For 30 years all of Oracle Corp., maker of the database software that drives thousands of big businesses around the world, has revolved around its founder. Larry Ellison owns a 23% stake worth $18 billion, and he rarely sells.
He tweaks Oracle's print ads; he fiddles with its press releases; he peppers techies with arcane questions. "I've run engineering since Day One, and I still run engineering," he says.
But Ellison is turning 62 on August 17. Isn't it about time he identified a successor? Bill Gates, 11 years younger, managed to do that.
A litany of would-be heirs to Ellison have come and gone, forced out or departing at will when they clashed with the boss or stepped too far into his spotlight. His two most senior aides are ill suited to the job: the mysterious Safra Catz and the articulate glad-hander Charles Phillips.
- Slideshow: The World's Billionaires - Larry Ellison
- Slideshow: The 400 Richest Americans - Larry Ellison
"I don't know who would take over if something happened to Larry. I don't want the job," Catz, 43, tells Forbes in a rare interview. As co-president and chief financial officer, she is Ellison's enforcer and chief of staff.
"Without him, Oracle wouldn't be the same." ("She's seen what fame and fortune bring, and she's not impressed," Ellison says mordantly.)
Phillips, 46, co-president and a former Marine Corps captain who came to Ellison's attention in the early 1990s when he recommended Oracle stock as an analyst at Morgan Stanley, adds: "Larry will be here forever. We don't discuss succession. That's not my job."
But it is the job of the board, and it, too, has failed to define Oracle After Larry. "There is no successor to Larry, no heir apparent," says board Chairman Jeffrey Henley, who ran finance from 1991 to 2004. "We discuss the subject, but there is no perfect plan. Larry still wants total control."
It's a brazen attitude in this Sarbox era of abundant regulatory second-guessing and crackdowns on slack corporate governance. And it is precisely the way Ellison wants it. "It is silly to think you can groom someone. Jack Welch didn't have a successor. If you want to put me in his league, by all means, guilty as charged," he says.
That isn't quite true: At General Electric, Welch went to elaborate lengths over six years to groom a successor from a handful of clear finalists. And in a recent BusinessWeek column the star manager criticizes complacent boards for ignoring the succession issue.
Corporate governance finger-wagger Nell Minow of the Corporate Library concurs. "Founders don't like distinguishing themselves from the company. Competition and self-confidence often get in the way of succession planning, but shareholders deserve more." Her group gives Oracle a grade of "D" for governance.
Ellison eschews succession in favor of a trusted staff that faithfully takes his lead. He spends much of his time away from Oracle: He doesn't even have his own office at the headquarters in Redwood City, Calif.; Catz occupies it.
"Life is short. I'm not going to spend every minute of it with Oracle," he says. He puts in scant time with customers. "Just not interested," he says, leaving that role to the smooth-talking Phillips.
This summer Ellison is spending three weeks sailing off the coast of Spain to qualify for the June 2007 America's Cup yacht race, his second. He is tweaking the design for his next yacht, a 250-footer that will be cozier than his other one, the 454-foot Rising Sun.
He also is finishing up a Japanese-village-style estate on 40 acres in Woodside, Calif., where he lives with his fourth wife. (Two grown children from an earlier marriage are not in the business.)
"Some people say Larry is lazy," says director Donald Lucas, who joined the board when Oracle had 11 employees. "But really he empowers the people around him by delegating. And if he nominated a successor, everyone would be gunning for that person."
Yet Oracle needs to secure its future -- especially now, as old-guard software vendors get undercut by a raft of cheap newcomers in the open-source movement. For Oracle the threat comes from such outfits as MySQL and Ingres. DaimlerChrysler, Siemens and UPS use MySQL, spending (so says MySQL) one-tenth of what they would have spent on the Oracle alternative.
"Innovative companies aren't buying Oracle's stuff. They're flocking to open source because it is better and cheaper," says MySQL Chief Executive Marten Mickos.
The threat explains Oracle's tepid valuation on Wall Street. At $15, its shares go for 19 times trailing earnings, scarcely better than the market's 16. Rival SAP enjoys a price/earnings ratio of 27. Its shares have barely budged in the past three years, notwithstanding that earnings per share have increased by 49%.
"Wall Street is often the last to know," says Ellison, who can be quick to lecture. He predicts earnings growth of 20% a year for the rest of the decade.
That goal could prove to be elusive, for Ellison's main database business, still nearly two-thirds of revenue, grows only 3% a year, according to Citigroup. Just about every big company that needs a database has one.
"We've already got what we want," says Nationwide Insurance Vice President Richard Wohleber. Michael Vincent of Dutch bank ING adds: "If anything, our spending will decrease."
To offset that inevitable slowdown in new database sales, Ellison has been betting big on applications software. Databases store the bits of a company's information, everything from auto part inventory levels to personnel records.
Applications organize, present and track that data, offering up insights like which customers are the most profitable ones and which salespeople land the best deals.
Oracle spent a decade trying to build the applications business in-house, to no avail; its homegrown products were user-unfriendly and riddled with bugs.
Yet Ellison personally has invested in two applications firms that have fared far better. Three years ago he reversed course and embarked on a shopping spree. It has been a big spree, totaling $19 billion and delivering 21 software companies, $4.6 billion in extra annual revenue and 18,000 additional employees. (Some 7,000 were canned, and the Oracle payroll now is at 56,000, up 34% from 2002.)
Among his prizes is PeopleSoft, a vendor of software to run personnel departments, which was won in a hostile $11 billion takeover.
When Ellison made a move on PeopleSoft, it was run by one of his myriad protégés-turned-castoffs, Craig Conway. In a similar vein was Siebel Systems, founded and run by another Oracle alum, Thomas Siebel. Siebel software helps salespeople sell.
Competitors paint Ellison's buyout binge as a capitulation. "This is a desperate act. In the Wild West, when you're riding a dying horse, you get off and shoot it. In the software industry, you tie it to another dead horse and see if it'll go anywhere," says Shai Agassi, president of SAP.
Ellison counters that SAP moves too slowly to update for the Internet age and will suffer for it. "The only thing more risky than moving to the next generation of technology is not moving to the next generation," he says. "So we'll see who's taking the risk and who isn't."
Now Ellison must make these pieces of software work together. Oracle has two versions of financial-tracking software, three sets of code for customer data and dozens of niche products. Ellison vows to continue investing in older versions of programs Oracle has acquired, even as engineers work on a grand unified design dubbed Fusion. It may be ready by the end of 2008.
For every dollar corporate customers spend on new software, they spend $6 implementing it and getting it to talk to other programs. Ellison says Fusion can greatly reduce this pain.
Other software execs see a daunting challenge, especially in the time frame Ellison promises. "This is the most complex undertaking ever attempted in software," says William Coleman, who cofounded Oracle rival BEA Systems and now runs server-tech firm Cassatt.
Some customers won't wait.
"All of Oracle's software would take years to integrate with what we already have. Instead, I'll have my guys build the software we actually need to run our business," says Alan Buffington, tech vice president at financial services firm KeyCorp. "We're operating in days or months, not years. We won't get our competitive advantage from Oracle."
Lawrence Joseph Ellison was born in Manhattan on August 17, 1944. He never knew his biological father; his unwed mother, Florence Spellman, left him to be raised in Chicago by her aunt, Lillian, and Lillian's husband, Louis Ellison, who had emigrated from Russia, renaming himself for Ellis Island.
The Ellisons adopted the infant, but Ellison says his great-uncle repeatedly told him he would never amount to much. (Larry reached out to his biological mother only in the the early 1990s, buying her a house near Oracle's headquarters; she died of cancer in 1999.)
He was a mediocre student. "A Latin teacher once told me that the 'F' she gave me would ruin my life. I didn't believe it and told her so. A little letter in a square box? It all seemed so silly."
He dropped out of the University of Chicago in the 1960s, headed to Berkeley, Calif. and by the mid-1970s began working on a database project. Code name: Oracle. Client: the Central Intelligence Agency.
In 1977, Ellison founded the company with Robert Miner and Edward Oates, naming it after the CIA job. (Miner quit in 1992 and died of lung cancer shortly after. Oates retired in 1995 and sold the last of his 2% stake last year.)
Ellison has been in control for three decades, and only once was that even remotely in question, after an accounting debacle back in 1990. Sales slowed, expenses spiraled upward, Oracle posted its first losses, and the stock tanked; the company had to borrow $170 million to keep afloat.
The board never really considered replacing Ellison: After all, he had no successor. The board implored him to recruit more seasoned execs, and soon afterward Jeff Henley joined as chief financial officer and Raymond Lane joined to run sales.
By 1995, Ellison had a vision of the future, prescient but premature. He began talking about a time when computers, smart TVs, videophones and other gadgets would link up via a worldwide network of information and entertainment services. Oracle would power this new era.
It didn't. A wireless Web browser for consumers that Oracle launched in 1999 was quietly folded into its business server line. "We used to joke that Larry is brilliant but time-dyslexic," says Mark Porter, who worked at Oracle in the 1990s on video-on-demand technology that never debuted.
In 1997, Oracle issued a new set of homegrown applications, which flopped. Lane, who a year earlier had risen to president and was seen as Ellison's heir apparent, began advising the boss that only a drastic change in leadership would save the applications business.
Ellison responded by naming himself head of apps and slowly stripping Lane of his responsibilities. By June 2000 Ellison all but fired him, taking away Lane's title of president.
Lane quit and is now a general partner at venture capital firm Kleiner Perkins Caufield & Byers.
In Lane's wake, 14-year veteran Gary Bloom hoped to become Ellison's number two but was spurned. He quit a few months later to run Veritas, now owned by Symantec.
"This is a team, and Larry is the only captain," says the board's Lucas. "If someone wants to pop up and announce they're the star -- poof! You're out."
It had happened before. Tom Siebel, a star Oracle salesman, left in 1990 to found Siebel Systems when Ellison rejected his idea for a sales-force product. Terence Garnett rose quickly to head marketing, then quit in 1994 after clashing with Ellison. As a venture capitalist he now backs rival Ingres.
More recently Oracle has blown through two finance chiefs in two years. Harry You lasted eight months and left to run BearingPoint. Greg B. Maffei, a former Microsoft finance chief, lasted only four months under Ellison. He now runs John C. Malone's Liberty Media.
"Larry's personality mandates that he's in charge, so he can't have a successor," says alumnus Marc Benioff, who left in 1999 to found Salesforce.com with a $2 million investment from Ellison that today is a 4% personal stake worth $100 million. "But one day he'll have a revelation, look outside for talent -- and it will likely be a former Oracle executive."
By the time Lane and Bloom had quit in 2000 Safra Catz had taken control. A dealmaker at the brokerage Donaldson, Lufkin & Jenrette, she had covered the software industry since 1986 and was a close friend of Ellison.
He was enchanted by her intellect (Though Larry likes to say she graduated in the top of her class at Harvard Law, she attended that school for just her final year. Her degree is from the University of Pennsylvania), and 13 years later he brought her to Oracle.
They speak five times a day. "Her power isn't that she has a lot of people working for her; she doesn't," says Henley, the board chairman. "Her power is that she's on the same wavelength as Larry."
Catz gets tough when the sales force wants to discount prices. She's the efficiency queen. She folded 70 employee databases into a single one to scrutinize labor costs. Oracle had dozens of call centers when she arrived; now it has just five scattered across several time zones.
Oracle is now neck and neck with Microsoft in profitability, with an operating (Ebitda, that is) margin of 42% in its last fiscal year (versus 43% at Microsoft). That margin is up from the 26% it was when Catz arrived.
By mid-2003 Charles Phillips had quit Morgan Stanley to run sales, cozying up to customers in the "I'm listening" role that Ellison avoids. Catz and Ellison began to look at acquisitions.
Ellison had insisted he wouldn't buy his way into applications, saying acquisitions were disastrous, given the technical challenges and the fact that engineers often flee after a merger. His reluctance faded amid Oracle's faltering in-house efforts and his own view that the software industry was slowing down and might never reaccelerate.
"Everyone's top- and bottom-line growth had slowed. IT budgets were in decline. The funny thing was that this was patently obvious, but no one seemed to notice," he says. In April 2003 he met with the board to pore over a list of eight targets. Oracle spent 18 months in pursuit of PeopleSoft, prevailing over the objection of the U.S. government and pressing the PeopleSoft board into firing Ellison's old charge Craig Conway. The final price of $11 billion was more than double Oracle's initial offer.
The PeopleSoft deal closed in January 2005, and Oracle had bought eight more companies by last September (Retek for $665 million, Iflex for $725 million, and such firms as Oblix, TripleHop and TimesTen), culminating in a bid for Siebel. That deal closed in January. By June 2006 Oracle locked up 11 more.
The rash of dealmaking doubled Oracle's share of core business applications (excluding desktop apps like spreadsheets) to 10%. SAP has 28%. The cost of that expansion can be seen on Oracle's balance sheet. Over the past two years it has spent $14 billion in cash, not quite half of that money borrowed in the bond market.
Ellison, ever in character, says it was all worth it. Some 95% of the newly acquired customers stayed on; Oracle now serves 30,000 applications accounts.
"Keeping these customers is an immensely profitable business for us," he says. Catz adds: "We now have a much larger presence in all these accounts. Lots of SAP customers are now Oracle customers because they were running Siebel or PeopleSoft."
Between databases and applications, Oracle gets a steady annual residual of $6.6 billion a year in support revenue, almost half of its $14.4 billion total. For every $1 million in new software sales, Oracle gets $220,000 a year in upkeep revenue over a five-year stretch. That predictable stream is roughly 40% more profitable than writing and selling new software. And this software, once in, rarely gets ripped out.
"Open source would be interesting if we could start from scratch, but we're with Oracle for the next 20 years. It hurts too much to switch," says Kevin Horner of Alcoa, which went all-Oracle in 2000.
A key advantage of his flurry of deals is that, Ellison says, "Our applications engineering team is twice as big as it used to be. And it's not just size, it's the experience they have of being first. Banking, telecom and government. We weren't first before we bought these companies."
Bradley Wright, tech chief at Houston engineering firm Carter & Burgess, agrees. "Now Oracle has the best minds in the industry," he says. He recently chose Oracle to manage contracts and expenses.
Ellison is betting his biggest customers will enthusiastically embrace Fusion. Most of Oracle's 15,000 developers now work on new software, borrowing the best features the company acquired.
The building blocks of Fusion were completed last year, says John Wookey, Oracle's applications chief, and by 2008 they will become a seamless amalgam of databases, applications and the middleware that stitches them together. Part of the selling proposition: unified security and user access; quick upgrades.
Some clients are dubious. "Fusion is just another new release with similar features. We won't spend more because of it," says David Ernst, chief technologist at the California State University system, Oracle's biggest college account.
But David M. Rudzinsky, tech chief of medical devices firm Hologic, is more optimistic. "Customizing the software is now Oracle's job. Before, that was our problem," he says.
Ellison, for his part, will just keep on sailing -- solo, when it comes to the succession question
Back in 1998, when he sailed his yacht, the Sayonara, in the Sydney to Hobart race, he and his crew endured 630 miles through rough seas and hurricane-force winds off Australia. Ellison was lucky to come out alive; five boats sank and six sailors died. Next April he will be on the seas again sailing in the America's Cup qualifiers.
"I'm a major in-the-moment person," he says. "If this is what I'm doing now, I'm not going to be lusting for the next day or something that's going to happen two weeks from now. I'm just going to enjoy what's happening now." One wonders whether his shareholders feel quite the same way.