As you might imagine I don’t spend a lot of time listening to what Gartner Group says. In fact CBDI membership registrations strongly suggest the opposite is true! However one of our members gave us feedback that at the recent Gartner conference in Cannes there was almost no coverage on SOA!
It’s not hard to figure out what’s happening. Gartner has a market maturity model that predicts every major IT trend will go through what they refer to as the “trough of disillusionment”. Gartner’s own Hype Cycle shows the Technology Trigger being followed by Peak of Inflated Expectations before Disillusionment, which may be the precursor to the Slope of Enlightenment. We might imagine they are preparing their case!
What’s really interesting here is that the Hype Cycle is specifically defined as the graphic representation of the maturity and adoption of specific technologies. Yet hopefully most people know by now that SOA is not a technology. SOA is an architectural style that is merely enabled by technology.
What’s really going on here is that many of the major vendors are realizing they have difficulty in “selling” SOA. While they all have lots of infrastructure technology to support SOA, they are having real difficulty is selling large scale SOA infrastructures or SOA projects because SOA is a very long term deal. It is an incremental implementation that is very often, particularly in the early stages, only indirectly linked to business projects and programs. So a diversion to more profitable activity would be welcomed by the vendors and make the world go round. We have all been talking about SOA for far too long!!
My own observation is that most organizations are dabbling with SOA and most of them are finding it harder than they ever realized. An interesting experience was related to me very recently where a major organization developed shared services for their core business. From a technical perspective the program was a huge success, but the IT and program managers felt unable to make any changes to the organization. Consequently even before the first stages of the program went live the shared services were being customized and replicated by individual projects. The result was that the SOA aspects of the delivery have been a complete failure.
Most enterprises don’t get this far. They are dabbling with individual projects and thinking they can evolve the architecture. But they find they never do - if the architecture and governance is not put in place up front, there is never a case to go back and retrofit.
I think to characterize this situation as disillusionment is frankly unintelligent - it smacks of marketing spin designed to attract attention. I prefer to see this situation as one of early learning. We are all learning, and sometimes we do need to make mistakes to learn. I suggest many organizations are currently deploying SOA in a tactical or applied manner. In practice there are good business gains being made, although the quality of the delivered SOA is often woefully inadequate, primarily because there is no enterprise level reference model and architecture on which good governance can be based. So individual projects are making up policy, which is something of an oxymoron.
But I am starting to discern a change happening where not just architects, because typically they have understood the need for a while, but CTOs and CIOs are starting to recognize the need for a strategic approach to SOA. The strategic approach recognizes the need for SOA capability (think CMMI) development across the spectrum of activity – addressing organizational, management, process, project issues as well as architecture and infrastructure. A well designed roadmap uses appropriate patterns to phase activity in line with business demand and a balance of all these areas of capability.
To support this we are also hearing strong interest in how to create the business case for SOA. At the recent CBDI Summit Meetings this was much interest in this topic - a sure sign that it is being taken seriously, rather than just as a technology issue.
My message to the spin masters that are talking up the trough of disillusionment and the slope of enlightenment and similar ideas would be to read John Bunyan for inspiration. The religious fervour they clearly bring to their thinking would fit better in the Slough of Despond, a deep bog in John Bunyan's novel Pilgrim's Progress, into which the character Christian sinks under the weight of his sins and his sense of guilt. To reiterate, we should not be disillusioned, even with sub-optimal SOA deliveries, they are probably important learning exercises and will allow the enterprise to move to a more manageable, repeatable SOA environment.