You're studying a whiteboard drawing of your company's information technology backbone, and your eyes are starting to glaze over. You've got a spaghetti-like maze of Internet connections and older, legacy networks and systems that fairly screams with complexity and inefficiency. You talk to a Web services vendor, and suddenly you discern a brilliant set of connections between systems that can rescue valuable information from the burial grounds of obscurity—as if all you had to do was provide simple links between systems to let them "talk" more rapidly and efficiently to each other.

Sound too good to be true? In some cases, it may be. Vendors may smooth-talk you into thinking that it's easy to administer new software known as Web services, designed to allow different computer systems to communicate seamlessly through the Web: Simply layer them over your poor, balkanized legacy systems, and suddenly everything connects. It's not easy to find anyone who will speak negatively about Web services nowadays. But buyer beware. Though many companies, from General Motors Corp. to Aetna Inc., are testing Web services software—touted by some as the next big wave of the Internet—technology analysts and companies alike are finding it's not quite ready for prime time.

Why bother with Web services at all? In many situations, the software can speed access to information that your company has bottled up in legacy systems and increase communication between the various departments of a company. Further, Web services can be fairly easy to install—if you have a newer infrastructure without layer upon layer of legacy systems. Because Web services are nimble, says Whit Andrews, a research director at Gartner Inc., companies that have small IT departments and "are aggressive at adopting new technology should absolutely deploy Web services today."

Indeed, according to IDC, the Framingham, Mass.-based technology research firm, the worldwide market for Web services is projected to rise at a torrid clip to $21 billion by 2007 from $787 million last year—despite the economic downturn. Technology consultant A.T. Kearney Inc., says 70 percent of all companies are at least experimenting with Web services, with more to follow. But to justify the investment, it's essential to know the kinds of tasks you'd like the new system to deliver: Believe it or not, many companies have charged into Web services based on the promise of simplicity and efficiency, only to discover later that a lack of strategic planning at the top was limiting success.

Which companies stand to gain the most from Web services? If you have legacy data locked up in older applications, and those applications support frequently changing business processes, then Web services are, at least, worth exploring. According to Alejandro Danylyszyn, senior manager for e-technologies integration at Deloitte Consulting, companies with complex supply chains, or which regularly hand off data between systems—financial services, energy trading, high-tech manufacturing and telecommunications, to name a few—are also ripe for Web services. However, says Danylyszyn, in order to switch over to Web services widely throughout the company, "you're going to have to solve two common problems of early technology adoption: security and performance."

Security standards—which would ensure common ways to verify the authenticity of applications that invoke Web services—are not widely supported. And "wordy" standards like XML can slow Web service performance, since long sets of instructions need to be interpreted along the way. Another drawback: Web services may not be cheap. "Like any investment in technology, you have a high [cost] curve upfront," adds Danylyszyn. "After that curve starts coming down, that's when the payback period begins. Many companies, especially in this economy, are not willing to make that hump at first." If you have to purchase a development environment and Web services platform servers, train programmers who don't know object-oriented coding, and migrate a large number of applications, Web services may require a substantial investment.

In some cases, Web services may end up being simply another much-ballyhooed technology in search of a solution. "The business vision should be driving where you need to implement Web services, instead of saying, 'Hey, there's a need for Web services here,' and trying to identify how to use it," says Chee Yuen Yap, CIO for JTC Corp., a developer of industrial real estate in Singapore.

Tell your business constituents:

We need a clear and strategic vision of the kinds of services you need.

Ask your data management guru:

How hard is it now to pull together the business data that our company needs?

Ask your development team:

What would it cost to create a Web services system that works?

This article was originally published on 04-01-2003
eWeek eWeek

Have the latest technology news and resources emailed to you everyday.