Can IT Save SOA and Itself by Learning from Web 2.0?

Enterprise architect, business strategist and blogger Dion Hinchcliffe is one of the more significant voices in the Web 2.0 world, at least from an IT perspective. I point that out specifically because I don't want you to confuse him with the Social Media gurus who are all the rage among those prone to marketing mumbo jumbo.

No, Hinchcliffe is first a technologist, with an eye toward helping IT survive in a rapidly changing cloud world. In fact, while he did not coin the term "Web-oriented architecture," he is arguably the first to expand upon the concept in his much-referenced 2008 post, "What Is WOA? It's The Future of Service-Oriented Architecture (SOA)."

Recently, Hinchcliffe revisited the issue in two lengthy posts for eBizQ. At first glance, these pieces look like familiar ground for Hinchcliffe, since they address how SOA needs to be more like WOA. But what's new and interesting is we know more about SOA-and it's weaknesses.

In the first post-"Where is The Future of SOA Headed? Where the Web Goes," - Hinchcliffe looks at three gloomy IT trends:

1. Modern software architecture, even with SOA, isn't keeping pace with the rate of change required in most organizations. Alignment with the business has always been a challenge, but now it's "becoming an endemic problem," he writes.
2. The business just isn't consuming the services IT's offering. "Without consumption and uptake, SOA cannot access the 'return' in ROI," he notes.
3. On a related note, despite the promises and gains made with SOA, the majority of enterprise data is "submerged and inaccessible to most business users," he adds.

The rub is, business users are consuming services and finding IT services that can keep pace-they're just finding them outside internal IT, in the cloud, with Salesforce.com and Amazon Web Services. The key to finding success with SOA and the business is to become more like these Web-based companies, Hinchcliffe says.

Oddly enough, this means loosening up in some ways. For instance, internal IT divisions should support a broader range of services. Forget the whole SOAP versus REST issue-support both. And while you're at it, toss in RSS, ATOM, semantic web and Linked Data-in other words, the full service continuum. In his follow-up post, "The Services Continuum: Expanding Our Notion of SOA and Enterprise Data Services," Hinchcliffe writes:

"While we might have at most a few hundred well-defined services in a typical organization today, there are literally millions of them on the open Web and many more inside our organizations under this larger umbrella. Whether we are trying to integrate systems more quickly, discover and get access to information more effectively, or rapidly building business apps based on these services (mashups), this more expansive and heterogeneous notion of services gives us a much bigger and richer pallette to work with to meet our needs."

In other words, you've got to move to a more Web-oriented architecture, where services are offered on a more self-serve basis to a broader group, where service delivery is easy, cheap and fast, and where business users can get access to the data and supply chains, he says.

To do that, you'll need to support accessing information via the browser and even-gasp-supporting enterprise mashups and user distributable widgets, according to Hinchclif

It's an interesting thesis, and he paints a compelling picture of the direction internal IT should be moving.

But I can't help but wonder if it's realistic to expect internal IT to move in the direction of SaaS companies. So many things slow down enterprise and government IT, from HR policies to the budget cycle to internal politics. And in some ways, IT is shelter from the impact of the market in a way Web 2.0 companies are not - a fact Hinchcliffe even acknowledges in the first article.

It's hard to imagine a lithe, large IT department in these places-even if SOA and the business depends on it. But I suppose the real question isn't whether IT can change for the business or SOA, but whether internal IT can change to save itself.