We have SOA because we bought the products but now we need governance. Obviously the biggest problem with this is a view that SOA is something you buy. Thankfully the vendors who have been preaching this to these clients weren’t able to convince them that SOA governance is also something you simply buy. It’s critical in these situations to reinforce the idea that SOA governance, even more than SOA in general, isn’t something you buy but something you do. Growing a garden isn’t simply buying all of the right materials, but is also the effort you put into using those materials to create a beautiful garden. I think in these situations it will be important to evaluate what products and tools the client has, what capabilities those products have in relation to governance, and how those tools might be leveraged, even if only in the short-term. The reality is that if the products are already bought, the organization won’t have to go through the process of lobbying for money. Hopefully they have some product that can become a valuable part of their governance development process. As long as we can get the commitment they may need to reassess what products they may need in support of governance in the not-so-distant future, they can start doing governance today.

We need all the details of governance today! The driving force behind this motivation is that those tasked with addressing the need for SOA governance want to be able to report to their superiors that “we are gardeners because we finished the garden.” The attitude that an outside consultant will have to change is that all details have to be in place before they can claim this. An important aspect to focus on in this situation is that an element of governance is understanding how the governance mechanisms will grow over time. We can also focus on how they might be able to use the governance structures they already have in place and simply make small adaptations to adjust to SOA. If we can get these clients to focus on developing a plan for how they will develop SOA governance mechanisms, they can understand how to grow the maturity of the governance itself. Then they can say “we have SOA governance and we have a plan for making it better!”

Can you give us the governance artifacts from another company? Besides the obvious non-disclosure issues, this is the “can you give us someone else’s garden?” mentality. The thought process is that if other organizations have already gone through the process and suffered the growing pains, perhaps we can avoid the same thing by dropping in place what they did. Although templates for service contracts, policies, and methodology might provide some guidance, it’s important to focus on the fact that these will have no value until they’ve been customized for and validated on the client’s environment. As long as we can get the client to commit to asking, “would this template work for us?” we can offer some generic artifacts to start the process, like using bulbs or young plants rather than seeds.

How do we bring everyone to the same level of maturity? The issue with this question is that the client thinks they should focus on those elements that lag behind while ignoring or even holding back those elements or groups that may be more mature. The positive part here is that they recognize they’ll need to pay special attention to those groups that lag behind the broader level of maturity. Rather than trying to get the entire garden to grow at the same rate, you’ll have to get them to become experts in the dual role of growing the garden from scratch and tending the established garden. In fact, in the early stages of SOA adoption it’s probably more important to identify and leverage those groups or areas that are more mature as opportunities to be more effective with the limited resources initially available. Establishing a “selective SOA” process can help identify those areas that can provide more immediate returns, even if that only includes abstract (and more difficult to measure) benefits like “better connectivity” and “more robust architecture.” Another aspect of this approach is not to try to force SOA on those who may not see the value. If we see potential for value but those involved with the projects don’t, we should move on to the next potential valuable project rather than try to “force a fit” to SOA that carries more risk of failure, something we want to avoid early in the adoption process. Later on, when we can show off the beautiful gardens we can grow, it will be easier to turn these people into gardeners.

We think this will be part of our SOA governance, how do we make it happen? In this situation the client is presenting the SOA consultant with a specific project they would like to do in support of what they currently believe to be SOA governance. It’s important not to shoot this down unless they’re completely off base, after all, we don’t want to throw out a plant they think is beautiful simply because we don’t happen to like it. Hopefully they’ve heard enough about SOA governance that this project is something that can be an element, just not the entirety, of their SOA governance mechanisms. If we take the approach of rejecting their ideas and trying to force them down a different path, we’ll be doing the same thing we’re trying to tell them not to do for their early projects. It’s more important to focus on how to turn their project into a successful part of their early governance, hopefully also using the products they already have. It will be more important to assess the lifespan of that project as an element of their SOA governance, whether the project needs to be enhanced, phased out, or adjusted in their future governance plan, and what other projects might be good to focus on to evolve the governance structure towards further maturity.

Although these examples certainly don’t cover every situation an SOA consultant might find himself in, I think it provides several common situations that may occur in most engagements.

Too often clients see the “garden” as being an SOA-enabled enterprise but don’t know how to get there. The vendors are going to concentrate on “don’t we have great plants!” and not be as concerned with how beautiful everything looks when taken as a whole. As we help get clients started with how they’re going to become an SOA-enabled enterprise, I think it’s important to reinforce the good things that they’re already telling you and turn that into something they can “grow” now.