Oracle OpenWorld: memoirs of an Oracle

Oracle is 30 this year! Time for Larry Ellison to take an amble down memory lane for tales of dodgy accounting, promoted pizza boys and bluffing the CIA.

By Stuart Lauchlan, reporting from Oracle OpenWorld in San Franciso

What do you call a software firm that makes the pizza delivery boy its first CFO, sells the CIA software that hasn't been written yet and names itself after a failed IT project? Oracle, of course. Those were just some of the highlights recalled by a rather endearingly human Larry Ellison on Sunday as he took a (sometimes rambling) wander down memory lane to mark 30 years of Oracle.

Three decades ago in 1977 the firm was a handful of individuals working out of a little office; today it's a $20 billion a year company with over 300,000 customers around the world. This week, 42,000 representatives of those companies took over the centre of San Francisco for the annual OpenWorld jamboree - and they really did take it over. Entire streets are closed off by the city authorities to make room for party tents to accomadate thousands of IT types being entertained by the likes of Stevie Nicks and Billy Joel.

But on Sunday, the star of the show was the man himself, the longest serving CEO of a software company, Lawrence J Ellison III. So much has been written about him and his business methods over the decades - not all of it complimentary - that it's temping to assume you've seen the entire Ellison 'picture'. But the Ellison of Sunday evening was a markedly human individual, paying tribute to colleagues from the past, recalling anecdotes from the early days of Oracle and giggling outlandishly at his own jokes, not all of which worked for the rest of the audience, but which clearly amused him. The 'other software billionaire' has seldom seemed more down to earth.

The other Mr Oracle

He began - as well he might - by paying tribute to Bob Miner, the other founder of Oracle whose role in the creation of the software giant has been all too often overshadowed by the force of Ellison's charisma and personality. By chance, Sunday was the anniversary of Miner's death and his widow was in the audience to hear her late husband's business partner dedicate the evening to him. Ellison recalled her being frustrated when he used to turn up at the Miners house at dinner time several evenings a week in a bid to persuade Bob to come on board with him to set up Oracle. She finally told her husband to go and do it and they'd all pay the price!

The price was eventually worth paying as Oracle became the phenomenal powerhouse it is today. But in those early days, things were very different. Ellison recalled that Miner was more fortunate than him in so far as he had savings in the bank he could live off. Ellison on the other hand was so cash strapped during Oracle's early days that Don Lucas - an Oracle board member since 1981 – took over his house mortgage to save him from a bank foreclosure.

Oracle's origins come from a set of specific circumstances and more than a little brazen opportunism. In the 1970s, Ellison and Miner were working for a firm called Ampex which was building advanced tape drives. Ampex was contracted by the CIA to create what was in its day a "huge terabit mass storage system on tape drives."

"We looked at the mainframe marketplace and saw that the biggest independent software firms were database companies. We determined to do the same thing, but for minicomputers." Larry Ellison, CEO, Oracle

Soon after this, Ellison changed allegiances to a rival firm called Precision which used laser technology to store data – as vice president of software. One of his first tasks was to contract out development of some the firm's software, most notably the creation of database software.

Miner was also persuaded to jump ship and bid for the business. The bid was successful - but also a sign of the almost joyously haphazard way Oracle would be run in its early years. “The spec was written and we bid $300,000 dollars,” Ellison recalled. “We won easily because the next lowest bid was $2 million. But I was so keen to win the business."

Winning the bid led to the formal setting up of Software Development Laboratories (SDL). "We capitalised the company," said Ellison. "I put in $1,200, we had $2,000 in the bank, and were sitting pretty – working out of a modest office at Silicon Valley that Precision had rented to us.”

The key decision that was taken that made all the difference was to base product development on research work on relational technology that was being undertaken by IBM with its System R project and to pitch that at the mid-range market, not the traditional mainframe database market that was dominated by the likes of Cullinet.

“We looked at the mainframe marketplace and saw that the biggest independent software firms were database companies. We determined to do the same thing, but for minicomputers. We thought: ‘to heck with this old stuff; let's build the first commercial relational database using the IBM research documents as our specification.'”

Having made this decision, the next step was to go off and sell a product that hadn't yet been built! More than that, sell it it to the US Government - if you're going to bluff it out, bluff it out big! “The only people I knew were the guys I worked with at CIA," admitted Ellison. "So I went to the one customer I knew. The CIA liked it and I made my first sale of $48,000. I went back to the office and told the guys that they had to finish building the product because I'd just sold it to the Government!"

Crashing at 150 miles an hour

The name changed to Oracle soon after. The original terabit database project for the CIA had been known as Project Oracle. It had failed which was in the event good news for SDL as it meant that in 1982 Ellison and Co could nick the name. "Because the project failed, the name was available," giggled Ellison. The first version of the database was Release 2.0. "Who would buy Version 1 of a database from four guys in California?" Ellison demanded, not entirely unreasonably.

The firm won more business from the intelligence community and began to expand its headcount. Notable recruits included Tom Siebel, who later became one of a long list of senior executive who fell out with Ellison and went off to form their own firms. Siebel's was of course Siebel Systems and led the CRM marketplace. Ellison offered a few magnanimous crumbs in Siebel's direction, conceding that he had set up a "brilliant" company. But then magnanimity comes easier once you've bought out your rival presumably!

There were also some conciliatory noises in the direction of former Oracle president Ray Lane who Ellison said had been "the best field guy in the United States" and who led Oracle to $10 billion a year in revenue by the late 1990s. Lane and Ellison had a hugely acrimonious falling out in the end and despite rumours that he might not make an appearance, Lane was also among the couterie of Oracle alumni who sat at the front of the audience on Sunday, and the very mention of his name brought a huge round of applause from the customers in the hall.

There was also some mild rewriting of history. The infamous Oracle crash of 1991 was cited, but prefaced with a claim that it was part of the early 1990s recession. Well, perhaps that played a part, but the out of control sales frenzy culture cultivated directly by Ellison and the frankly reckless book-keeping and accounting practices at the time were far bigger factors. And Ellison knew it: "I drove the company into a brick wall at 150 miles per hour!"

Much speculation surrounds the events of the week or two following Oracle's first quarterly loss in 1991 as Oracle's then head of operations Geoff Squire wielded the axe and carried out the necessary blood letting to keep the firm afloat. (Squire's role in Oracle incidentally was only noted in a brief, throwaway line. But at least he got a name check. Others, such as EMEA president Pier Carlo Falotti went unnoted...). But this week Ellison recalled that he had been warned that Oracle might go bust and admitted that he had been worried he would be sacked!

In the event, mass redundancies resulted as well as the appointment of a new CFO - Jeff Henley. Oracle's accounting practices had also been suspect, right from the early days. "We were trying to get a bank loan once so they said send in your financial statements," Ellison explained, adding that the problem was that no-one in the company understood balance sheets. After copying other firms balance sheets, the answer arrived in the shape of the regular pizza delivery boy who it turned out was studying accounting at U.C. Berkeley. Oracle hired him. "He said he wouldn't quit college, but he'd help us do our books," chuckled Ellison. "We said, 'take whatever you need [for pay], we'd never know.'"

They were very different days. No software start-up could get away with stuff like this in a post-Enron world. It was also a very different sort of keynote for the slick CEO who's more used to panther-like prowling of the stage, outlining his latest 'vision thing' and shamelessly baiting the competition. Instead, we got Ellison alone on stage, standing in the spotlight, clutching a set of notes from which he kept deviating and losing his place, but delivering what was clearly a personal testimony about the firm with which he has become synonymous.

So what in the end was his conclusion about the key to the firm's success? Was it the technology? Was it the various management teams, culminating in the current powerhouse pairing of Charles Phillips and Safra Catz? Or was it the inevitable customers without whom etc etc etc? Maybe it was all three. Or maybe the real answer lies in the throwaway line about the early days of the company: "We had absolutely no adult supervision whatsoever!"

Happy birthday Larry!

Related stories