Clarifying Oracle's Fusion Strategy, Thursday, March 30, 2006
Jim Shepherd

Oracle first unveiled its Fusion strategy about 15 months ago at the event that celebrated the conclusion of its PeopleSoft acquisition. Since that time, Oracle’s marketing team has valiantly struggled to refine and clarify the Fusion message, while the internal landscape has been constantly changing. The Fusion name is now also being used for the middleware suite, the overall software architecture, and as a code name for the next generation ERP product. Meanwhile, management has acquired 10 or 11 additional software companies.

This turmoil has led to lots of complaints among Oracle customers and prospects that they just don’t understand what Fusion is or what it means for them, and it has also created an opportunity for Oracle’s competitors to spread disinformation. Here’s some clarification.

The real Fusion strategy

AMR Research has spent some time recently looking at how the real Fusion application strategy has evolved and at what Oracle’s development team has been working on. When we stripped away all the technology jargon, marketing spin, and competitive posturing, we found that there was a rather clear and straightforward plan to create a next-generation, service-oriented application suite.

The new product will be based on E-Business Suite (EBS), but it will incorporate functionality, technology, and usability features from the Siebel, PeopleSoft, and J.D. Edwards products. Contrary to popular opinion, it will not be built by merging the code from those products. Oracle has consistently said that it would not combine code lines, but unfortunately the name “fusion” seems to suggest this blending of ingredients, and competitors have been happy to perpetuate the misconception. The transformation of EBS to this next-generation, service-oriented architecture (SOA)-based product line is underway, but it will be a work in progress and take several years to complete. 

We believe that the EBS release 12.0, due out at the end of 2006, will be the last release under the E-Business Suite name. We expect Oracle will unveil its next-generation ERP product in 2007, but perhaps not under the name Fusion since the middleware suite is already using that brand. The EBS 12.0 release will use the Fusion middleware stack. It will feature some aspects of Oracle’s Fusion Architecture, including additional web services, better standards adherence, and far more Java-based code. Each successive release of the next-generation product will move the product further down this architectural and technical path, but will also incorporate functionality from the acquired product lines. 

Continued development

While EBS is evolving into the next-generation ERP suite, the Fusion strategy also calls for continued development in the PeopleSoft, Siebel, and J.D. Edwards product lines. These products will see periodic functional enhancements, the incorporation of Fusion technology, and undoubtedly the development of tools that will assist with migration to the next-generation product and/or integration with other applications within Oracle’s ever-expanding portfolio.

At the moment, Oracle seems to be wisely leaving its newly acquired banking and retail application out of the Fusion strategy discussions. We have no doubt that it will eventually be migrated to the new architecture, but Oracle doesn’t need that kind of distraction right now.

Slow pace OK with users

The slow development shouldn’t be a problem. The vast majority of enterprise application customers and prospects are in no particular hurry for next-generation products. They like what they hear about the potential for SOA to improve flexibility, but most are rightly skeptical about new technology.

When we talk to Oracle application customers, whether newly acquired or long-time EBS users, they seem relieved that Oracle is taking an evolutionary approach to developing a new product line and happy that a substantial percentage of the development resources are still being invested in the products they currently run.

Many of these companies still remember the very painful transition from host-based to client/server applications in the early 1990s, and they are in no hurry to repeat that experience. The good news is that the shift to SOA does not require the kind of complete product rewrite that client/server entailed, but there is still the potential for lots of nasty quality and performance surprises (see the AMR Research Alert article “Avoid the Mistakes of Client/Server When Adopting SOA”). Oracle is smart to have delivered a reassuring message about the long-term support and enhancement of the existing products while positioning the migration to the next-generation product as entirely optional.

Strategy is sound, marketing needs work

As usual, Oracle has done a much better job with its product strategy and development than it has with its marketing. A very disciplined and effective process seems to be in place to evaluate and compare the various product lines, sort out customer and market priorities, and allocate development resources to the appropriate projects. At the engineering level, the collaboration among the various acquired teams seems to be remarkably good.

Unfortunately, many customers and most potential customers still don’t understand Fusion. The marketing has tended to be a combination of grandiose statements like “we’re more than halfway to Fusion” and thousands of pages of dense whitepapers filled with technology jargon. Where the Oracle executives have gotten out and talked to individual customers and small groups, the strategy seems to have played quite well. But for the majority of customers, Fusion still seems to cause more concern than enthusiasm.