Identifying the Workflow Potential of Business Processes

Though the activities of business organizations in the field of workflow management are facing an enormous growth, a methodical solution for an essential problem occurring in many workflow projects is still lacking: the identification of those processes that can be supported by workflow applications in a profitable way. In this paper we present a structured framework, which can serve as a guideline for the evaluation of processes during the selection and introduction of a workflow management system. This framework contains three groups of criteria: technical, organizational and economic. Designed as a scoring model, this approach enables the user to evaluate the workflow potential of business processes systematically. First, the underlying concepts and design of the framework are outlined. In the following we describe, how the framework can be adjusted to individual needs through a three-step aggregation process. In order to demonstrate the practical relevance its application within an industrial workflow project is described. 1. Current practice of workflow projects Workflows are processes temporal and logical sequences of activities that are necessary to perform operations on economically relevant objects whose control logic lies within the control sphere of an information system. Every workflow is based on a process model that has been enhanced with additional attributes supporting its automation, the so called workflow model. A workflow management system is an information system that supports the execution (and optionally the specification) of workflow models. From the point of view of the workflow management system the elementary process activities are encapsulated, i.e. the workflow management system has an external view on the automated or manual execution of the functions but it is not concerned with the internal functionality of the activities. With the rapidly growing interest in workflow management, an increasing number of corporate projects involving the implementation of workflow-based applications [1] can be observed. In this area, companies face 0-7695-0001-3/99 $10 several problems such as fitting a new information system into the given IT infrastructure or the selection of the process candidates for workflow automation. While technical questions are extensively discussed in the literature, there is a lack of an elaborate and detailed solution for the identification of workflow candidates among the business processes of the enterprise. The identification of business processes with high workflow potential, i.e. those processes that can be supported by workflow applications profitably, is an essential issue within a workflow project because the selected processes and their system environments are the core determinants of the technical and business related requirements for workflow systems [2]. A delay in, or even a neglect of, this task leads inevitably to severe scheduling problems or, in the worst case, to the failure of the workflow project. The absence of an appropriate discussion of the workflow-relevant processes is also reflected in the common practice of workflow projects: Without a detailed analysis of the processes the selection of the workflow system is often reduced to technical criteria. Consequently, the functionality of the workflow management system dominates the business requirements of the processes that are subject to workflow automation. This goes along with the popular belief, that processes already optimized during BPR projects do not need to be changed anymore [3]. As a consequence, workflow projects often have failed or have been realized just to a minimal extent. Indications for this are the low number of workflow-based applications that are actually applied in day-to-day business, many non-operative pilot installations and minor applications for backoffice processes such as vacation day permission or travel expenses settlement (cf. [4], [3]). These facts are even more important as the mere introduction of a workflow management system does not necessarily guarantee any economic advantage. On the contrary, the purchase of complex workflow management systems and (to a much higher extent) their introduction, operation and maintenance, induces immense costs. Therefore, it has to be realized that workflow management cannot make business better in general, but that an economic and technical, methodically sound analysis of