In the musical Chicago, attorney Billy Flynn shows client Roxy Heart how to razzle dazzle a jury to prove she's the naïve victim of "liquor and jazz". Larry Ellison seems to have watched and learned.
On Tuesday Oracle mounted a production of its own during a call with Wall Street, drawing on some familiar dramatic devices to explain results that were down - though not disastrously so, given the financial climate.
The problem for Oracle is that its world is not divorced from the real world. And when things go awry in the real world - such as a recession - Oracle has to dance harder to prove the efficacy of its mega-M&A strategy for permanent growth and excellent margins all around the stack.
Oracle saw a seven per cent drop in revenue for the fourth quarter ending May 31 to $1.8bn (£1.1bn) on revenue that fell five per cent to $6.8bn (£4.1bn). Earnings per diluted share were down a penny to $0.39 (£0.24).
Sales of software licenses and overall software revenue - that includes support and updates - were also down 13 per cent and three per cent respectively to $2.7bn (£1.6bn) and $5.7bn (£3.5bn).
For the year, Oracle saw a one per cent bump in net income - effectively flat - to $5.6bn (£3.4bn) on revenue that grew four per cent to $23.5bn (£14.3bn). EPS grew three cents to $1.09 (£0.66). Software sales fell five per cent to $7.1bn (£4.3bn) but overall software revenue grew six per cent to $18.8bn (£11.4bn).
But enough with the numbers - it's time for the theatrics.
First, it was the magical cloak of moving numbers. Oracle breaks out the impact that currency fluctuations have on its results each year and quarter - it calls this "constant currency".
Announcing its latest results, Oracle made repeated and uncharacteristically heavy references to how much better its results would have been had it not been for those pesky real-world currency fluctuations.
With constant currency factored in, Oracle said that its quarterly net income would have grown five per cent and revenue four per cent, while new software licenses would only have dropped a mere four per cent. For the year, net income would have jumped 11 per cent on revenue that grew 10 per cent, while overall software revenue would have seen a 12 per cent growth.
Oracle explained its reason for including constant currency - as it does each quarter and year - as providing "a framework for assessing how our underlying businesses performed excluding the effect of foreign currency rate fluctuations."
Unfortunately for Oracle, it's judged - like every public company - not on performance in some parallel universe, but on what happened in this reality.
Which leads us to theatrical device number two: deploy smokescreen of customer wins at the expense of the competition - a standard Oracle tactic.
For this the company rolled out its big guns: chief executive Larry Ellison and presidents Charles Phillips and Safra Catz.
On the firing line was SAP, which - as ever - just couldn't seem to stop losing business to Oracle in its own backyard, Europe. Also losing to Oracle were Teradata, IBM, and open source - Oracle claimed a doubling-down on WebLogic by US bank Wells Fargo, a customer that had dallied with open-source middleware before Oracle bought the WebLogic products with BEA Systems.
Salesforce.com took multiple direct hits from Ellison, who said that Oracle would make all its applications on-demand and on-premises using the same code base, and that Oracle could become the number-one supplier of on-demand applications.
Ellison defined "on-demand" to mean versions of his company's applications hosted and operated by Oracle either on customers' sites or off-site, or in a software-as-a-service model via the browser. "Our on-demand business is much smaller than on-premises business, but is growing faster. You will see a shift over the course of a decade," Ellison promised.
Asked by one analyst whether Oracle is getting into cloud computing, Ellison said: "a little bit".
Then it was time to wow The Street with technology: Ellison declared he was there to talk about Exadata, the server appliance launched last year and built with HP. Ellison listed early Exadata customers including "a Californian smartphone manufacturer" - translation: Qualcomm - plus RIM, Amtrak, Thompson Reuters, and Barclays Capital in the UK who'd dumped Teradata or IBM DB2 on the mainframe in different cases.
Ellison claimed these as actual customer wins, not freebies, saying that when Oracle had offered Exadata on a try-before-you-buy basis, customers bought because the results were so good.
In one of the few nods to the impending Sun Microsystems acquisition, Ellison said that the goal is to provide a complete and integrated stack, noting that Sun "gets down into the hardware. Exadata is our first experiment in the area and it's going very well."
Sometimes, though, the big guns overheat and hit things they shouldn't.
While blasting Salesforce.com, Ellison strayed into talking about late delivery of Oracle's Fusion brand of software. Now, it seems, Fusion Applications will be "announced" later this year and "customers will try them and we will start delivering them next year." That sounded like Fusion applications will actually be launched in 2010, and backs what we've reported.
Fusion's been a long time coming and has long been dropped from Oracle's central messaging and talking points. Delivery was initially promised for 2008, but in November 2007 that became "first modules" during calendar 2008. Separately, Oracle's Fusion Middleware 11g, also due in calendar 2008, is now set for an official launch on July 1 in Washington DC. ®