Information Week recently surveyed 270 business technology professionals about the state of service-oriented architecture, and their assessment of the situation is that, while SOA is experiencing trouble, it's also not dead yet.
The whole is SOA dead or not discussion reminds me of that oft-quoted scene from "Monty Python and the Holy Grail," wherein a customer tries to toss a sick, but alive, person on the dead cart during the Plague.
Analyst: Here's one.
SOA: I'm not dead. I'm getting better.
Analyst: No, you're not. You'll be stone dead in a moment.
SOA: I think I'll go for a walk. I feel happy! I feel happy!
Analyst: You're not fooling anyone, you know.
It almost makes you wish someone would put SOA out of its misery.
But, of course, SOA isn't alive in fact, and so it's not going to die. It's an architectural practice and, while companies may be a bit skittish about it-only 23 percent said their organization had deployed SOA-it's still an option and companies are still pursuing it. In fact, another 29 percent said they were experimenting or in development with SOA, according to the Information Week article explaining the survey results.
Of course, when Anne Thomas Manes of the Burton Group declared SOA dead, she didn't mean SOA was without value. When I interviewed Manes in January, she explained that she meant SOA was dead as a means to a blank check from the business.
In this week's podcast, SOA consultant and expert David Linthicum reiterated this point and gave his own assessment of SOA's health:
I don't think SOA is ever going to die, to be clear. It's a core architectural pattern, it's going to have a great value in the world of enterprise architecture going forward and it's always going to be part of that. It has been before SOA appeared as a buzz word and it will continue after that, because it just makes sense to leverage that kind of an architectural patterns to solve some of the IT issues within your organization.
In fact, Linthicum said he's actually seeing a renewed interest in SOA this month. He noted the renewed interests differs slightly than in previous years in a few interesting ways:
Now, that last one is intriguing. The interest is coming from the top - exactly the opposite of what you would expect if SOA is really dead as a business initiative.
And yet: Companies aren't quite happy with the results of SOA. Information Week points out that the percent of overall software reuse is only marginally higher after SOA-a 39 percent reuse rate after SOA versus the pre-SOA 32 percent reuse. And while 41 percent surveyed said SOA performed as expected, 34 percent still weren't far enough along to to deploy production applications with SOA yet. A Burton Group report estimated that only 20 percent of organizations are seeing the anticipated benefits from enterprise SOA initiatives, according to this ebizQ article.
Really, the question isn't whether SOA is dead or alive. That was just a fun distraction. The question continues to be: What makes some companies succeed with SOA while others falter or fail?
That's the very question Sandy Rogers, program director for SOA, Web services and Integration research at IDC, has been researching recently.
Rogers was featured in last week's InfoWorld SOA Report podcast. Her research focused on organizations that are succeeding with SOA and offers a different perspective on the issue.
She found people and "management dimensions" are the most important criteria for success with SOA, and explained how that can be divided up into seven, inter-related themes:
According to Rogers, those who succeed with SOA treat it as a business initiative:
SOA is really a business discipline and the most successful individuals and enterprises adopting SOA treated it as a business initiative and not necessarily just a technology approach and that means getting not only their particular team involved in trying to align to the business, but it's really about bringing all different stakeholders and getting their involvement as soon as possible in both understanding and laying out the vision of where SOA and what SOA can bring to their enterprise.
That sounds like to me we're back to involving the business in SOA. This has been a long-standing debate about SOA-do you talk to the business about SOA or not. Of course, as Manes pointed out, just because you're doing SOA doesn't mean you have to "talk SOA" to the business. You can talk about other, more tangible business benefits, knowing all along SOA is the engine that will achieve those benefits.
Still, I wish Rogers had talked more about how those successful SOA companies involved the business in SOA.
At any rate, SOA's still alive and kicking, even if it looks a bit grim. As Linthicum noted in his podcast:
So the prediction of SOA is neither rosy nor dismal. I think service-oriented architecture is going to continue to add value within enterprise architecture and people are going to continue to practice it. I don't think it's going to be a big focus, like it was in 2005, 2006, and 2007. The hype is going to hopefully calm down around service-oriented architecture and it's going to bake into enterprise architecture and it's going to make enterprise architecture even that much more effective.
I think we'd all be ready for that eventuality.