Proceed With Caution

The Sand Hill Group report finds Web services are usually implemented in three stages:

Educate: Most companies conducted pilots and trials to become familiar with the technology and to maximise learning about its advantages and disadvantages and the business problems it can address.

Power the engine of value: Conduct small, focused implementations that address a specific business goal. These discrete projects will combine to serve as the engine of value from Web services.

Realise global benefits: As more applications become enabled by Web services through completion of discrete projects, developing new applications and integration points takes less time, effort and money. These benefits accumulate quickly across the company, making its IT systems faster and more flexible.

The report finds companies planning a Web services journey can improve their chances of success by following in the footsteps of their pioneering peers. And getting started is relatively easy. It begins with the education phase then moves on to IT managers working closely with business executives to identify business cases of inefficiency around the company that Web services technology might address.

"The marketing group at one major consumer packaged-goods company asked for IT's help in improving application integration with its ad agencies," the report says. "The collaboration tended to be completely ad hoc and haphazard, with several constituencies on either end exchanging information, creative materials - such as photographs, copy and video clips - ideas and schedules. The Web-services solution first laid a formal message structure to capture these interactions and helped to streamline collaboration by propagating relevant information to interested parties. The benefits included faster time to market with marketing campaigns and lower clerical costs needed to support interactions with the ad agencies."

Rangaswami says it is vital for organisations to have a business case before proceeding to Web services adoption. He says the 60 case studies were nearly all being done by the line-of-business executives, as opposed to the IT department.

"So our 60 case studies show that IT departments usually are involved in pilots, and implementations have been for the most part driven by the business side of the house," he says. "This is a very big finding for us: that you need IT to help you with the pilot projects, but justifying the full-blown implementations is definitely led by the line-of-business or the business executives."

The business case proved critical for the 60 Web services projects in the Sand Hill study. Most aimed to solve business problems: 75 per cent were aimed at increasing operating profitability, and the remaining quarter focused on reducing IT's total cost of ownership. Business goals included reducing operating costs (38 per cent) and increasing revenue (22 per cent).

"More than half of the projects with business goals were aimed at improving customer or channel relations," the report says. "One government health-care provider with 1000 points of care, each with its own home-grown system, used Web services to get a unified view of its customers and present them with a uniform experience.

"Projects that aim to lower operating costs involved a variety of objectives. Some tried to provide a unified view of parts and finished goods inventories across plants and/or divisions. This increases negotiating power, reduces parts duplication and makes it easier to shift parts and products from one plant to another as market demand changes. Giving employees self-service benefits administration was also popular."

And since most customer data is distributed in multiple data sources across many access points, the majority of projects were aimed at customers.