The Aberdeen Group, which prides itself on "fact based" research (primarily conducted through surveys and benchmarks) recently released a study reporting that "nine of every 10 companies are adopting or have adopted service-oriented architectures and will exit 2006 with SOA planning, design, and programming experience."
This report was based on a survey of more than 120 IT and business professionals, seems to indicate that SOA technology is on the rise, especially among the biggest enterprises.
Yes, the title of the article is intended to be ironic. The statistic is a deliberately misinterpreted reading of the Aberdeen study. Clearly taking the 90% result as a metric of SOA "completion" rather than adoption does not do justice to the research itself. This way of looking at the problem is perhaps as worthy of skepticism as Oracle's claim that Fusion is halfway to completion. The headline is put in those terms in order to draw attention to a phenomenon that also seems to be on the rise: SOA Fatigue.
In a discussion with Carl Ververs senior enterprise architect at ThoughtWorks a global IT consultancy, InfoQ learned that a number of their customers are expressing the equivalent of "are we there yet"? Perhaps they are forgetting Judith Hurwitz's first principle of SOA--that it is a ten year journey. Although analysts like Anne Thomas Manes of The Burton Group have been good about reminding us that "when it comes right down to it, you don't need any new products to do SOA." However, the urge to follow vendor SOA hype coupled with the fact that SOA takes time and effort leads some pundits to conclude that SOA is in a trough of dissillusionment.
Getting the benefits from SOA costs money. It also requires IT organizations to take responsibility for producing business outcomes. "Redesigning business processes, high IT integration costs, and customization challenges are eating up 40% of the IT budget in integration expenditures," said Peter S. Kastner, Vice President and Research Director for Enterprise Integration at Aberdeen and author of the report. "SOA is broadly seen as a real technology step forward, with the largest companies, who have the biggest integration problems, leading the way."
The three approaches to SOA highlighted in the study are SOA Lite (lightweight open source centric "mashups"), SOA ERP (deep dive Enterprise Resource Planning integration) and Enterprise SOA.
One of the potential problems in such diverse and broad segmentation of SOA is that it could result in the "big tent" effect where many unrelated projects all are labeled SOA for the purpose of appearing on a resume or satisfying the "pointy haired boss". This relates to the antipattern called "Defensive SOA" described in the InfoQ SOA Antipatterns article.
To clarify, this article is not a criticism of the original research report, which is factual in nature. It merely cautions against jumping to conclusions about the state of SOA and the progress organizations are making towards its goals and benefits.
AberdeenGroup, Inc. provides fact-based research and insights focused on the global, technology-driven value chain. Aberdeen’s benchmarking, market and solution assessments, sales acceleration programs, and conferences support Global 5000 value chain and technology executives -- and the solution providers who serve them.
