Newsletter Sign-up

Decades ago, wine maker Paul Masson's spokesman, Orson Welles, proudly proclaimed, "We will sell no wine before its time." After all, a good wine requires more than a quality grape; it requires an adequate fermentation process, proper storage, time to age and careful bottling before the consumer can truly enjoy its exquisite flavor.

Similar to the wine-making process, Web services have needed time to age and also require careful bottling in order to effectively produce a quality result that meets the lofty promises touted by technology vendors over the past half-dozen years.

While there are as many definitions of Web services as people you ask, for the sake of this discussion, let's agree on a basic definition. Web services are a set of self-contained applications, interfaces and communication protocols developed and delivered based on a set of XML standards and agreed-upon programming rules. Theoretically, these Web services could be reused, accessed, shared and executed via the Internet when and where needed by any system capable of calling the service.

The ultimate endgame with Web services is to achieve true ubiquitous computing by using a cafeteria approach, allowing insurers to pick and choose services in order to easily assemble them into complete dynamic business processes. Early obstacles, such as Web-services security and the burden of legacy systems, combined with the growing pains common with new technologies, have limited progress to date. But careful planning and adherence to an architectural discipline will allow insurance companies to reap the Web-services rewards.

Until recently, the majority of Web services have been implemented in their simplest form with some insurers using services as a wrapper for their legacy system to expose the older technology to the Internet--a technique commonly referred to as Web-enabling. While the resulting user interface may be based on Web services, the underlying technology remains the legacy architecture with all its maintenance and integration challenges.

Web services have also been commonly used to add new functionality modules to the existing environment in order to extend legacy environments and offer services for reuse by multiple systems. Once again, new Internet-available and reusable features are a "nice to have," but the old technology continues to require the same level of attention it has in the past, with the new Web services requiring additional human resources with a proficiency in the technologies.

The most aggressive level of Web-services adoption involves insurers rearchitecting their entire technology infrastructure. Web services are ready for prime time at this extensive level of implementation, but success is highly dependent upon an insurer's commitment, careful planning and design, and well-governed development and deployment. Allowing Web services to proliferate without a well-thought-out and well-implemented plan could very well create an even greater level of inefficiency and complexity.

Factors for insurers to consider as they put their Web-services strategy into play and look to use these technologies to their full potential include:

* Integration of disparate systems

* Instant and real-time access to information

* Reuse of business functionality

* Internet-distributed services and information

INTEGRATION OF DISPARATE SYSTEMS

Lack of standards has long been a major stumbling block keeping insurers from effectively integrating disparate systems. The tight coupling required to communicate between systems was an inhibitor due to the time and cost needed to develop and maintain the necessary translators. Even with the integration in place, translation errors, and errors due to manual entries, continued to plague legacy technologies.

As an ideal integrator, Web services are able to facilitate interoperability across platforms with their open, standard protocols.

While Web services are capable of delivering on the "integrate it all" promise, insurers will initially be challenged to develop the proficiency in these newer technologies in order to effectively implement the integration points. Early on, there may be a heavier reliance on the technology vendors than insurers would prefer. However, standards bodies and integration vendors also continue to mature their offerings, which will make achieving the goal of proficiency easier over time.

Once all the systems are connected into a monolithic network of interoperability, instant and real-time access to information is a logical expectation. But, meeting this expectation is highly dependent on the design of the network. Because a user request may be traveling through a series of systems and integration points with the assembling and deconstruction of XML messages, the path the request takes through the network is critical.

Although Web services are relatively easy to build, the development teams must agree to consistency in issues such as security and authentication. Proper forethought in the network design, combined with good governance and architectural discipline, will determine the real-world definition of "instant" and "real-time."

And what about the data itself? Web services are able to support a single data source, but today most insurance companies have inconsistent data strewn across the enterprise, often in duplicate and triplicate. Investing up front to ensure quality data and reduced data duplications will be a significant challenge for many. But what's the point of instant and real-time access to data if it's the wrong data?

With Web services, insurers can effectively reuse functionality developed as a service and can call that service as needed. Reuse of functionality is not a totally new concept in that older architectures accomplished the task by requiring insurers to cut and paste code into multiple programs in order to replicate a common function. This archaic approach brought a nightmare of maintenance and revision-control issues that are virtually nonexistent in a Web-services environment.

Although Web services offer a lot of potential for a cafeteria approach to building complete business processes by assembling services in a variety of configurations, the challenge comes in determining what functions are best componentized. For example, the functionality between policy and claims is so tightly coupled today that it becomes a development challenge to analyze these functions in a "Web-services light" and redesign them into isolated and easily consumable functions. This level of design requires only the most experienced architecture to dismantle the processes into truly functional components.

With traditional legacy environments, insurers had systems that typically resided within a homogeneous corporate network, with little network reach beyond the company's geographic or organizational boundaries. The Internet and Web services have changed all that, with the distribution model now based on the reach of the Internet--virtually limitless.

This offers a plethora of potential opportunities for insurers to reach out to consumers, channel partners and trusted suppliers, but with the rewards always come the risks. Web-services security continues to offer a formidable challenge in that there is no easy method to manage security across multiple organizations, leaving Web services potentially vulnerable against cyberterrorists and hackers.

Despite all these words of warning, Web services truly are ready for prime time, but it is the companies who do the up-front design work, develop the skills and maintain the strongest governance over the development effort that will have the greatest success.

For most insurers, the risks and complexities of Web services are worth the effort given the tremendous potential. Insurers need to evaluate the risks and the benefits to make an informed decision about the extent of their Web-services implementation. They need to take into account how well it fits their business and their technical competency and how committed they are to the architectural discipline required.

Like a fine wine, Web-services technologies may have matured with age, but they are still an acquired taste.

DAVID MEYER is vice president, engineering, for InsureWorx Inc., based in Emeryville, Calif.

April 15, 2006

Copyright 2006© LRP Publications