Web Services: “Why do it?” and “Why do it Now?”.
The unique attractiveness of Web services lies in its pervasiveness and language independence, in its inherent capability to support loosely coupled connections, and in the fact that Web services can be implemented gradually over time, one step at a time.
Pervasiveness and language independence is achieved by relying on the Web (and its protocols). As the central middleware layer, the Web infrastructure has been endorsed by all major systems vendors, for all major devices, software- and hardware platforms. Loose coupling is archived by a layer of abstraction, the ability to generate proxies, (where required, even dynamically at execution time) and the ability to map formats easily via style sheets. Gradual implementation means that (different from past innovations, like the switch from mainframe to client-server technologies) Web services can be adopted over time, while measurable benefits can be achieved often within a matter of months.
Taking all the above benefits together, it becomes increasingly hard for decision makers to ignore the presence and the business opportunities associated with Web services. The potential business benefits are significant, and cover a wide spectrum. On one side of the spectrum is the opportunity to significantly reduce the total cost of ownership (TCO) of the existing infrastructure. According to leading studies, a huge portion of IT budgets those days is allocated to just keep existing infrastructures operational. The tight coupling of today’s applications requires significant efforts in redesign and regression testing, triggered often by even minor changes (like upgrades, extensions, or even just repairs) in some part of the system. Gradually loosening up this coupling via a message-based architecture will lead to (often significant) reductions in maintenance costs. As another example, a new level of service to both internal as well as external clients, will furthermore help to reduce TCO while simultaneously increasing competitiveness and leveraging existing assets. Crucial information like shipping information, when exposed securely via Web services, will allow vendors to seamlessly integrate into a client’s planning applications. This does create true competitive advantage, while reducing the support needs via call-centers for the now-empowered customer. (It should be noted, that the loose coupling of Web services at some point of time also likely will reduce the costs to switch to competing vendors for the customer. So, offering Web services will not by itself lead to sustainable advantage. However, not only will late adopters be penalized, the first mover’s advantage can be significant). Similarly to the above-mentioned external services, internally deployed services allow for non-intrusive adoption of new business processes, allowing businesses to stay flexible (while controlling the feared beast called scope creep, by focusing on small but clearly defined ‘baby steps’). Finally, new services or even business models will evolve around Web services. I would surely expect the majority of market research data providers to start offering their data soon in form of Web services, which should allow a significant boost in sales volumes, as more customers (especially those customers with smaller IT budgets, who could not afford incorporating those data before due to high IT interfacing costs) will decide to incorporate this information into their own business intelligence infrastructure. And there are other more prominent examples, too many to mention here.
Does it all sounds too good to be true? You may ask, given all the benefits mentioned above, so why isn't every corporation ‘doing Web services’ already? The reasons are two-fold. On one side, the Web had to overcome its ‘hype’ stigma from the internet bubble. Reliable research and market education, as well as well documented success stories have meanwhile helped to overcome this stigma, and have allowed the return of the Web into the focus of strategic planners in respectable organizations. On the other side, some of the technologies are fairly young, and standards have been emerging over last years, and are partially still under development. The WS-I (Web Services Interoperability Organization, with SAP as one of its founding members) focuses on the enabling of interoperability. It does so by building on the established de-facto standards like SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language), however by filtering on a core set of options and features to be supported by compliant providers. This is been done by reviewing real world use cases business scenarios, to assure the relevance of the standard for the real world. With its 25++ years of business software expertise, SAP is a significant player. More formal standardizations efforts are driven by bodies like W3C (World Wide Web Consortium) and OASIS (Organization for the Advancement of Structured Information Standard). One of the deliverables of the combined efforts is the SOAP 1.2 standard, which widely enhances and improves the SOAP 1.1 standard, by reducing ambiguities and resolving omissions. Similarly, also an updated WSDL standard is under review. With regard to web services registries (the place to register, to discover and to access Web services descriptions), progress has been made via an informal consortium called UDDI.org. In its latest specifications, especially the flexible support for taxonomies (a key ingredient) has been greatly enhanced. While several public registries are available (including one operated by SAP), some analysts expect the real power to lie in the private registries (perhaps somewhat a similar trend as we saw with e-exchanges), although equally based on UDDI.
Given that some of the standards are still evolving, the valid question does arise, whether to wait or to take action now. I would recommend taking action now for several reasons: First of all, the current standards (while still being improved) already offer a solid base to execute against, and have been endorsed by many vendors. Speaking on behalf of SAP, SAP NetWeaver offers inherent support for Web services. It is straight-forward to develop new Web services, or to build Web service wrappers against existing RFCs or BAPIs. SAP NetWeaver components like EP or BW expose some of their key capabilities as web services, and are at the same time able to consume a wide variety of web services. XI as a message based infrastructure represents the state-of-the-art XML based middleware. Web AS comes equipped with a UDDI registry. Furthermore, Web AS also seamlessly allows for integrating the different platforms, like J2EE, .NET or Websphere (remember, the ability to integrate those different platforms is one of the beauties of Web services). Secondly, the first mover’s advantage can me immense, and so not only when developing entirely new business models. Web services allow you to reduce customer’s pain points, while often simultaneously reducing support costs. (In a similar way, as e-banking did 5 years ago. Speaking from my own experience: Being on the road a lot, I had switched to a major bank which had endorsed e-banking early-on. While many banks offer similar service, I did not see a reason so far to switch banks again. Clearly, the first mover’s advantage was crucial in this case). Finally, it is worthwhile to remember that it’s a historically proven fact that typically standards do not drive innovation, but standards follow innovation. It is up to the leaders in the community to take action now. For the IT vendors, to drive technology, to sort out what’s feasible and what’s practical, and to pro-actively participate in the relevant bodies. For corporations, to evaluate and adopt those technologies, under clear business aspects, while working hand-in-hand with the leading, recognized and dependable business technology providers. And for consultants and thought-leaders, to familiarize themselves with the available technologies (like SAP NetWeaver) and trends, to engage in fruitful discussions, and to implement bullet-proof and scalable solutions based on web services.
Links and recommended further readings:
