Developing Taxonomies for Business Process Engineering

In many business environments, we find extensive business process structures that consist of many individual processes, each with a complex composition of activities. The elements in the processes are often based on an ad-hoc, existing way of working, which is not always properly documented. The processes evolve over time, not rarely on a per-process basis. Consequently, process definitions diverge and the use of process elements within and between process definitions becomes misaligned. To address this issue, we propose the use of catalogs of standardized process building blocks in business process engineering. Different from approaches using patterns, we base our catalogs on foundational parts (which we call primitives) organized in three dimensions: business process activities, objects manipulated by activities, and actors performing activities starting with the semantics of processes, not the syntax. To provide a solid basis for the structuring of each of the dimensions (and hence the organization of the foundational parts in the catalogs), we use taxonomies. In this paper, we discuss the development of these taxonomies. We apply a slightly modified existing taxonomy development method, which uses both deductive and inductive steps. We discuss the development of one taxonomy in detail, basing the inductive steps on processes from a complex, real-world case organization. In doing so, we make a first step towards a business process engineering approach that is centered on a process-content-first point of view, aligned with the needs of the process management practice.