Mapping LARGS criteria and relationships for supplier selection using a fuzzy hybrid approach

This paper provides a detailed mapping of the Lean, Agile, Resilient, Green, and Sustainable (LARGS) criteria for supplier selection, identifies the most relevant criteria, and evaluates the relationships between them in a decision-making process. Using fuzzy DELPHI to collect experts’ opinions, 16 key LARGS criteria were identified starting from a set of more than 50 factors derived from the literature. Then, applying a combination of fuzzy DEMATEL-ANP and gathering again the opinions of experts, the complex interrelationships among those criteria were mapped, highlighting how these connections can impact the final weighting of each criterion in a supplier selection process. From a theoretical perspective the proposed analysis aims to identify the most important criteria for supplier selection within a LARGS framework and to show how these weighted criteria connect each other in a decision-making process. The findings show that “emergency order processing” and “Environmental Management System” are the primary factors influencing the network of interrelations among criteria, while “pollution control”, “reputation” and “vulnerability” are the most critical factors in the selection of LARGS suppliers. The final weights for the criteria, obtained taking into account the relationships among the factors and LARGS perspectives, were tested in a case study, to appraise their effect in real supplier selection processes. From a practical viewpoint, the outcomes clearly show that considering the relationships among the LARGS criteria can change the ranking of suppliers, and thus, those relationships play a critical role in selection process; this confirms the importance of a comprehensive evaluation of suppliers to achieve effective and sustainable supply chain management. Valuable insights for decision-makers, to optimize potential supplier selection processes within increasingly complex, dynamic and competitive supply chains, can be derived.