Why Your SOA Effort May Fail

I'm getting ready for my StarEast trip, but want to mention an interesting article I read this week titled, "SOA failures traced to people, process issues" on Networkworld.com (http://www.networkworld.com/news/2008/043008-interop-soa.html).

The article quotes Anne Thomas Manes of the Burton Group in response to a question about what overriding message IT executives need to hear about Service-oriented Architecture (SOA).
"The problem's not technology, Howard said. People and processes are at the heart of what's wrong with SOA as it currently exists in enterprises."

I found it interesting that in an audience of around 300 people, only 6 indicated that their SOA efforts were proceeding well.

I can add my own observation that adds support to Manes' comments, that is, people in my course on testing SOA seem to be much more interested in the technology aspects than they do the people and process aspects. This has been bothering me for some time now.

I agree that the technology seems to be progressing more than the human aspects. For example, getting the business folks to work better with the technology people is a big challenge in some companies. In fact, I think this is the big challenge of getting people to adopt agile methods as well.

The acticle goes on to state, "IT departments implement a SOA program that may be technically proficient but doesn't meet the needs of business users, Chris Howard said, noting that Burton Group is researching SOA successes and failures through interviews with IT pros and business executives at dozens of clients. Business executives often conclude that IT pros exaggerate predictions of reusability or underestimate project cost, Howard said. IT professionals are generally bad at presenting the business case for SOA, and need to get better at explaining the long-term benefits in cost and flexibility to CEOs, he said."

Interesting stuff, and it lines up with what I see as well.

Take a read and see what you think.

On a different note, yesterday (May 3) was the 9th anniversary of the F5 tornado which ripped across Oklahoma. It was a record-setting event, with sustained winds of over 318 mph and 40 deaths. There were 675 reported injuries. The damage estimate was 1.2 billion dollars. The outbreak spawned 66 tornadoes. The main tornado passed about 1 mile south of our home and was the closest I've ever been to a tornado. It was an awesome display of the fury of a tornado.

So today when the sirens sounded at noon, as they always do here on Saturdays, it brought back the feelings of taking cover and praying. (By the way, I learned that prayers get real short when an F5 tornado is bearing down on you!) That was quite a day for sure. I still use this as an example in my talk, "The Risks of Risk-Based Testing" as the kind of risk that is so far outside of the bounds that you don't even know how to plan for it.

I'll be posting all this coming week from StarEast, so stay tuned for updates!