The significant growth in e-commerce sales has brought major challenges to logistic activity planning. Although several studies have analyzed the cost of transporting goods, only a few have addressed the multiple layers in logistics distribution. This chapter investigates the order delivery planning problem (middle mile and last mile) inspired by an e-commerce company in Latin America, aiming at reducing transportation costs. First, we model a network where the last mile is through direct delivery. This network is instrumental, for example, when the delivery region has low demand and the customer requires express delivery. A second model considering routing delivery in the last mile is also introduced. A first experiment with the proposed distribution-transportation models presents an analytical investigation considering customers within low-demand regions. This experiment is to assess the trade-off between transportation costs and delivery time. A second experiment involved scenarios in denser regions. The scenarios are divided into three sets: (i) assigning all customers to a distribution center (full assignment), (ii) assigning some customers to a distribution center (partial assignment), and (iii) assigning customers to any distribution center. The results showed that routing delivery significantly reduced transport costs, reaching the mark of 83% in some scenarios compared to direct delivery.