I know what you're thinking: SOA hype has reached an absurd level and now someone is literally proclaiming that it will change the world - but bear with me for a minute. Anyone who has been around corporate IT for the last five years or so has seen an avalanche of development work sent offshore for two primary reasons: cheaper unit labor cost and the flat-out inability to find qualified American developers. Also, the mainstay system development model whereby business units built up app silos, which served to minimize reuse and increase integration complexity, demanded a way to deal with the cost of its own inefficiency.

Offshoring was perfect - you could justify using twice as many developers as necessary when they cost one third as much. The trend hasn't been just limited to corporate IT either. Offshore workers are handling many other business functions, though these are mostly lower-level functions like call centers and help desks.

On a human level, I think offshoring has been a great thing. It has uplifted third-world economies and given opportunity to people whose talent means more now than their location. Furthermore, it sure seems that countries that benefit from this type of globalization are more stable than those that don't. Unfortunately, the results for business have been more mixed. Poor management and architecture has lead to many project failures along the way.

Corporate IT is now entering a new era. It's hard to find a large- or medium-sized enterprise that is not at some point along the SOA adoption curve. Some have charted a course for agile business process management while others are being dragged along as their platforms and tools make the transition under them. However what makes SOA a tectonic shift in enterprise computing is that it is being driven by the business. For once the move to retool is not being driven by techies or platform vendor salesmen. Business managers have grasped that implementing SOA standards serves to manage complexity, which in turn drives cost down; however to business, SOA is more about competitive advantage through agile management of processes and rules than pure cost containment.

"Agility" in SOA terms can be best described as a measure of the cost (dollar cost and opportunity cost) of implementing changes in IT systems. The old IT release cycle that never kept up is at one end of the spectrum, and the future of near-instantaneous changes affected by business analysts without IT staff intervention is at the other. The ability to quickly add or change business processes brings companies closer to their customers, partners, employees, and anyone else with whom they choose to interact. The more agile the enterprise, the closer the bond is - and the greater the competitive advantage.

Remember that we are talking about the future. No business analyst today composes new business processes and puts them into production, but that is exactly where the art is headed. SOA standards have always promised complexity management and cost savings by relegating to the infrastructure functions that are currently developed over and over again in applications. Additionally, now we are seeing that concept taken to the nth degree with the new Service Component Architecture (SCA) standard that promises declarative, on-demand infrastructure for services. Infrastructure pieces will be boiled down to checkboxes and radio buttons on the analyst's process management session as they compose new services and rules. To conceptualize this scenario, think of the analyst enabling a service for external federation via a checkbox and then being assured that it will be governed under the company's trust and access control policies. To be sure, a great deal of effort from IT will have gone into delivering that assurance to the analyst, but all in the infrastructure - none at the level of the services themselves.

So how does this developing trend affect offshoring? It greatly diminishes the need for application developers, and instead separates corporate IT into two worlds: infrastructure management and business logic management. These are the two areas that have remained largely onshore for various risk management reasons. Instead, it seems clear that we will see a shift to higher-order functions being handled offshore - published as services and incorporated into business processes of trading partners without respect to locale. Also, remember that we are talking about corporate IT. The infrastructure elements themselves will no doubt continue to be developed in large measure outside of Europe and America as they are today.

Everything from financial services to energy trading to macadamia nut fulfillment will be published to service registries and anxiously consumed by business analysts around the world to save money on transactions or to introduce new functions. This will usher in an era where countries can trade on their intellectual capital at the service fulfillment level and usher out an era where they have to send that capital overseas or otherwise sell it piecemeal to foreign firms. Return on investments in education and technology will be much quicker in this climate, globally. Moreover, it will no doubt refocus governments on being reliable service fulfillment platforms. Too utopian you say? Possibly, but it seems clear that SOA is going to serve as a key facilitator in accelerating business globalization, and that the old offshore model will fade away.