Is SOA In Trouble?

I recently got an e-mail from ZapThink expecting more of their usual excellent commentary on the service oriented architecture  (SOA) market.  What I found instead was an impassioned screed asking whether SOA would ever become the predominant approach to enterprise architecture or "succumb to the pressures of confusion, misdirection, and ignorance that assail it, and become a tired label that signifies little more than a set of product features?"

Lest you wonder what all the excitement is about, SOA is an approach to application development that delivers functions to end users not as standalone applications, but as sets of "services" (such as "look up customer record" or "calculate credit rating") that can be used by multiple applications. The main objectives is to make it easier for large organizations to share information about applications, and with their business partners, and to allow developers to re-use the code needed to perform these standard services rather than re-writing them from scratch each time they're needed.   

Suffice it to say SOA is somewhat of a niche market, so it was surprising to see a consulting firm that makes a living from that market question it's viability so openly. It was also interesting to see ZapThink's Managing Partner Jason Bloomberg (the author of the post) frankly ask readers whether ZapThink was "part of the problem" by being a cheerleader for SOA.

ZapThink is facing a classic problem that comes with marketing or understanding a new idea (SOA) rather than a standalone product such as a general ledger application or a service such as on-site PC repair. All the problems ZapThink raises about SOA -- CIOs don't understand it, consulting firms lack the skills to implement it, entrenched vendors lie about how their existing products work with it -- are endemic to writing about or marketing an attempted conceptual breakthrough like SOA. 

What's an analyst (or a PR or a marketing firm) to do? Stay calm and keep your eye on the ball. Where the ideas behind SOA have value (the flexibility of exposing application services over the network, software reuse, open "directories" of services and data anyone can use) applaud them. Where products from newcomers or legacy players make it easier to reach the goals of SOA, praise them. When vendors, CIOs, consulting firms or end users misunderstand or distort the SOA message, correct them.

But remember that you're primarily talking about a set of ideas and principles. Even if SOA fails, the best of its ideas and principles will survive in the next round of technology acronyms. That's how we humans make progress. If you did a good job explaining or marketing these ideas in their SOA incarnation, you'll do fine the next time around as well.