Is Oracle's Fusion Coming Together?
A year ago, executives from software giant Oracle (ORCL) stood before customers, analysts, and the press to announce Project Fusion. That's the name they gave an ambitious plan aimed at stitching together Oracle's own applications with software acquired in a $19 billion spending spree on companies including PeopleSoft and J.D. Edwards (see BW Online, 1/19/05, "Oracle Lays Out Its Game Plan"). After a year of being pretty mum on details, Oracle marked the anniversary on Jan. 18 with a two-hour presentation, followed by a cocktail reception at San Francisco's City Hall.
In many ways, it was characteristic of most Oracle events. There was a red carpet with the Oracle logo draped over the City Hall steps. Along walls hung banners emblazoned with effusive praise from customers about how Oracle's technology would take them into the future. And no one knew until the last minute whether Oracle's unpredictable chief executive, Larry Ellison, would show. At the very end, Oracle President Charles Phillips announced that Ellison had the flu and couldn't make it after all -- despite promises by staff just two hours earlier that he'd be there.
So Phillips hosted the event, along with John Wookey and Thomas Kurian -- the Oracle executives in charge of putting the applications together and creating the so-called Fusion Middleware, the foundational code that will make it all work together smoothly (see BW, 1/23/06, "The Hardest Job in Silicon Valley"). Their message? The massive undertaking is halfway done, and the hard part is over. "A year ago, no one would have guessed we would have made as much progress as we have," Phillips crowed.
Phillips sought to debunk what he called "myths" about Fusion, which he noted is no longer called "Project Fusion" because so much work has been done, and the word project sounds "exploratory." For example, Oracle is not actually merging code of all of the disparate applications; rather, it's taking the best features and ideas and rewriting the code, based heavily on Oracle's eBusiness Suite.
Other notions he aimed to dispel: A Fusion upgrade would be costly and painful for customers, and a project of this scale had never been done before. "It has. We're just releasing a new product," he said, adding that Oracle releases new products all the time. "If we hadn't made [a single] acquisition, we would still need a next-generation" product, he added. His overall soothing message: Fusion isn't as scary as it sounds, and customers won't be forced to do anything.
Some analysts took solace from the presentation, which included product demonstrations. Pat Walravens of JMP Securities said he got a clearer sense of where Oracle was going and what progress it had made. "This road map should help customers plan and budget for the transition to the Fusion suite in the next two to three years," he wrote in a Jan. 19 research report. Similarly, Jason Maynard of Credit Suisse First Boston thinks Oracle is progressing well on its plan and says the middleware products could prove a valuable edge over arch-rival SAP (SAP).
Still, others came away unclear on how much progress Oracle had made. Charles di Bona of Bernstein & Co. was one of several attendees who questioned how the management was measuring they were "halfway" to Fusion, particularly since the final suite won't be delivered until 2008 and developers are just now starting to write the code. "Is Fusion on schedule?" he wrote in a Jan. 19 report. "The answer remains somewhat a matter of interpretation." Oracle's stock traded up modestly on Jan. 19, rising 1.5% to $12.52 in mid-morning trading.
Di Bona also questioned how PeopleSoft and J.D. Edwards customers would take to an "upgrade" of a product that has a completely rewritten code base. That point was seized on by Shai Agassi, a member of the executive board and president of the product group at SAP. He argued that if Fusion bore little resemblance to J.D. Edwards or PeopleSoft, it could hardly be called an upgrade. His point: The more the code is rewritten, the more it becomes a new application altogether. And the newer the software, the harder it may be for customers, especially large corporations, to switch from one to the other.
"They didn't want to say up front, 'You're going to have to migrate from your software,' because those words would have created a tremor," Agassi says. "Now it's been a year, and what they're saying is, We're not going to take that code, we're just going to take some of the ideas. You can't do automatic migrations because it's rewritten. If I'm looking at a whole new implementation, why is it going to Oracle?"
The market leader, SAP doesn't want to cede its 21% share. Oracle will have 11% when it finishes the acquisition of Siebel.
Oracle counters that it's giving customers an opportunity to ease into Fusion, by implementing certain features of the new application set in new versions of J.D. Edwards, PeopleSoft, and Oracle applications coming out this year. And Fusion Middleware -- the foundation for the new applications -- can already work with the old applications inherited from each company.
Wookey urged customers to start the process by upgrading to these new versions when they come out this year. He noted that worried customers had been holding off upgrading because of uncertainty over whether there was a clear path to upgrade from say, the new PeopleSoft suite to Fusion. "There is," Wookey said, confidently. For the last year customers have been waiting to see what Oracle would deliver. Now that Oracle has given them details, the market will monitor sales of new software licenses and Fusion Middleware in the coming quarters to see whether clients are getting on board or still sitting on the fence.
Meantime, SAP is planning to release a whole new application suite about a year before Fusion is complete, giving would-be clients yet another option. As for Wall Street, many still have the same impression they had a year earlier: It's a compelling vision that could very well knock SAP on its heels, but it will be tough to pull off.