Target setting and benchmarking by the possibility of selecting the closest alternative benchmark

Data envelopment analysis (DEA) stands out from other performance evaluation methods because it focuses on providing benchmarks. Benchmarking plays a vital role in improvement the performance of inefficient units, especially when classical models are combined with managerial perspectives, potentially resulting in a more feasible benchmark or target. Sometimes, choosing certain efficient units in the system as a benchmark and target isn't feasible from a management standpoint for inefficient units. This happens for various reasons, including the exclusivity of an efficient unit, the outlier status of an efficient unit, insufficient access to resources, or discrepancies with reality. Such units shouldn't be chosen as benchmark and target for inefficient units within the system. Therefore, disregarding such units in the benchmarking process based on the manager's preference and opinion can contribute to better alignment with reality. Implementing this strategy reduces the disparity between actual and efficient performance. This paper presents a model designed to address these considerations. The model takes into consideration managerial opinions, enabling managers to identify and ignore specific efficient units during benchmarking. It then proposes alternative benchmarks closest in similarity to improve efficiency for the inefficient units.