Workflow-based Process Controlling

The exact and timely analysis of automated business processes through the analysis of protocol data is one factor for the use of workflow management systems in organizations. While in the past the automated routing and distribution of work have been prevailing factors in the marketing of workflow applications, the side-effect of having actual and precise business data for analysis yields many opportunities for controlling and managing business operations that have unfortunately been neglected in the past. We outline the economic aspects of workflow-based process monitoring and controlling and the current state of the art in monitoring facilities provided by current workflow management systems and existing standards. After a discussion of the three evaluation perspectives, sample evaluation methods for each perspective are discussed. The integration of workflow audit trail data into existing controlling environments and implementation issues are outlined before an outlook on further research is given. EVOLUTION OF PROCESS INFORMATION SYSTEMS Workflow-based Monitoring and Controlling The need to serve customers in global markets and the tendency toward smaller, more flexible, less hierarchical organizations leads to an increasing spatial distribution of companies. This distribution of formerly centralized enterprises on the one side and the (temporary) integration of different companies (virtual enterprises) on the other side lead to the necessity of an organizational and functional connection of the distributed workplaces. Distributed information systems and Internet-based workflows are the enablers for these kinds of ventures. Recent standards such as Wf-XML [1] foster the use of workflow management systems for these cross-organizational ventures. In combination with concepts such as Customer Relationship Management (CRM), workflow systems enable companies to streamline their processes while maintaining a clear customer focus. The prediction, that in the next few years, companies implementing CRM strategies will have to rely on solutions composed of best-of-breed components (cf. [2]) strengthens the role of workflow management systems as the “glue” in modern business application architectures. During the enactment of workflow-supported business processes, the automation of coordination increases the efficiency of process execuSECTION ONE—THE WORLD OF WORKFLOW 62 tion through the elimination of transport times, the automation of routing decisions and the monitoring of deadlines. However, the actual benefits of workflow management are usually described in vague terms. While the costs for the selection and introduction of the system can be calculated rather precisely, the benefits of workflow automation are more difficult to determine. Besides quantitative benefits such as decreasing cycle times, reduced personnel cost and (if workflow and document management technologies are combined) document storage space and paper cost, qualitative aspects have to be taken into account as well. These figures include shorter time to market due to process improvement, higher process quality due to decreased error rates and faster response to customer inquiries. Moreover, while most costs from the introduction of a workflow management system are generated during the system introduction phase, the benefits are generated over a much longer period of time. In order to enable a cost-benefit analysis during the requirements analysis phase of a workflow project, time-adjusted methods of investment analysis are necessary. The historical data of workflow instance execution provides a valuable source for the analysis of the economic impact of workflow management systems. This historical data can be evaluated either in realtime (monitoring) or during an ex-post analysis (controlling). Workflow monitoring deals with the analysis of workflow instances at run-time. Active monitoring of the current state of workflow instances can serve numerous purposes, such as the generation of exception reports for overdue work items or early warning reports for potentially overdue work items (Management-by-Exception, cf. e. g. [3]). Passive monitoring upon request can deliver status information about running workflow instances, e. g. for answering a customer inquiry about the status of an order (cf. e. g. [4], [5]). Workflow monitoring can also be divided into technical and organizational monitoring. While technical monitoring is used for performance measurement (e. g. response time, system load etc.), organizational monitoring measures the organizational efficiency (e. g. idle times, workload analysis etc.). Workflow-based controlling aims at the ex-post analysis of the audit trail data of process enactment (sometimes called strategic process controlling). Here, the single workflow instances are aggregated according to different dimension schemes as described in section 3. Workflow-based controlling is useful for the detection of long-term developments in workflow enactment and the review of already existing workflow implementations. In order to identify deviations in process execution, the audit trail data is often compared to target data derived from corresponding business process models. The efficient integration of audit trail information within existing repositories such as Data Warehouses is an important factor in the usefulness of the results of workflow-based controlling applications. The positioning of the two phases is shown in figure 1. WORKFLOW-BASED PROCESS CONTROLLING 63 Business Process Analysis and Design Workflow Model Implementation (Build Time) Workflow Execution (Run Time) Workflow-based Controlling Business Process Model