Supply network of any manufacturing product is generally multi-tier in nature. Suppliers to manufacturer are called Tier-I suppliers, and suppliers to Tier-I supplier are referred as Tier-II suppliers. In general, decisions on order quantity allocation at each level of sourcing are independent. Thus, overall cost performance of supply network may be sub-optimal in presence of dependency. In this paper, considering dependency, a generalized solution approach is proposed for a single product sourcing. The approach consists of two steps i.e. supplier selection and order quantity allocation at different stages. Mixed Integer Nonlinear Programming is used to determine solution assuming multi-echelon inventory policy. Specific cost advantage metric considering independent decentralized and a multi-tier centralized decision policy is compared and contrasted. A full factorial simulation experiment is used to identify key decision variables and associated interactions.