Defining replenishment policies for perishable products is an important activity, particularly where suppliers have a range of products. As product ranges increase, consumers can substitute products if their preferred product is out of stock. Such substitution considered simultaneously as perishability makes it difficult to achieve balanced results over different departments/companies in the face of fluctuating demand. Given these circumstances, a financially calculated replenishment policy makes communicating the impact of operational changes difficult. In contrast, non-financial measures improve the communication between departments and staff (e.g., between warehousing, procurement, and sales), and allows them to set operational targets from broad corporate strategies. The first objective is to use non-financial performance measures to define the most favourable replenishment policy in a two-echelon model with multiple perishable and substitutable products. The second objective is to evaluate and explore the importance and interactions of the input factors (i.e., consumer demand, product lifetime, and substitution) in a perishable and substitutable inventory management model with sensitivity testing using MANOVA. Developing the framework consisted of three steps. First, the discrete event simulation (DES) was built and run for each of a given set of replenishment policies. The performance of the inventory model under each replenishment policy was measured by three conflicting non-financial performance measures; specifically, average inventory, fill rate, and order rate variance ratio. Second, the analytic hierarchy process (AHP) method was used to weight the importance of each performance measure. Third, the data envelopment analysis (DEA) method was used to evaluate and rank the performance of each replenishment policy. Then, the most favourable replenishment policy, which has the lowest DEA Cook's superefficiency score, was chosen. The results showed that the consumer demand, product lifetime, and substitution inputs to the model have large effects on retailers' and supplier performance; however, only the interaction between consumer demand and product lifetime had a similarly large effect on firms' performance. Suppliers are more greatly affected by the bullwhip effect in the model; in contrast, the effects on the retailers is smaller. Moreover, this research also shows that, in the studied context, the most favourable replenishment policy is stable under changes in the weights of performance measures. This study contributes to inventory management theory by being the first research to develop a non-financial framework and demonstrate that it is comparability to financial approaches for perishable and substitutable inventory. For managers, this study contributes by providing a framework (based on non-financial measures) to develop or modify replenishment policies to balance service level/cost in contexts with perishable and substitutable products. The framework is particularly relevant for suppliers, as they are more impacted by fluctuating demand. The non-financial approach also enables managers to evaluate the effectiveness of other supplementary techniques (e.g., forecasting techniques) in the inventory management when making a business case.