One of the challenges in working with companies in their early explorations into new technologies and methodologies is the inevitable backlash that occurs regarding change. Many have argued that at best, SOA brings nothing new to the table, and at worst it will fail to achieve the desired results for the enterprise (i.e. agility, reuse, interoperability, etc.) just as previous attempts such as OO, CORBA, and Web services have failed.

While I certainly believe that SOA is more of an evolution than a revolution, there are several key aspects that give me hope with respect to success for the service orientation of the enterprise.

• Business alignment — While previous attempts around system integration and componentization have had some success at achieving improved capabilities at a very low-level, many failed to achieve any sort of sustainable ROI. A principal reason for this was the lack of business alignment. Developing more flexible, interoperable, and reusable capabilities is great, but if they are not aligned with the business and where it is headed strategically, then failure is inevitable.
• Runtime governance — We’ve made a lot of strides in the areas of technology project management, management of design and user requirements, as well as configuration management of software. Unfortuantely, once we push our systems out into production, all of our plans and controls go out the window. The enablement of runtime governance of services that are managed by automatically enforceable policies is tremendous. This gives us a new dimension to controlling, influencing, and protecting the way in which enterprise components are used and represents a significant different over previous initiatives.
• Gravity — There is sometime to be said for the fact that people have really gotten behind SOA. Vendors, think tanks, standards groups, public and private sector organizations have really taken up the SOA banner and run with it. There have been plenty of good ideas in the past, but without a critical mass, they are doomed for the history books (err… blogs).
• Internet – The Internet has changed the face of computing, business, and social interaction. SOA is inherently network-aware and able to take advantage of standards-based distributed computing frameworks.

Finally, a colleague passed along a great post from IBM Fellow Kerrie Holley that describes several aspects regarding how SOA is different from previous approaches: What is Different about SOA than previous approaches?