If you're working on a service-oriented architecture (SOA) project within your organization, the bad news is that an economic slowdown may result in budget cuts and even the cancellation of projects. The good news, according to analysts at Burton Group Inc. is that you and your SOA vision can survive if you follow a strategy they call "Think big, take small steps."
Anne Thomas Manes, director of research at Burton, laid out the problem and the solution in a Webinar this week.
While not mincing words about the perils of an economic slowdown and the pressures on SOA architects to do more with less, she said, "If you want to be successful at significantly improving the value that the software delivery organization provides to business, you're going to have to do it in slow gradual steps and you have to maintain momentum over the long-term."
For anyone who might be in denial, Manes put it bluntly: "We have an impending recession here in the United States and it is very definitely impacting the economic situation all around the world."
That can't be good, as Tony Soprano used to say. Manes didn't sugar coat what a recession might mean to IT.
"You're going to see budgets shrink," she said. "You're going to be asked to do more with less. One of the most dangerous things is we're going to see funding for major initiatives evaporate."
Other trends putting pressure on IT, Manes said, include globalization, outsourcing and auditing required by new federal government regulations. All this signals changes in the way future applications are going to be developed and delivered, she said. These changes are unavoidable, but not necessarily un-survivable.
"What this comes down to is change is mandatory," Manes said. "You cannot continue using a business as usual approach because it's just not going to survive in the future. The company is not going to survive if the company is not able to support these requirements."
Speaking to an audience of information technology pros, Manes asserted that in this difficult business environment, technology is not the answer, or at least not the whole answer.
"Being technology people we think this wonderful shiny technology is going to solve all our problems," she said. Then she offered a brief history of technologies failure to "solve all our problems.
"I was there back in 1990 when CORBA first came on the scene," she said. "We were all excited because CORBA was going to solve all the world's problems. And you know what? It didn't. And the next time around it was Web services. And you know what? That hasn't solved the world's problems either. And now, it's REST. REST is going to solve the world. But you know it doesn't work that way."
For IT professionals that want to survive and perhaps even thrive in a recession, Manes advocated giving up on the idea that technology alone is going to fix business problems.
"That's because change is a function of three different things," she said. "It's a product of people, process and technology. Unless you're willing to actually change all of those you're not going to get a different result. Of the three, technology is by far the least important. What matters is you go out and look at organizational structures, mindset, approach and your processes. That's where you are going to make changes in the kind of quality you produce."
This is not IT business as usual where the solution is to buy the latest vendor products after the salesperson promises they will turn application lead into gold. Manes was not advocating pulling the plug on old systems and replacing them with new buzzword-compliant software.
"Change is tough," she said. "You can't just throw everything out and start all over again."
The Burton strategy for getting through an economic downturn is to not to give up on the goal of reaching SOA nirvana, but don't expect to get there overnight on a shoestring budget.
"Our advice for this is think big about where you want to go, what needs improvement, identify what your priorities are, and then deliver in small incremental steps, where each step delivers real value to the business," Manes explained.
Think big, small step SOA
It probably couldn't be coming at a less inopportune time, but Chris Haddad, vice president and service director at Burton, said everybody seems to have an opinion of what constitutes SOA.
"The redefinition of the SOA landscape is beginning to take a toll," he said. "Everybody seems to have their own definition on how to go about a project."
Lack of clarity as to what SOA is and where it is going, architects and developers wonder if they should sit this dance out and wait for the next insanely great technological breakthrough to solve all their problems.
"Enterprise architects and people down in the trenches working on delivering solutions are scratching their heads and wondering does service-oriented architecture equal business process management," Haddad said. "Do I have to use the newfangled middleware that conforms to the WS-* stack if I'm going to perform a SOA project? Is an ESB required?"
As Manes argued earlier in the presentation, new technology does not always clarify the situation.Based on five years of talking to Burton clients doing SOA, Haddad said, "They're also scratching their heads and wondering where's REST and how does that fit in? Should I just skip service-oriented architecture entirely and focus on delivering RESTful services?"
After listening to clients questions, Haddad said, he tried to look at SOA from a new angle.
"For the last eight months I've been taking a step back and focusing on core design patterns that can increase the quality and value of your application portfolio, and how you deliver software assets to the end user," he said.
His conclusion was the SOA needs to get back to basics.
"If we desire to reduce complexity and therefore become a more agile IT delivery shop we need to focus on sharing and reusing the assets that exists today," Haddad said, "consolidating functionality into a smaller set of moving parts, and conforming projects to common standards."
The back to basics approach means SOA projects need to refocus on providing rich service descriptions with governance and measurable metrics, he said. However, for all the focus on the details, the Burton strategy, also sees the big picture.
"One of the major impediments to grasping the concept of service-oriented architecture is that it is not a discrete project," Haddad said.
This is where the Burton strategy of "Think Big; Take Small Steps," gets a little tricky because it requires maintaining a simultaneous vision where the big SOA picture is always in mind, even as a developer works on a small Web services project.
"When I first went out and talked to companies about their SOA initiative," Haddad said, "I wanted to talk about the technology. But that's not what is really going to improve that state of IT. We need to use SOA as a catalyst that can transform IT as we know it and force us to take a portfolio oriented view of our software assets instead of continuing to focus on single siloed projects. We also must focus on the small pragmatic steps that we can take to affect change."
To make this big picture, small project strategy work, Burton is advocating taking the smallest SOA detail seriously.
"Consider that each well-design service is a single small step towards an improved application architecture," Haddad said. "It's a grass roots approach where each project develops services that are then publicized throughout the organization and potentially reused across teams and projects."
Armed with this think big/small step philosophy, Haddad urged SOA architects not to give up on SOA in a recessionary climate.
"Don't give in to SOA fatigue," he said. "Success in your SOA adoption will require maintaining momentum over the long term."