Virtual Enterprise (VE) is a temporary platform for individual enterprises to collaborate with each other, sharing their core competencies to fulfill a customer demand. In order to improve the customer satisfaction, the most successful VEs select their consortium's members based on customer's preferences. There is quite extensive literature in the field of partner selection in VE, each proposing a new approach to evaluate and select the most appropriate partners among pool of enterprises. However, none of the studies in literature recommend which partner selection methodology should be used in each project with a particular customer attitude. In this study an algorithm is proposed which classifies the customers into three categories; passive, standard and assertive. Three different approaches; Fuzzy Logic-FAHP TOPSIS and Goal programming are used for each customer type respectively. This classification is beneficial since the problem's characteristics; such as vagueness of data, change as the customer's attitude varies. The results certify that, adopting this algorithm not only helps the VE to select the most appropriate partners based on customer preferences, but also the model adapts itself to each customer's attitude. As a result, the overall system flexibility is significantly improved. (C) 2016 The Authors. Published by Elsevier B.V.