For the same reason that political parties hold their rallies around this time of year (squeezing them in between the end of the holiday season and the start of the Christmas wind-down), this is also the peak time of year for vendors to hold their user conferences, and to brief analysts on their past successes and future strategies.
Having been briefed by a significant number of SOA technology vendors, it is apparent that there is a remarkable consistency in the messages now being sent to the market: if you want to do SOA (and the implicit assumption here is that everyone wants to do SOA), then you need to decide how you are going to implement governance over the SOA implementation in your organisation.
This will not come as much of a surprise to any of the analysts attending these briefings. Again, there has been considerable consistency between rival analyst organisations in pointing out the sort of mess that will be inevitable if governance is left as an afterthought.
However, it is a very different message to the one being expounded by vendors at the beginning of the SOA wave. Then, the message was that once you have 'wrappered' legacy applications to expose the services, and provided a means of orchestrating the flow, you would then have a rosy future recombining all of the services in innovative and exciting ways to the delight of the business world.
Of course, there is an element of truth in both of these message styles. There is an achievable destination of recombination and re-use of services that is both desirable and rewarding, but also there is a deep and ugly pit awaiting those that take the short cut, ignoring the toll-road marked 'governance'. But really we have known this all along. The early messaging was designed to get organisations started down the SOA road, and recognising some early benefits before they had to commit to serious expenditure.
However, another consistent message, coming from case studies of user organisations that have already delivered successful SOA projects is 'I wish we had started SOA governance earlier'.
Now it could be seen that this change of messaging is just reflecting the new-found wisdom that governance should be addressed early on in an SOA deployment. It could also be seen in the light that until recently several substantial vendors really did not have much to offer in the way of technology to support SOA governance.
Again, it could also be seen as a way of differentiating between those vendors that just provide a run-time platform for SOA, and those that can provide a soup-to-nuts service.
The market certainly needs ways of differentiating between offerings.
It is quite easy to arrive at a list of 25 or more vendors that can each provide a viable run-time platform for SOA, and each of them will have an area of strength. As with many situations, the initial decision is likely to be whether to assemble a solution from best-of-breed components or to buy a ready-made suite. Before the larger SOA platform vendors fleshed out their suites by adding governance capabilities, the only real option was to partner with one of the much smaller number of SOA governance specialist vendors.
This would have had the effect of a small number of point-solution vendors having a disproportionate impact on the evolution of the market. So it has been convenient, in some cases, to let the inconvenient issue of governance remain in the background.
The market has now matured to the extent that most of the larger vendors can provide both the run-time platform and viable governance tools. This should be of great benefit to those getting started in SOA, provided that the entry-point can be kept cost effective for a low-impact project.
It is inevitable that there will be a great consolidation in the number of vendors operating in the SOA space. In order to retain a long-term future, vendors will need to be able to provide a full solution, either with a single-integrated suite, or with a well-integrated set of partner offerings. By well-integrated, this must mean much more than the simple capability for the products to interoperate, and should include as a minimum, integrated development and management interfaces and the consistent use of a shared repository.
The next 12 months will be a telling time for SOA technology vendors.
It is likely to be much more difficult to reach a total of 25 or more platform vendors, in fact it is likely to be difficult to find 15 that have a truly viable revenue stream from providing SOA platform technology. Ultimately, markets seem to consolidate down to four or five vendors that between them account for 80% of the market revenue, with no other vendor having more than a 2% market share. This might be adequate for a smaller vendor to keep afloat, but will not permit the growth levels needed to make it immune from market setbacks.
Butler Group is a ComputerWire sister company and part of the Datamonitor Group. Find them at www.butlergroup.com