist at Kleiner Perkins Caufield & Byers. "That's exactly the opposite of the company I was running. I'm embarrassed by it."

Ellison says he's doing his best to make things better for customers. The hosting alternative eliminates headaches for customers. Plus, he's pushing a new database technology called "rapid application clustering." The idea: A bunch of databases can run on smaller computers and act like one big database. It's billed as less expensive because it runs on cheaper computers, and it's more dependable: If one computer fails, the others keep running.

The success of these new gambits is anything but assured. Ellison might have a tough time selling clusters of small databases--since he has long insisted that large, centralized databases are best for big corporations. Still, he's forging ahead with what he does best--creating new technology. If Ellison doesn't get something going soon, it's going to be hard to blow the wind back in Oracle's sails.

By Jim Kerstetter in Redwood Shores, Calif., with Steve Hamm in New York and Andrew Park in Dallas

    Before it's here, it's on the Bloomberg Terminal.