There is a disconnect between organizations that respond to surveys indicating that they are adopting service-oriented...
architecture (SOA) and those that are successfully doing SOA, says Anne Thomas Manes, vice president and research director for Burton Group Inc.
The difference between SOA adoption rates and actual success in achieving the SOA goal of having business people be active participants in the process has caused a debate in the blogsphere. An analogy for the difference between adoption and success is the oft-repeated story of government bodies adopting an annual budget only to report 12 months later that they are running a sizeable deficit. Adoption does not equal success.
"Yes -- that's exactly it," replied Manes when SearchSOA asked if the analogy fit. "A remarkable number of companies are 'adopting,' but typically only from a technical perspective."
She outlined a "common scenario" where SOA adoption doesn't lead to success:
- An IT group launches the SOA initiative.
- It executes 1-3 pilots to prove that the technology works.
- It deploys an SOA infrastructure, including ESBs, WSM, XML gateways, and/or registry.
- It may develop a knowledge center of best practices.
- It may set up a repository and some preliminary governance processes.
- And the initiative stalls because the IT group can't get the business to engage.
The problem at this point is cultural rather than technical, Manes explained. IT runs into resistance from the business side.
"The business units don't want to become service providers and they don't want to be service consumers if the services are provided by a different business unit," she said. "In some cases the business units may build lots of services, but they are one-off services, used in one application, or used to build one point-to-point connection."
At this point what has been accomplished is closer to old fashioned application integration rather than SOA.
As Manes puts it: "These folks are doing service-oriented integration (SOI) rather than SOA."
This adoption versus success conundrum was raised by Manes earlier this month in her blog about the difficulty of finding real SOA success stories, not just integration projects.
Based on eight interviews she did with organizations that had adopted SOA, she said she only found one where the promise of SOA was being realized.
Referencing a recent article by Ron Schmelzer, senior analyst for ZapThink LLC., she noted that the other seven organizations were doing versions of 1990s enterprise application integration (EAI), which Schmelzer labels EAI 2.0 as opposed to SOA.
Of EAI 2.0, she wrote: "It's a better form of EAI than in the past, but it's still fundamentally focused on integrating application silos (EAI) rather than dismantling the silos (SOA)."
Manes got other SOA bloggers' attention when she wrote: "It has become clear to me that SOA is not working in most organizations."
As analyst Joe McKendrick noted on his ZDNet blog: "When Anne talks, people listen."
McKendrick's take on the situation is that it may just be too early in the brief history of SOA for implementations to achieve much more than tactical successes.
"Anne's conclusions fly in the face of other studies and vendor case studies that say companies are seeing some success from their SOA efforts," McKendrick notes. "However, these studies and case studies are projecting tactical successes (such as cutting development time for certain types of interfaces for specific functions), versus more strategic successes that impact the business as a whole."
In McKendrick's view it will take time to educate the business side to see the value of participating in the larger vision of what SOA could become.
But in an Application Development Trends roundup of recent blogsphere commentary, editor Becky Nagel found a more negative take on the business-side view of SOA from Eric Roch, chief technologist with consulting firm Perficient Inc. In his blog he says that business strategy experts who are listened to by CEOs and CIOs are advancing a skeptical or even negative take on SOA.
Roch writes that while architects talk technology, "business experts are telling our CEOs that the technology is not going to solve any of their problems!"
This reflects the troubling aspect of the problem of realizing SOA's potential, which Manes summarized for SearchSOA: "The impediments are cultural, not technical."