Hold off on Web services

One of the few hot technologies in our industry today is Web services. Unfortunately, making the promise of Web services a reality will be much harder than anyone is willing to admit.

The premise of Web services is that applications should be made available to organizations as services rather than stand-alone software. The value proposition of Web services is that it would be easier to distribute application functionality across organizations if they used a common language or protocol. Making it easier to share applications makes it easier to conduct business across organizations.

The first problem with Web services is the term itself. In today's world, Web services are more likely to be sets of products than sets of services. Initial Web services implementations let companies wrap specific applications in industry-standard protocols so they can be delivered via the Web, but not in an easily automated fashion.

While the idea of establishing a common language or protocol to facilitate industrywide communications is not new or unreasonable, it represents another problem with Web services. The broader adoption of Web services depends on a new set of industry standards and products that are not going to happen overnight. With a number of megaplayers each determined to support its own flavor of Web services, and upstarts unveiling products ahead of full standards approval, finding truly standard technologies to support Web services is risky.

The third indication that the Web services market is not what it seems is that analysts, hardware and software vendors, and specialized service providers are talking about its virtues more than enterprise users and venture capitalists are. When double-digit analyst forecasts and exaggerated vendor promises outnumber customer testimonials and venture capital commitments, it's a sign of a technology still looking for a market.

The reality of the Web services market for the foreseeable future is that early adopters are more likely to use a set of products than to leverage a set of integrated services. This means they are also more likely to spend more time, money and resources developing and deploying their initial Web service implementations than they anticipated. And given the embryonic nature of Web services, there will be few consultants or systems integrators with expertise to turn to for help.

Before adopting Web services, ask yourself the following questions. First, what is the business issue you are trying to solve and can today's Web services products satisfy your business objectives? Second, how can you isolate your initial Web services deployment to test their effectiveness without jeopardizing your ongoing operations? Third, when do your business partners plan to adopt Web services?

In other words, wait until there are real standards, products and services before you make any significant investment in Web services.

Kaplan is managing director of THINKstrategies, a consultancy in Wellesley, Mass. He can be reached at jkaplan@thinkstrategies.com.