One of the missing ingredients in every vendor's push toward a service-oriented architecture has been sufficient proof that the concept really works. That proof could take many forms, but one of the most compelling would be a critical mass of composite applications that leverage the integration technology and plug-and-play features of service architectures. There are lots of definitions of what quantity constitutes a critical mass. Yet it's clear that by any definition there just haven't been enough composite applications -- packaged or customized -- to make the SOA dream a reality.
There's a lot more that needs to take place before the collective visions of SOA, composites, and the process-driven enterprise now being promulgated by the leading enterprise software vendors really take hold. (For more on these issues, see Dave Brousell's Take 1 column)
That said, I'm happy to report that 2006 will be the year of the composite application -- or at least the year when software vendors start to offer a lot of new applications that are either packaged composite applications or the tools to make them a reality. Whether customers line up to buy these applications will depend on many factors, from how functionally compelling they are to how well developed the supporting service architectures are. But, at a minimum, there's now going to be something to line up for.
Two things are happening that promise to make 2006 the beginning of the composite era, an era that is starting off very much in SAP's favor. The first is that SAP's plans for building its ecosystem are now coming to fruition. In addition to its own efforts, SAP now has over 1,200 partners building composites or supporting those who do. The count for composite applications that run on NetWeaver should hit 1,000 by the end of 2006, SAP expects.
This represents an important proof of concept for SAP, which has been a leader in the composite applications market that it helped start with its initial xApps. And it opens the door for other vendors to follow in SAP's footsteps. If SAP can make it safe for customers to buy composite applications, it will be a lot easier for the rest of the vendor community to put more effort behind composites.
The second major factor is that end-users are starting to see the value of the service architectures on which composites can be based -- and implementing these architectures themselves. SAP now has over 6,000 deployments of its NetWeaver service architecture under its belt, and counts an average of five NetWeaver go-lives every day. While NetWeaver is not a prerequisite for implementing a composite application, using it as the platform for one helps ensure that the application will take full advantage of the service architecture's functionality.
What will the software market look like with a thousand or more composite applications available to buy? It promises to provide innovation at a lower price point. That innovation will be more industry-specific, and, within industries, more tailored to customer needs than before. It will be a market with a lower time to value, a lower total cost of ownership, and a much better return on investment.
Of course, there will be a few gotchas -- you'll need a NetWeaver or its equivalent to get the best bang for your composite buck. Yet, if 2006 lives up to its promise, the year of the composite will be everything we've wanted in enterprise software, but have been hard-pressed to find. Until now.