SOA in a Recession?

Over the past week, my colleagues Joe McKendrick and David Linthicum have wondered aloud about prospects for SOA projects as the economy hits a possible recession and IT budgets head for certain cuts. In fact, all this kicked off with a post from SOA Consortium director Brenda Michelson that consortium members remained optimistic that their SOA projects would survive the cut this year.

Unless you’re living under a rock or a trust fund, it’s hard not to ignore the warning signs of doom and gloom. Ironically, forecasts for IT budget cuts come after a year when the IT sector finally got its budgetary head firmly above the waterline since 2001.

We had a chance to catch up with Linthicum and McKendrick and discuss this very point during a prep call for an upcoming panel session at the next Open Group Enterprise Architects Practitioners Conference. We concurred that “special” (read: strategic or architectural) projects are likely to take a back seat to projects that are tactical or aimed at simply keeping the lights on. That’s exactly what transpired during previous recessions of the early 1990s and 2000s.

Our take on the SOA Consortium’s grass roots optimism is that the sample is a bit self-selected: these are companies that are likely further along in SOA adoption, and therefore more likely to budget staff time to participating in such a group. And once a set of practice or architecture gets rooted, it no longer fits the definition of “special project,” which is Linthicum’s contention.

Nonetheless, if there is a 2008 recession, for IT organizations it will likely not be as severe as that of 2001 because there is a heckuva lot less bloat to cut. A more apt comparison might be the early 1990s when the term “downsizing” and “outsourcing” first entered the vocabulary. Yeah, that’s still happening today.

But more to the point, the recession of 1990s occurred just as IT was on the verge of another great architectural migration, in that case to client/server. While investments in client/server were obviously slow back at that time (we had a hard time coming up with real-life installations while writing for Computerworld back then), when dollars flowed back several years later, they did so with a vengeance. And that was before IT organizations got caught up in the Y2K and Internet booms. OK, we don’t expect SOA to explode in the same way as its predecessors, but that’s because it’s not 1995 anymore. SOA projects for non-early adopters will likely slow this year, but they won’t get permanently derailed.

We’ll be plumbing this and related topics further as we join Linthicum and McKendrick on a SOA Reality Check panel moderated by InfoWorld editor-in-chief Eric Knorr at the Open Group’s 18th EA Practitioner’s later this month in San Francisco.