Considering the downturn, there are new things that are killing SOA, and perhaps a few things that have been killing SOA all along. Keep up with me on this one.
Some venture capitalists (VCs) are killing SOA. Money is tight, and some VCs have taken to micro-managing some smaller SOA technology companies in their portfolios. Unfortunately, they don't understand the market and are making the wrong calls in many instances. What seemed cool at the last cocktail party is perhaps not the right thing to do within a particular SOA technology company. I'm seeing all sorts of silly things out there now, including some SOA companies that are in the process of closing their doors.
[ Learn more about SOA deployments: What actually works. ]
Sarbanes-Oxley is killing SOA. SOA technology companies need access to capital. Either you do that through VC, or you do so through the public markets. However, going public is not advisable in most instances now that they regulate the crap out of you. The Sarbanes-Oxley reporting rules are expensive and counterproductive. Thus, most SOA technology companies stay with the private equity players or don't raise money at all. This has two negative effects. First, as private citizens we're not given the opportunity to invest in emerging SOA technology players. Second, the emerging SOA technology companies don't have access to the same pubic markets as their much larger publicly traded competition. Kill SOX, and the technology market will boom.
Big consulting is killing SOA. Here we go again, but the problem continues. The larger consulting players that are typically systemic within most global 2000 and larger government organizations are running off cliffs with SOA on a daily basis. Bad advice, vendor-driven architecture, lack of SOA skills, and lack of an overall path to SOA are killing many a SOA project out there, and in most instances, with some better guidance, it did not have to be that way. The larger consulting organizations lead with the capable guys, and then drop off the kids to actually do the real work. The larger SOA consulting players need to invest in training, get a mentor, or stop playing the game.
The lack of SOA skills is killing SOA. SOA is not development, nor is it traditional enterprise architecture; it's, well, SOA. Thus you can't do SOA unless you have the knowledge around the proper approaches, methodologies, and right enabling technology and standards. Most that do SOA don't, and those folks fail. The lack of SOA talent is clearly killing SOA.
SOA-in-a-box is killing SOA. Seems that the vendors are not hyping this as much as they have been, but the whole SOA-in-a-box concept is still out there. Technology is never a SOA solution; it's a means to implement architecture. Indeed, your architecture should be independent of the enabling technology and standards, and those that lead with technology, typically are not doing architecture and end up with a patch more so than an architecture. Sorry, there is no SOA-in-a-box.