Article Preview
TopIntroduction
Service-Oriented Architectures provide major advantages for today's enterprise information systems by presenting the interfaces that loosely coupled connections require (Papazoglou, 2007). Papazoglou and Van den Heuvel (2006) provide an early overview of methods and techniques used in service-oriented design. In their view, (web) service design and development is about identifying the right services, organizing them in a manageable hierarchy of composite services and choreographing them together for supporting a business process. Following SOAD (Zimmerman et al., 2004), they distinguish between top-down, bottom-up and meet-in-the-middle approaches and discuss major principles of service design such as low coupling and high cohesion. Although these criteria are relevant, the approach considers service design mainly as a software development problem, which in our view is not sufficient. As stated in Nayak et al. (2007), “the current trend toward a service-oriented enterprise necessitates a formal characterization of business architecture that reflects service-oriented business thinking”. The starting point for design should be the business level at which services can be identified that provide value to customers and can be offered in an economically viable way. It is at the business level that business concerns should be dealt with first.
Value creation has always been the most important business concern, but only recently, business researchers have started to recognize the intimate relationship between value creation and service provisioning. Vargo and colleagues (Vargo & Lusch, 2004; Alter, 2008) have introduced the concept of “service-dominant (S-D) logic”. S-D logic focuses on service provision in contrast to goods production (G-D logic). The shift from G-D to S-D logic is one from a value proposition consisting of operand (passive) resources to one consisting of operant (active) resources. Instead of only seeing value being created within companies that exchange the means for this value creation from one to another, it emphasizes the value being created between companies (or companies and consumers). Grönroos (2008) goes even further by saying that value is primarily created at and by the customer, and the company is just co-creator.
If value creation is the main business concern, it is clear that neither software models nor business processes provide the right level of abstraction for business value modeling. What we need is a modeling language that allows to identify business services and to reason about them from a value perspective. At the same time, we want the language to be a basis for process design and system development, so that business and IT are aligned rather than evolving in isolation. The Model Driven Architecture (MDA) provides a useful framework, as it allows traceability between diagrams at different levels. Following an MDA approach, business service models can be positioned at the CIM (computation-independent) level, and related to process models at the PIM (Platform Independent level). At this level, the efficient and effective coordination of business services is a major concern. PIM models (e.g., a BPMN model) can be mapped to an implementation platform, using available software components wherever possible.
A model-driven approach to service modeling has also been proposed by SOD-M (De Castro, 2009). SOD-M uses e3value models at the business (CIM) level and UML at the PIM and PSM level. Although e3value is indeed a useful starting-point for analyzing business models (cf. Weigand et al., 2009), it is also limited in the sense that it focuses on the interactions in the value network while largely ignoring the production processes, and has not been set up for “service-oriented business thinking” in the first place.