Selecting the best supplier is a recurrent organizational challenge that occurs in a supply chain (SC) as a result of the presence of complex variables, restrictive criteria, and conflicting priorities. Since an SC network is often developed with ambiguous conditions and information due to the industrialization of society and the intricacy of market competitiveness, fuzzy decision-making models are more effective. This paper proposes a two-stage decision-making model to select suppliers and to estimate cost-effective order numbers per supplier. The initial stage of the proposed model involves identifying fuzzy linguistic variables, interpreting appropriate decision criteria for evaluating suppliers, and modelling fuzzy technique for order preference and similarity to ideal solution (TOPSIS) method. The goal of fuzzy TOPSIS method is to attenuate the ambiguous expert inputs. In the second stage, economic order quantity is determined and assigned to each supplier using TOPSIS scores as inputs for a linear programming (LP) model. Different constraints, including demand, density qualification, acidity qualification, price, and capacity are formulated using the LP model. The mathematical model seeks to optimize total value of purchasing. The model is implemented in a dairy company to show its applicability and effectiveness. It has been found that supplier A1 and supplier A4 need to deliver 8000 kg of dry milk to the company, while supplier A5 needs to supply only 3500 kg. It is expected that the obtained results will assist organizations in developing a methodical strategy for addressing order allocation and supplier selection problems in more a realistic context.