SOA is not an all-or-nothing proposition. SOA bundles many different technologies (see the Appendix, "Standards in SOA," for a list). It may not be possible to adopt every one of them in your business. Some of the benefits you have outlined in the business case may not be achievable for technical reasons. For example, if you think you can interact with your suppliers using Web Services, you need to make sure that:
Most of your suppliers offer a Web Service today.
You are able to consume these Web Services. This sounds unnecessary, but any number of issues can make it very difficult to use a Web Service. Unless you actually try it out, you will never know.
There are mainly two ways to test whether a business goal will be technically achievable through SOA. You can build a proof-of-concept solution. Such a solution should attempt to use the actual SOA platform of your organization and the services of your partners. That is perhaps the most conclusive way. However, procuring funding for throw-away software can be difficult. Alternatively, you can perform thorough research of your business partners' SOA infrastructure and the capabilities of your own SOA platform. This will yield a "best effort" estimate of whether a specific SOA technology will work well in the context of your business.
The finding from this exercise will feed back into the business case. For failure of a test can mean an unrealized business benefit of SOA. Exhibit 5.1 shows the feedback ...