Design of Data-Driven Decision Support Systems for Business Process Standardization

Increasingly dynamic environments require organizations to engage in business process standardization (BPS) in response to environmental change. However, BPS depends on numerous contingency factors from different layers of the organization, such as strategy, business models (BMs), business processes (BPs) and application systems that need to be well-understood (“comprehended”) and taken into account by decision-makers for selecting appropriate standard BP designs that fit the organization. Besides, common approaches to BPS are non-data-driven and frequently do not exploit increasingly avail-able data in organizations. Therefore, this thesis addresses the following research ques-tion: “How to design data-driven decision support systems to increase the comprehen-sion of contingency factors on business process standardization?”.
Theoretically grounded in organizational contingency theory (OCT), this thesis address-es the research question by conducting three design science research (DSR) projects to design data-driven decision support systems (DSSs) for SAP R/3 and S/4 HANA ERP systems that increase comprehension of BPS contingency factors. ... mehrThe thesis conducts the DSR projects at an industry partner within the context of a BPS and SAP S/4 HANA transformation program at a global manufacturing corporation.
DSR project 1 designs a data-driven “Business Model Mining” system that automatical-ly “mines” BMs from data in application systems and represents results in an interactive “Business Model Canvas” (BMC) BI dashboard to comprehend BM-related BPS con-tingency factors. The project derives generic design requirements and a blueprint con-ceptualization for BMM systems and suggests an open, standardized reference data model for BMM. The project implements the software artifact “Business Model Miner” in Microsoft Azure / PowerBI and demonstrates technical feasibility by using data from an educational SAP S/4 HANA system, an open reference dataset, and three real-life SAP R/3 ERP systems. A field evaluation with 21 managers at the industry partner finds differences between tool results and BMCs created by managers and thus the po-tential for a complementary role of BMM tools to enrich the comprehension of BMs. A further controlled laboratory experiment with 142 students finds significant beneficial impacts on subjective and objective comprehension in terms of effectiveness, efficiency, and relative efficiency.
Second, DSR project 2 designs a data-driven process mining DSS “KeyPro” to semi-automatically discover and prioritize the set of BPs occurring in an organization from log data to concentrate BPS initiatives on important BPs given limited organizational resources. The project derives objective and quantifiable BP importance metrics from BM and BPM literature and implements KeyPro for SAP R/3 ERP and S/4 HANA sys-tems in Microsoft SQL Server / Azure and interactive PowerBI dashboards. A field evaluation with 52 managers compares BPs detected manually by decision-makers against BPs discovered by KeyPro and reveals significant differences and a complemen-tary role of the artifact to deliver additional insights into the set of BPs in the organiza-tion. Finally, a controlled laboratory experiment with 30 students identifies the dash-boards with the lowest comprehension for further development.
Third, OCT requires organizations to select a standard BP design that matches contin-gencies. Thus, DSR project 3 designs a process mining DSS to select a standard BP from a repository of different alternative designs based on the similarity of BPS contin-gency factors between the as-is process and the to-be standard processes. DSR project 3 thus derives four different process model variants for representing BPS contingency factors that vary according to determinant factors of process model comprehension (PMC) identified in PMC literature. A controlled laboratory evaluation with 150 stu-dents identifies significant differences in PMC. Based on laboratory findings, the DSS is implemented in the BPM platform “Apromore” to select standard BP reference mod-els from the SAP Best Practices Explorer for SAP S/4 HANA and applied for the pur-chase-to-pay and order-to-cash process of a manufacturing company.