The road to Web services is paved with potential opportunities, not the least of which is improved collaboration with customers, partners and suppliers. But the deafening hype combined with tall tales can make it difficult to separate fact from fiction.

It's becoming the El Dorado of IT - the vision of the real-time enterprise (RTE), where systems communicate seamlessly and quickly and organisations extract staggering efficiencies and deliver high levels of service. The notion itself may be neither radically innovative nor groundbreaking, but proponents are finding that achieving the RTE is proving to be almost as elusive as it would be desirable, and the dream of a real-time business remains just that - a dream.

Analysts say operational inefficiencies are largely responsible for preventing companies from achieving this vision. Now, a new study from the Sand Hill Group shows how companies can use Web services technology and services-oriented architectures to speed progress towards the RTE.

M R Rangaswami, co-founder of the Sand Hill Group, says his firm conducted the study in order to gain a sense of the process through which companies implement Web services technology, and to learn about those implementations. To this end it reviewed 60 Web services projects and interviewed 117 enterprise software decision makers (distilling more than 200 hours of conversations) made up of Global 2000 corporate customers, application vendors, EAI vendors, platform vendors and systems integrators.

Rangaswami says not only did the study show companies are gaining significant value from a strategy that experiments with Web services without breaking the bank, it reveals that despite conventional wisdom, many of the most successful applications of this new technology take place outside the corporate firewall, with fully one-third of all projects in the study externally focused.

"There's been a lot of concern about Web services being used outside the firewall, and we think that is not necessarily justified," he says. "We found a third of the projects being done were external projects, which connected suppliers and customers, and even though security was an issue, people found workarounds. They were definitely finding use for it outside, if the business case was justified.

The study also identified a potential emerging market for what Sand Hill Group calls "EAI Lite" for internal, and "B2B Lite" for external use, based on standards and open systems. Rangaswami says 44 per cent of projects surveyed were either small EAI Lite or B2B Lite projects connecting just a few systems as opposed to trying to connect multiple systems, and mainly using the proprietary J2EE or .Net languages.

The research has shown Web services are going to become, if not a market unto itself, a technology that will become part of the fabric of the organisation, he says.

Business Drives

The report notes confusion surrounds the term "Web services" even among experienced technologists. It defines Web services not as a Web application but as a new way to integrate enterprise applications that is faster, easier and less expensive than traditional integration technologies.

"Basically, it involves breaking down a large application into smaller components and making these components serviceable, so that other applications can request a 'service' without knowing the system's innards. These service requests take place using open, Web-based standards, which today include the Simple Object Access Protocol (SOAP) and Web Services Description Language," the report says.

"But our study at the Sand Hill Group identified one major distinction: successful Web services implementations are strictly business-driven. Enterprise customers say the technology will survive because it makes a business impact, rather than being adopted for technology's sake. The pioneering companies we interviewed are realising genuine business benefits and plan to continue on the road to adopting Web services."

Web services projects have already delivered meaningful business value to many leading enterprises. Although plenty of organisations are still gathering information, evaluating or prototyping, the leading organisations are already developing production applications or are already relying on Web services to help run the businesses. Most vendors are already investing heavily in the technology, and leading businesses see the technology as capable of helping them increase profitability, reduce operating costs and improve customer or channel relations.

The Sand Hill Group found pioneering Web services enterprises were very positive about the technology. "The earliest customers said they have Web services projects in production and are realising significant benefits," the report notes. "Further, Web services projects are mushrooming across these enterprises as organisations learn from their early successes." Surprisingly, the study did not find tales of failed or significantly delayed projects, nor did it hear of unhappy customers. The participants believe that 2003 will be the year when Web Services goes mainstream and is adopted by most corporations.

There is no doubt Web services are a hot topic for debate. A recent Cutter Edge report from the Cutter Consortium notes that while there is plenty of hype about the issue, there has been a dearth of accurate data about what - if anything - organisations are actually doing with Web services.

Cutter tried to find out with its own survey in early 2003. Its report - based on 240 responses - found:

• Adoption is proceeding rapidly, with nearly two-thirds of respondents already writing software - whether prototype or production.
• Twice as many organisations are Web services providers as are Web services consumers.
• Significantly more Web services are being used inside corporate firewalls than outside. Nevertheless, more than half of all respondents are providing Web services outside the firewall, while nearly half are consuming external Web services.
• More of the companies in the sample are members of older standards bodies - such as the World Wide Web Consortium (W3C), the Object Management Group (OMG) and Java Community Process (JCP) - than of OASIS, RosettaNet and the Web Services Interoperability Organisation (WS-I). This may be simply because there has been more time to join the longer-established groups or because Web services are still being used in relatively simple ways.