In her recent blog, Anne Thomas Manes announced "SOA Is Dead-Long Live Services". She alleges that SOA has been killed by the economic depression because enterprises cannot afford major investments in IT transformation. She envisions web-based information services to be the survivors. The disillusionment with SOA is a result of misunderstanding perpetrated by software vendors anxious to leverage an industry trend for quick sales. Those who believe SOA is dead will be the real casualties.
SOA is a business architecture enabled by technology-it is not simply an application of web services technology. The OASIS SOA Reference Model is based on access to business capabilities across organizational boundaries. It is about relationships between organizations. It also is the basis for synergy between business and IT. The Internet and related technologies enable easy access to such capabilities independent of the particular technology used to implement the operation of the service.
According to a related article, "SOA Gets an Obituary" in Yahoo.Tech:
Interviewed Monday afternoon, Manes said successful SOA implementations have resulted from major IT transformation efforts rather than just slapping a bunch of interfaces on applications. ‘Those companies have seen spectacular results from these efforts, but in those circumstances, SOA was part of something much bigger,' Manes said.
Upon closer examination, I am sure the major transformations referenced above were really business transformations supported by technology. The real benefits of SOA come from improvements in the operation and agility of the business.
Technically, the key here is to establish a standard infrastructure in which services can be easily accessed and shared, and to establish standard message formats and data definitions by which service providers and service consumers can interact. IT should focus on enabling business by supporting the implementation and integration of business services.
From a business perspective, the key is to identify sharable capabilities and provide well-defined interfaces by which they can be shared and used in multiple business contexts. So, for example, the same billing operation (computer application and associated business activities) can be used by multiple lines of business. However, the transformation goes beyond just putting a façade on an old billing system, because the old billing system is most likely designed for a particular line of business-its functionality may need to be generalized. Furthermore, the billing operation no longer belongs to one of the lines of business; it belongs to the enterprise. Business savings will result from economies of scale in the billing operation, not just the elimination of redundant billing applications.
The consolidation of key capabilities is where a transformation to SOA starts. Such solutions as the shared billing system are the low-hanging fruit. Outsourcing of commodity services can provide other low-investment, high-ROI opportunities. You don't need to rush out, buy a lot of software and transform your IT infrastructure.
In the long term, SOA provides not only economies of scale, but agility. Our SOA Maturity Model (EDS an HP Company) defines a progression toward agility. It addresses both technical and business changes that must occur over a long journey. At the end of the journey, the agile enterprise embraces SOA so that sharable capabilities are made accessible as services. These capabilities include not only capabilities for delivery of customer products and services, but also capabilities that support operation of the enterprise such as accounting and human resource management services.
These capabilities (the shared business units) become building blocks for future business endeavors. Instead of creating a new organizational silo to address a new line of business, existing capabilities are incorporated into a new value chain, and only missing capabilities must be developed and integrated. The enterprise can adapt to challenges and opportunities much more quickly. Improvements to existing capabilities will have less pervasive effects because the affected functionality is more likely consolidated in fewer and smaller organizational units.
Far from being wiped out by the bad economy, SOA will enable smart enterprises to achieve economies of scale in the short term and agility in the long term. The SOA Maturity Model provides guidance for short-term ROI and continued improvement as the enterprise progresses to higher levels of maturity and agility. Those who abandon SOA will become competitively disadvantaged.
For more on SOA for business, the agile enterprise, and the SOA Maturity Model, see my book, Building the Agile Enterprise with SOA, BPM and MBM.
Enterprises attempting to implement a Service Oriented Architecture (SOA) are confronted with an initial investment in SOA infrastructure. This represents a financial burden for the first one or more service implementation projects, and thus a major hurdle for the SOA strategy.
In general, initiatives to move to SOA start with the IT organization. The enterprise is typically segmented into operational silos and initiatives are managed as projects within those silos. For SOA, the IT organization tends to focus on establishing the SOA technological capability, particularly the infrastructure. This is reasonable since there can be IT cost savings, and it would be desirable to build services on an infrastructure that will be there in the long term. However, this should not stand in the way of adopting SOA.
Infrastructure is needed because shared services will be accessed across silos. A SOA infrastructure should support ad hoc integration of many services across organizational boundaries. The challenge is to establish this networking capability at a time when initial participants see the improved infrastructure as increased costs with little value. It's like trying to establish a telephone network with the cost of the network carried by the first telephone installation. The infrastructure investment anticipates a return when the number of network participants reaches critical mass.
Unfortunately, most executive leaders will be skeptical about the potential business value of SOA. SOA is promoted primarily as a new technology with potential IT cost savings. Technical people will often focus on the technical challenges-the infrastructure, design and integration of services, the performance of distributed operations and the reuse of software. Software vendors are anxious to sell their products and emphasize the technical capabilities. However, the real benefits of SOA are in enabling improvements to the efficiency and agility of the enterprise. This must be demonstrated to executives, not only for the infrastructure investment, but for the changes that will be needed in organizations and governance. The SOA Consortium has captured industry case studies illustrating how individual enterprises have realized business value from SOA.
A SOA maturity assessment tool developed by EDS addresses the need for an enterprise to coordinate the development of both technical and business maturity. It is based on 5 maturity levels, similar to the CMMI (Capability Maturity Model Integration) used to assess an organization's software development maturity. A SOA maturity assessment will not only clarify an organization's level of readiness for SOA, but it will identify specific weaknesses that should be resolved before aspiring to more advanced levels. Sean Wiley of EDS and Andrew Sutherland of Oracle discussed the importance of the SOA Maturity Model.
In early stages, the most important aspect of a SOA strategy is to demonstrate the business value of SOA. That business value comes from the consolidation and sharing of business capabilities across the enterprise. There may be some business value in the consolidation of computer applications and supporting technology, but the real business value should come from transformation of the enterprise to exploit shared capabilities. Not only can consolidations yield economies of scale, but they can improve operating consistency and control, enabling problems and opportunities to be addressed more quickly and reliably. My book, Building the Agile Enterprise with SOA, BPM and MBM, describes the strategic business impact of SOA and provides additional information on the SOA Maturity Model.
Most large enterprises today have evolved as a result of mergers and acquisitions or additions of new lines of business. These often exist as relatively independent organizations each managing many of the same business capabilities. This may have been okay in the past, but in today's highly competitive and dynamic marketplace and with the current capabilities of information technology, this decentralization and autonomy is inefficient and inflexible
If a SOA infrastructure is not there for early consolidations, there will be rework later. The fundamental infrastructure requirement is support for loose coupling through asynchronous messaging-most enterprises already have message oriented middleware that can address this need. The lack of the preferred SOA infrastructure does not prevent the development and integration of well-defined service interfaces. The formatting and exchange of messages should be isolated in an interface layer to facilitate future adaptation. Good service design can minimize the cost of later rework.
If the IT organization can leverage initial project funding to implement a SOA infrastructure, so much the better, but infrastructure funding should not be a show stopper. The real challenge is to find redundant capabilities that can be consolidated to yield significant business value. That requires understanding the business and overcoming the political hurdles that will be fundamental to SOA transformation. That's where the energy should be focused.