HOME | SPEAKING | PUBLICATIONS | ABOUT JSB | CONTACT ME

Foreword for Understanding the Power (and Limitations) of Web Services by Anne Manes

by John Hagel, III; John Seely Brown
February 12, 2003

This book performs a valuable service for managers seeking to harness the business potential of Web services technology. Web services represents a major step in the direction of service-oriented architectures and these, in turn, will provide a foundation for extraordinary innovation in business practices and industry structures.
As with any new technology, there is enormous confusion, generated by the hype machines of major technology vendors and assorted enthusiasts. Will Web services enable dynamic composition of applications from hundreds of micro-services? Will Web services truly allow all applications to finally connect seamlessly with all other applications? Will we finally overcome the barriers created by generations of proprietary technology layered on top of and around each other? Will XML really provide a lingua franca that can transcend all the fragmented dialects that make communication among applications so challenging?

This confusion is dangerous for several reasons. At the very least, it will slow down adoption of the technology. Managers have been burned by exaggerated claims from previous generations of technology many times in the past. They are understandably gun shy at this point and develop an allergic reaction to any suggestion of hype surrounding a new technology.

But there is an even bigger danger. Some managers might actually believe the hype and launch implementation efforts that far exceed the capabilities of the technology. If they are lucky, the implementation effort will merely fall short of expectations and fail to deliver the anticipated return on investment. A few will be truly unlucky and find that they have compromised key business activities because of limitations in the technology. If that happens, the broader impact of Web services technology will be seriously compromised as a backlash gathers force and creates a serious obstacle for anyone interested in implementing this technology.

If Web services technology is the first step in the direction of much more ambitious service-oriented architectures, it is essential that the first step be a solid one, free from stumbles and moving the business briskly forward. For that to happen, we need absolute clarity regarding the capabilities of the technology and, even more importantly, the limitations of the technology as it exists today.
This book by Anne Thomas Manes does precisely that. Bringing a real practitioner’s experience to the task, Anne carefully walks managers through the fundamentals of the technology, taking particular care to point out the limitations. She clearly sees the potential of the technology, but she also understands that, to realize this potential, expectations must be managed carefully and potential adopters need to be clear about the performance boundaries of the technology.

With this book in the hands of CIO’s and business strategy executives, we will be much more likely to see accelerated adoption of Web services. Confident that they understand the technology, these executives will be more likely to deploy Web services in production environments. Alert to the limitations of the technology, they will be assured that Web services will deliver the functionality required to generate real business impact. As the compelling return on investment supporting this technology becomes apparent, executives will broaden the deployment of Web services and the technology providers will be motivated to continue to invest heavily in enhancing Web services capabilities.

This is a deceptively disruptive technology. Enterprises are adopting Web services to support very mundane technology “plumbing” in order to deliver tangible operating savings. Nothing very disruptive about that. Most Web services enthusiasts find such implementations very uninteresting. On the other hand, most business executives are very excited about the prospect of achieving substantial operating savings with modest investment and relatively short lead-times measured in months instead of years. Achieving these near-term savings does not require any major changes in business practices or organizations.

As the technology penetrates the enterprise, it begins to create more degrees of freedom for managers. Exploiting this freedom will require disruptive change in terms of business practices and business organization. Those who have the understanding and commitment necessary to confront these disruptions will reap the real economic rewards of Web services.

What is the source of this disruptive power in Web services technology? It begins with the notion of contract-driven integration embedded in key Web services standards. Web services standards define the foundations for a service-oriented architecture. As Anne points out, one of the three basic artifacts of this architecture is the service contract. The Web Services Description Language (WSDL), a key early Web services standard, specifies the format for this service contract. This service contract describes what will be offered (the service type), what procedures must be used to access and use the service (the binding) and where the service can be found (the endpoint).

Now, service contracts in themselves are not a major breakthrough, as Anne points out. Previous generations of service-oriented architectures also used standardized ways of describing service contracts. The big difference, though, is that these earlier forms of service contracts only supported one way of communicating with the service and required all users to install the same middleware in order to locate the service to begin with. In other words, these service contracts were still tightly coupled — they required all users to conform to the same communication protocols and the same middleware platforms.

Service contracts implemented with Web service standards support much more loosely coupled connections. Services can be located, accessed and used over a variety of middleware platforms. The standards assume that providers and users of Web services will operate across a broad range of technology platforms and therefore seek to provide maximum flexibility in creating and supporting connections.

This flexibility is key to the power of Web services technology and related service-oriented architectures. The real opportunity (and challenge) is to extend this philosophy of loose coupling to asynchronous transactions. As Anne points out, one of the key sources of operating inefficiency in business processes involves the difficulty in automating long-running, loosely coupled asynchronous transactions — multi-step business processes that span extended periods of days or weeks. The operating inefficiencies are especially significant when these business processes span multiple enterprises.

By addressing this opportunity, Web services and service-oriented architectures will provide the technology foundation required for a very different approach to management of business processes. In contrast to the hard-wired approaches used throughout business today, we will see the spread of loosely coupled business processes. These business processes will be designed in a modular fashion so that orchestrators of these processes can quickly assemble the right set of activities to tailor the process to the needs of specific customers or products.

Of course, to realize the potential of this technology foundation, additional elements will need to fall into place. In particular, we would highlight the need for service grids and for a very different management approach based on trust rather than control.

Service grids assemble a diverse set of specialized enabling services required to support the connections across applications enabled by Web services technology. For example, specialized enabling services would include security, directory services, data transformation and performance auditing. These services are essential to deliver mission critical functionality.

Rather than implementing them at each node, creating significant complexity and expense for each party participating in the connection, service grids would make these services available on a shared basis. The various enabling services required to compose a service grid are likely to be assembled in a federation from a variety of sources. Some of the more specialized ones may be provided from within enterprises participating in the connections while most of them will likely be sourced from specialized third parties that view these services as their core business.

This would create significant efficiency in terms of avoiding replication of functionality at each node. It would also provide a powerful platform for accelerated learning and performance improvement, offering providers of these shared services much broader visibility into the requirements for performance in many different operating environments simultaneously. Service grids are only now beginning to emerge. They will need to evolve significantly in order to handle the challenges of long-running, loosely coupled business processes. For more information about service grids, see our white paper “Service Grids: The Missing Link in Web Services” available at www.johnhagel.com.

Service grids are essential to enhance the performance of the technology itself. But to really harness the economic value of the technology, management will need to shift to a very different management approach. Most executives today are experienced practitioners of a control-based management approach. Shaped within the enterprise, this management approach assumes a single point of control. It specifies activities in great detail and then seeks to monitor those activities at a granular level to ensure quick intervention if actions deviate from their prescribed course. In part, previous generations of technology shaped this management approach. Tightly coupled technology platforms required detailed specification of activities in advance and strict adherence to these activity flows.

Two forces have been increasing stress on this control based management approach. First, companies have realized that enhancing value for the customer requires coordinating activities within business processes across multiple enterprises. Thus, the assumption of a single point of control, so essential to the effective functioning of control-based management approaches, becomes questionable. Federated models of decision-making become essential. Second, intensifying competition and accelerating change make it more difficult to specify detailed activities in advance. It becomes even more difficult to adhere to these specifications as countless unanticipated events and exceptions disrupt operations. Flexibility becomes essential.

These stresses are creating a need to move to a trust-based management approach. This approach starts with the assumption that multiple enterprises will need to come together to deliver value to customers. Rather than a single point of control, this approach focuses on the need for an orchestrator — someone who will identify and recruit participants with the required capabilities, clearly specify and communicate expectations regarding deliverables and create the appropriate incentive structures to motivate each participant to perform as expected. Rather than specifying activities, the orchestrator specifies outcomes. Rather than seeking complete transparency regarding all activities in the process, the orchestrator seeks selective visibility from each participant regarding events that might put outcomes in jeopardy.

Realizing the full economic of Web service technology will in the end require significant changes in management approaches. As more loosely coupled business processes emerge, shaped by trust-based management approaches, executives will realize they have far more options in terms of organizing their business. Activities that previously were internal to the enterprise will now be dispersed to enterprises that have far more capability. More distinctive activities will remain within the enterprise, but they will be made available to other enterprises on an as needed basis. As this process unfolds, enterprises will need to be rearchitected, in terms of business models, governance approaches and organizations. In parallel, industry architectures will rapidly evolve in response to the new economic and competitive dynamics generated from service-oriented technology architectures and loosely coupled business processes.

What started as a relatively benign way to achieve near-term operating savings will reveal its true disruptive potential. Substantial economic value will be created as these disruptions unfold. At the same time, substantial economic value will also be destroyed. Managers who understand the true business potential of this technology will be in the best position to create, rather than destroy, value.

But that is another story. For now, we need to focus on the first steps. Anne does a superb job of helping managers to understand this technology so that they can move with sure footing and avoid potentially harmful stumbles along the way.