Edging into Web services

Automating the flow of information between companies can be costly and complex. But Web services--when applied correctly--offer just the ticket to help CIOs tackle the job.

The real long-term prize of any business collaboration lies in mobilizing the assets of partners to deliver more value to customers. Emerging Web services technologies will play a crucial role in facilitating such collaboration.

To extract value from these technologies, it will be necessary to make hard business decisions about the way they are used instead of merely treating them as just another new software tool--for their power lies in the ability to make the systems of trading partners interact. Another specific feature of Web services is their ability to let companies start with small, focused initiatives to achieve quick wins at minimal risk and then to build on success as they evolve.

Technologically, every company is an island, operating with its own assortment of systems, applications, databases and communications technology. Few companies have the power or the will to force their trading partners to adopt common systems.

Conceivably, a company could replicate in-house the IT platforms used by its most important trading partners and thus enjoy a seamless flow of information. But the effort required would be prohibitively expensive and complicated. In theory, all of a company?s present and potential trading partners could adopt the same enterprise application suite, but this is basically a fantasy.

The few connections that companies do make between one another, largely using electronic data interchange (EDI) networks, are of limited use. These networks provide only for the exchange of specific transaction data--for instance, invoices and requests for proposals--and can?t be used to exchange the quantity and quality of information needed for higher-value collaboration, for example, on product development. EDI connections are also expensive and inflexible. As companies forge more and more links, they will find that using EDI to connect new business partners or to shed old ones is neither quick nor cheap.

The advent of Web services promises to let a company connect its applications to any number of trading partners relatively inexpensively and easily. Of course, Web services could also be used to link applications inside the company. But these are experimental technologies and for now most companies will prefer to stick to more expensive--but proven and reliable--approaches to integrate key internal systems. The best use of Web services currently lies in applications in which connectivity problems are more complex, efficiency gains are greater in the near term, and alternative ways of connecting companies are limited.

Saving money
Web services are essentially a number of Web-based standards and protocols that enable companies to connect applications and data directly to one another. The standards can be incorporated in a layer of software--an interface--that companies put atop an existing application, thereby allowing any other application with a similar interface to link with it and exchange data.

Writing this layer of middleware is far less expensive than customized code--about $30,000 for a modest connection between two applications, according to one financial services company, compared with $800,000 for the customized version. Moreover, code rooted in a feature of an application makes for a rigid connection: If the underlying application is changed, the customized connection must also be changed or even rewritten. Web services let companies tinker with the application while avoiding changes to the interface.

The best use of Web services currently lies in applications in which connectivity problems are more complex.
What this means for business is that a company like Nike, with many versions of products and a broad range of partners, will be able to connect its own technology to that of its suppliers more efficiently, reducing the need for employees to send, receive and reenter transaction data manually. Such a company could expand the amount and kind of data it exchanges with trading partners, thus not only improving the way both sides interact and collaborate but also transforming the way they develop, make and distribute products.

By using Web services to enhance collaboration in business alliances, some companies could even expand the value of the goods and services they deliver to customers. In addition, Nike would enjoy greater flexibility, so that when fashions change the company could add new suppliers and drop others quickly and inexpensively. Similarly, it would be able to connect more readily to the large and fragmented retailer network that sells its shoes.

Aid with outsourcing
Web services can also help companies outsource business processes. Whenever a company turns over the management of a process or function--manufacturing, logistics, or human-resources management, for instance--to an outside provider, the process presents a coordination challenge.

Technology companies that outsource manufacturing, for example, must ensure a seamless transfer of data from their product engineers to the manufacturer's engineers. To help such companies, one manufacturer is launching a Web services tool that will enable both groups of engineers to start collaborating in the development of new products at an early stage.

In initial trials, the company and selected clients managed to cut design-cycle times by more than half and to tackle such time-wasting problems as a 50 percent error rate in bills of materials generated by product-development engineers. By making it possible to identify design enhancements earlier, the collaborative approach also promises to reduce manufacturing costs substantially.

This achievement is just the beginning. Managers should understand that Web services still represent more of a potential boon than a real one. Only a small number of companies, among them Dell Computer, General Motors and Merrill Lynch, are using Web services protocols to connect selected activities to suppliers, dealers or customers.

Eventually, specialized companies that offer mission-critical managed services, such as security and specialized messaging, will be needed to ensure the reliability of Web services. If collaboration is to work, moreover, companies must define the business terms they share: Will product sizes be specified in centimeters or inches, for instance, and defined as a product?s height by length by width or in some other way? Without precise definitions, the exchange of information can?t be automated.

Web services won?t remain on the edge for long. Once companies have enjoyed the short-term advantages of reducing the inefficiencies on their boundaries, they will want to integrate internal applications.

Dell is one of the companies that has already started to expand its use of Web services to core activities. In 2000, the company began sending components specifications to its suppliers in a Web services format so that the suppliers? inventory-management systems could read the data automatically. That move helped Dell reduce the inventories of components at its far-flung assembly plants by more than 80 percent?down from 26 to 30 hours of production to 3 to 5 hours.

Some elements of the Web services architecture are not in place to date, so companies must be realistic about what they can achieve.
Dell didn?t stop there. Its suppliers serve many plants, and the company decided that if it could aggregate information across them, it could match production capacity more efficiently to demand and manage supply shipments accordingly. The snag was that because of the company?s rapid growth and geographical expansion, each Dell assembly plant had implemented its own manufacturing applications and database-management systems, with no effort at coordination. Information was thus shared manually. Rather than replace systems to achieve uniformity across plants, Dell used Web services to do the job, vastly cutting logistics costs throughout the production network.

Limits still exist
When managers plan their investments in Web services, they should follow three principles--leverage existing technology, implement investments in stages, and plug in new elements of the technology over time--and carefully balance the longer-term strategic and operational advantages that Web services can support.

When a company installs an enterprise platform, the sensible course is probably to rip out legacy systems and to replace them with applications that run on the new platform. The reverse is true with Web services. Here, the new technology is essentially an overlay, so the challenge is to get more value out of existing assets. Companies should also consider the additional capabilities that Web services alone can provide.

To benefit fully from a large application, a company must generally redesign many of its core systems. With the more modest and focused Web services approach, little or no restructuring is needed and an installation can be completed in days or weeks rather than months or years. This means that not only companies but also their trading partners can invest in stages.

The fact that Web services components can be added without redesigning an entire system means that a company can adopt them without having to fear that it will be overtaken by later adopters armed with more advanced technology. Indeed, companies that quickly build the institutional skills (such as the ability to manage relations with partners) that are needed to harness Web services will probably stay ahead of the curve.

Of course, there is a flip side to this virtue: Some elements of the Web services architecture are not in place to date, so companies must be realistic about what they can achieve. Their investments should be based on an objective understanding of the current state of these technologies, and they should avoid initiatives that depend on features not yet available.

Any company tempted to fill the gaps with older technologies should be wary of creating hybrids that will limit its options when Web services alternatives become available. Proprietary extensions to fill gaps in the features of Web services, for example, should be implemented as modules with clearly defined interfaces. In this way, it will be easier to replace the proprietary extensions with evolving Web services standards as they become available.

One further word of caution: A staged, pragmatic implementation of Web services is by no means without pitfalls. Executives lulled into complacency by the simple and mundane nature of Web services and by their early tactical implementations might overlook the broader opportunities and lose valuable time. It is management?s attitude that will ultimately determine who creates value with Web services.

A new technology architecture that first emerged from the cloud of the Internet will reshape the IT platforms of businesses and lay the groundwork for entirely new business opportunities. Assertive managers who recognize the potential of this architecture will use it at the edge of their companies to realize profits today--and explosive growth tomorrow.

For more insight, go to the McKinsey Quarterly Web site.

Copyright © 1992-2002 McKinsey & Company, Inc.

Autoplay: ON Autoplay: OFF