In-Depth

The Leap from Illusion to Fusion

• By Alan R. Earls
• November 1, 2005

The Big Idea

Virtual Certainty

• Traditional software development will almost entirely be supplanted by a virtual paradigm in which Web services are orchestrated dynamically to power composite applications.
• Bring in WSDL files from Web servers, and you create pipelines of activity. If you string enough of those routing and transformations together, pretty soon you have an entire enterprise application infrastructure.
• Organizations automate many once-manual business processes; simplify their app dev and maintenance by introducing modularity and effective reuse; and optimize their business processes to ensure they align with IT.

As the great English lexicographer Samuel Johnson once noted, "A second marriage is the triumph of hope over experience." Oracle, now well past its acquisition of Siebel and PeopleSoft/J.D. Edwards, is pinning its hopes for a successful marriage on a strategy it calls Project Fusion. Oracle aims to eventually rationalize its sprawling product line by combining all the best features of each constituent product into a new and much improved whole.

By some accounts, it's an overly ambitious goal, but many analysts-and those close to the user community-are optimistic Oracle will pull the rabbit out of the hat. For one thing, as Oracle is happy to point out, it will leverage a wide range of existing Oracle middleware products, recently renamed Fusion middleware. Beyond that, Oracle hopes to fund the development effort with ongoing maintenance revenue from its greatly enlarged customer base.

The skeptics, not surprisingly, include competitors such as BEA's VP of Solutions and Product Marketing Bill Roth, who says, "It is going to take nothing short of nuclear fusion to bring all those products together." Roth says he bases his opinion on the extensive differences among the existing products, starting with the user interface and extending deep into the code. "I know a company that had a terrible time just integrating PeopleSoft with their own apps, and everyone has heard Siebel horror stories," he says.

"Our perspective," Roth adds, "is that you need to start with integration as your design center and then build the applications, not the other way around."

SAP's Peter Graf, product marketing executive VP, is equally critical. To him Project Fusion is a "chameleon" that has continued to change its colors to fit Oracle's marketing needs since the original Webcast in which it was announced. "At first they were talking about an all new Java-based platform with Java-based functionality, but everyone who knows this software knows there is no way you can do it that way, especially in the time they are promising," he says. So Graf asserts, behind the curtains, Project Fusion had a metamorphosis and "those original promises are now being contradicted."

Big, really big, job
In fact, Oracle's message has shifted somewhat. However, for analysts and the leadership of Oracle's major user groups, that's just a result of Oracle's coming to grips with what by all accounts is a gargantuan undertaking.

"I think they will accomplish what they are setting out to do, the question is when and how much will it cost to follow that path," says Paul Hamerman, an analyst at Forrester Research.

Hamerman agrees there will be challenges for Oracle in integrating the functionality and in combining the products into a unified whole, but, he notes, "Oracle will not be rewriting from scratch." Instead, he says, Oracle leadership tells him it plans to use Oracle's E-Business Suite schema and its functions and then pull in elements from other products.

"Their latest acquisition complicates things a bit because they are now saying that Siebel will be the centerpiece of their future CRM strategy because it is a more comprehensive offering," Hamerman says.

And while getting all the pieces to fit might seem almost impossible to some, Hamerman says he believes the approach will work. "They are working within a fairly standard architecture based on J2EE as well as their middleware-though they have indicated they will also support Web- Sphere middleware," he says. That boils down to what Hamerman sees as a reasonable combination of Oracle proprietary and open-standard elements.

In general, Hamerman says the most stable elements are likely to be the financial and HR functionality. Much will depend, he says, on "the functional footprint" of E-Business Suite, and especially the Oracle E-Business Suite Human Resources Management component. What will change is the user interface because, "Oracle knows that's a weak spot in terms of usability, and they have indicated that what will eventually emerge will look more like PeopleSoft," Hamerman believes.

Although some question how Oracle will pay for all this, when they have also committed to maintaining and upgrading their existing offerings, Hamerman doesn't believe that's an issue. "The whole rationale of the acquisitions was customer retention, which means customers will continue to fund a high-margin revenue stream." What's more, Hamerman says, Oracle can begin to look at that customer base as almost a private market to which it can sell other Oracle products.

"I think the success of this effort will be based on how quickly they can get customers to adopt it," he says. Customers will be reluctant to be early adopters, especially if the new products appear to be buggy. So Oracle must prove out what it is offering, he says. Still, customers are likely to come aboard only gradually, Hamerman believes.

First, do no harm
Like Hamerman, Denis Pombriant, managing principal at Beagle Research, is trying to figure out what Project Fusion is about, based on limited information. "It is not real concrete at this point, and if you listened closely to the press conference at which they announced the Siebel acquisition, it seems that they don't have a coherent vision themselves," he says. However, adds Pombriant, that's because what they are working with is "still wet clay."

For Oracle, he says, the prime directive is to do no harm. The existing products have to work because there are too many big companies that aren't willing to simply write off their existing software and go to whatever Oracle suggests. More to the point, he notes, the software that's out there still works just fine. "I'm fond of reminding people that the Oracle announcement didn't cancel anyone's software license-people will go through a normal process of using their products, and at the appropriate time they will take another look at what to buy next." And that, he argues, is actually good for Oracle because "if they were asked to sell a Fusion product in the next 10 minutes, they couldn't."

Putting the mojo to work
But Oracle can sell middleware. "In classical Oracle form, they have managed to mystify the entire market by naming everything Fusion," says Jim Shepherd, an analyst at AMRResearch. He characterizes Oracle Fusion middleware, portrayed by the company as the mojo that can make Project Fusion work, as "Oracle drawing a circle around products they already had," which is why they were able to claim thousands of users only a month after they introduced it, he says.

In addition, he says, the components of Fusion middleware give Oracle the ability to offer development tools, a BPEL-engine, integration tools and more. Those offerings, Shepherd says, are not only useful for IT departments but "are also probably getting to the point where ISVs could consider using them as a platform," he says, which he predicts will probably happen over the next year.

When it comes to Project Fusion, Shepherd says Oracle is "smart enough to understand that you can't really integrate disparate code, especially when it has been built by different companies with different design philosophies. What they intend to build and what they talk about as Fusion (the application) "is a next-generation product which will use the functional design from the existing products but not the code," says Shepherd.

"SAP and others like to imply that Oracle is pouring millions of lines of code into a Cuisinart and they will be spitting out some kind of Frankenstein monster," says Shepherd. "I don't mean to imply that it will be easy or even that anyone has ever done this before-it is daunting- and just the fact that there is a huge degree of overlapping functionality makes it tougher," he says. Likewise, sorting out which functions to use from each application is going to be tough. Putting in place a reasonable migration path for customers will not be easy either. As a result, he says, Oracle will have to write a lot of code from scratch, "but the good news is that they have the functional design and the not-too-ugly news is that maybe they can sort out a data model that is reasonable," he says.

Unrealistic optimism, most likely
Shepherd says on the output end of the process, the trick will be constructing the new application "in a sophisticated form so that it is not so big that no one can possibly use it," says Shepherd. And the fact that Oracle says they are targeting delivery in 2008, Shepherd thinks is "wildly optimistic."

"First, no one knows how to do it; second, it is an enormous task; and third, you have to continue to maintain all your existing products to maintain your flow of maintenance revenues," he says.

But it will happen and, in his estimation, means that the handwriting is on the wall for other ISVs. "I think if you look at other industries, the implication here is that the number of large global OEMs that can survive in a given market is small," he says. Thus, he says, it is unlikely that that there will continue to be hundreds of supply chain vendors or 50 to 75 CRM vendors. "That just doesn't make sense," he says. Further, Shepherd says, "Customers have clearly indicated they want to buy single-vendor integration."

Making history, maybe
Marc Hebert a former Oracle CIO, and now marketing executive VP at systems integrator Sierra Atlantic, says his company has been involved in helping Oracle develop a BPEL-workflow product as part of the Fusion Middleware.

In his view of what lurks inside Project Fusion, "the middleware is the underlying integration platform," and the base of the Fusion application project is the Oracle database. Layered on top, the middleware is part of the application server layer. "The whole architecture is nice and tightly integrated, which is probably Oracle's strength in this strategy," he says.

Still, he admits, "the Fusion project itself is perhaps the biggest, most challenging software development project in the history of enterprise software, and it is very hard to say how it will turn out."

To its credit, Hebert says Oracle has done a very clean execution of its acquisitions, in particular by providing a clear roadmap and developing a strategy that gives them plenty of leeway. And they have announced roadmaps for existing products, which has calmed down their existing customers. So, he says, "If Project Fusion turns out to be 5 years out instead of 3, they can still keep their existing roadmaps going indefinitely so customers will have a normal, rational upgrade path."

Further, Hebert says, customers are buying it. "I have sat in on independent user group sessions, and I have seen the show of hands-customers are saying that everything is OK." Besides, he adds, there are few alternatives, which is part of Oracle's strategy.

Sunny side up
Hebert's assessment of customer comfort levels is echoed by John Matelski, chief security officer and deputy CIO for the City of Orlando, who also serves as president of Quest International Users Group (comprised of traditional Oracle customers as well as customers of the companies Oracle has acquired).

On a personal level, he says the City of Orlando has been a J.D. Edwards customer since 1997 (currently running the JDE 8.10 EnterpriseOne Financials) and has also been a longtime user of Oracle database products.

The PeopleSoft acquisition of JDE was not a pleasant experience. Based on the uncertainty that exists with any merger/acquisition, and the fact that only 18 months had transpired since the first merger, Matelski was very concerned about Oracle's potential acquisition of PeopleSoft.

"I was vigorously opposed to such an acquisition, not because of the merits- but rather the anticipated impact that it could have again on the City of Orlando's implementation." However, Matelski adds, to Oracle's credit, "They did an excellent job of communicating with the city and the user group/ customer community, and were able to quickly quell my concerns. Since then, he notes, "All of our interactions with Oracle have been very positive, and the customer interactions have improved immensely."

He also credits Oracle with a much better understanding of user groups, and an excellent Global customer support program, which encourages and sustains independent user group activities. "They also understand that user group support is not a marketing role, but rather a customer advocacy function," he says. Thus, the user group support model has provided "avenues into our mission-critical ERP vendor that never existed with PeopleSoft-not only do I have access to the product development managers that I always had access to, I now have access all the way up the corporate ladder," says Matelski. Further, he says the accessibility of senior executives, such as Juergen Rottler, John Wookey and Charles Phillips, is unprecedented. "All of these lines of communications have been created due to Oracle's customer focus," Matelski says.

Whereas PeopleSoft seemingly had no product strategy vision under the leadership of Ram Gupta, John Wookey seems to have the vision needed to move customers into the future, Matelski believes. "Though that vision will not prove out completely until we see a product developed, the fact that there is a vision for the future is refreshing," Matelski says.

Matelski is also on the executive board of the International Oracle User Council. The IOUC consists of leaders from the six currently recognized umbrella user group organizations that Oracle works with, including Quest, Oracle Applications User Group, Independent Oracle User Group, Oracle Development Tools User Group, Asia Pacific Oracle User Council and Europe/Middle East/Africa Asia Oracle User Council.

Matelski says a product development committee focused on the evolution of the Fusion product was created as a function of the IOUC. "I do not believe that our committee has been given access to any information that the general public has not received, however, we have been tasked to work closely with Oracle senior management to provide a two-way conduit between Oracle and the user group community (and by extension, the entire global customer base), to gather information and provide tactical and strategic feedback," he says.

Clarity requires a product
Looking ahead, Matelski says, with Fusion modules anticipated in 2007 and an initial product suite expected in 2008, the City of Orlando has 2 or 3 years to plan for potential upgrades. "My reality is, now that Oracle has indicated that they will support all of the PeopleSoft and J.D. Edwards products until at least 2013, the City is in no rush to move off our existing solution [and] quite frankly, we will conduct our due diligence, weigh the benefits of enhanced features and functionality against the cost of upgrading, and decide if/when it makes sense to make a leap," Matelski says.

"Based on the current and evolving ERP landscape, I have a great deal of confidence that moving to Fusion will make sense at some point of time," Matelski says, "However, until the 50,000- foot vision manifests itself in a groundlevel product suite with the features, functionalities and price points that make sense-we will stand pat on our current release," he says.

"Overall, despite my initial trepidations, the Oracle acquisition of People- Soft was the best thing that could have happened to JDE product users," he says.

Steve Hughes, president of the Oracle Application User Group, is also upbeat. He says his group formed the Fusion Council to pull together input from the user community. "The council is global in its nature and scope, and includes 80 companies that are representative of the various products and geographies," he says. The group recently held its second meeting, at Oracle OpenWorld, and is now coordinating survey efforts with Oracle to ensure a broad input of ideas.

"What we hear from the community is that they have heard the Fusion story and there is more excitement than anxiety," Hughes says. "They are eager to understand the components and functionality of all the products so that they can get a better understanding of where the project could end up," he adds.

"I don't know whether people are happy or not, but when there is a lot of confusion, sometimes people decide not to decide. The thing that will eventually propel the market is clarity, and there is no clarity until there is product," research analyst Pombriant adds.

On ADTmag.com
Oracle touts portlet factory for SOA dev
By Kathleen Ohlson

Oracle Fusion architecture strategy
By John K. Waters

Fusion Council
By ADT Staff