SOA adoption: who pays? who reuses?

Listen to many vendors and analysts, and you can be forgiven for thinking that everyone and their dog is now service-enabled. 

Last year, Yankee Group predicted that more than 70 percent of companies from many industries would be immersed in SOA by this time. AberdeenGroup says its more like 90 percent -- the other week, the analyst firm issued a survey report that said at least 90 percent of companies are on their way to SOA -- "nine of every 10 companies are adopting or have adopted service-oriented architectures and will exit 2006 with SOA planning, design, and programming experience." Evans Data also released survey results, but concludes that only 24 percent of companies have what they consider to be functioning SOAs -- but this is almost double from a year ago.

These statistics vary wildly, and lead one to conclude that SOA must be in the eye of the beholder, because one company's "SOA" may be, in someone else's view, a Spaghetti-Oriented Architecture or "JBOWS" (Just a bunch of Web services). 

But just as likely, one of the inherent properties of SOA is that is not a finished project that gets wrapped up in a bow after six months with its builders and proponents moving on to something else. SOA is an architecture, and architectures are ongoing, living, breathing entities (carrying on at least six of the seven life processes). There is no definitive end point, nor should there be. (Though David Linthicum asked a good question, 'how do you know when you're done?', pointing out that there may be a point where an SOA begins to show diminishing returns.)

A new qualitative research report out of Saugatuck Research questions how far along we really are with SOA. (Thanks to reader BCohen for pointing to this new report.) Saugatuck calls SOA adoption "modest" at this point,referring back to predictions just 18 months ago which pointed to strong uptake of SOA methodologies and related technologies.

Saugatuck conducted "more than a dozen deep-dive interviews," and came to the following conclusions:

• "Most firms who are deploying SOA are focused around two stages – either early stage planning, and/or trial SOA deployment focused around legacy application integration."

• "Sharing computing resources will require a change in the way line-of-business managers view, and use, IT."
• "Funding the upfront investment required for enterprise-wide SOA deployment is a key concern and potential inhibitor to SOA adoption."

Mark Koenig, author of the Saugatuck report, notes that most companies are in the early proof stages of SOA. More than two-thirds of the firms it has talked to "are still in planning or piloting stages of SOA deployment." Few, if any, of the companies it looked at "are leveraging the full technology and business benefits of SOA code or component re-use, or fully leveraging SOA as a new paradigm of designing, developing and deploying new applications."

In a survey conducted by WebServices.Org last year, we explored the depth and breadth of service reuse across 1,000 enterprises, and found only 21 percent were at a stage in which they were actually sharing services across two or more business units.

Saugatuck adds that the companies they spoke with "are focusing primarily on application and data integration – and for the most part are using SOA to help modernize legacy back-office applications in finance and accounting, and human resources. The overarching strategy around focusing on integration is to demonstrate initial success and ROI while creating reusable components for future SOA initiatives."

Saugatuck also tackles the up-front investment question, noting that funding SOA “start-up costs” is a major an area of concern. "Initial SOA deployments are typically 50%-to-100% more costly than traditional methods, especially at the enterprise level." 

In a previous post on this Weblog, I asked, who will be the first business unit that lays out the investment that will likely bring more benefits to others than itself? SOA is a cross-enterprise initiative that can benefit all business units that come into contact with it. Who’s going to foot the initial bill to help out the rest of the organization? When you’re talking about building an inventory of components that will be made available to any business unit that needs it, how will that business unit be made to pay for the service? Who will pay for its creation in the first place?

To get these investments going, Koenig suggests that education be provided to line-of-business executives, especially that while SOA deployments may take longer and be more costly up front, returns are far greater in the long run. Koenig also suggests that vendors provide financing to get projects started.