257 14 th International Research/Expert Conference ”Trends in the Development of Machinery and Associated Technology” TMT 2010, Mediterranean Cruise, 11-18 September 2010 A SOLUTION TO CUSTOMER SELECTION PROBLEM IN LOGISTICS USING THE ANALYTIC NETWORK PROCESS (ANP) Onur Yilmaz Yildiz Technical University Beşiktaş Yerleşkesi Endüstri Mühendisliği Bölümü 34349 Beşiktaş, Istanbul Turkey Ali F. Güneri Yildiz Technical University Beşiktaş Yerleşkesi Endüstri Mühendisliği Bölümü 34349 Beşiktaş, Istanbul Turkey Bahadir Gülsün Yildiz Technical University Beşiktaş Yerleşkesi Endüstri Mühendisliği Bölümü 34349 Beşiktaş, Istanbul Turkey Şenim Özgürler Yildiz Technical University Beşiktaş Yerleşkesi Endüstri Mühendisliği Bölümü 34349 Beşiktaş, Istanbul Turkey ABSTRACT Logistics is a fast growing business sector in Turkey as well as the rest of the world. The need businesses have to transport, store and deliver goods is now acknowledged as a core factor building the product cost. Thus companies offering service in this sector basically combine needs of different customers and cater to all of them to take down the logistics costs, and therefore the product costs. This kind of logistics service is called 3 rd Party Logistics (3PL) and some companies in this sector are distinguished with the high quality of their service, and thus, are dealing with growing demands from customers. At one point, the demand from the customers exceeds the capacity of the 3PL service supplier, forcing the supplier to select some customers from among others to work with. Since there are many criteria concerning the customer selection problem in logistics, this is handled with a multi-criteria decision making (MCDM) method. In this paper, we use analytical network process (ANP) to determine the best alternative to work with, for this method allows considering the importance of a criterion in a cluster over another one in a different cluster. We get the view of an expert professional in 3PL sector to determine the clusters and the sub-criteria of this problem and the relations between them, and make pairwise comparisons. Three customers are determined as alternatives. At the end of our work, the alternative with the highest rank is the customer which the company should select. Keywords: logistics, customer selection, analytic network process 1 INTRODUCTION There are a few studies that deal with customer selection so far. Yorke and Wallace used the Pareto curve for that purpose [1]. Kim and Lee took the existance of network externalities into consideration [2]. Bhatnagar et al. dealt with customer priorizitation within the scope of customer relationship management (CRM)[3]. We focus on the customer selection problem in the logistics sector. 3PLS (3 rd Party Logistics Service) is highly demanded by customers of various sectors. Behind this demand lies profitibality and the need