Organization design decisions: A card sorting approach

In this paper, we treat organization design as a sorting (or clustering) process. We used a card sorting methodology where participants were asked to decide who (in a set of people/roles) should be organized in the same team, given information about each person (or role’s) tasks and interdependencies toward other roles. We compared the participants’ decisions to an algorithmically derived solution, and also developed an index to measure the distance between the participants’ decision and the optimal solution. We found that the results were dependent on the complexity of the task: Performance was lower on the more complex tasks, consisting of 9 and 12 roles, than the simple task with 6 roles. We also found that the tendency to discard interdependencies between roles and use a simple heuristic (i.e., to sort the roles based on their titles) was somewhat stronger with the 9-role task compared to the simplest task; but the results did not support the hypothesis with regards to the 12-role task. Finally, the data suggest that the likelihood of identifying the optimal solution increases with the time spent on each task (this relationship was significant for the 9 and 12-role tasks). We believe that our methodology can be used to identify the “microfoundations” of organization design and we discuss how the methodology can be refined and serve as the basis for future studies.