Urban logistic companies suffer from complex supply chains with multiple facilities and checkpoints, which introduce high complexity to their operational planning. This is especially true on last mile trips, where these companies need to deliver hundreds of thousands of packages per day into urban areas. While planning the routes that will be used to deliver daily parcels, logistic companies use commercial solutions that aim to solve what is known in the literature as the Vehicle Route Problem (VRP). VRP solutions aim to optimize the paths of individual vehicles or a fleet of vehicles to reduce route costs, waiting times, emissions of pollutants and traffic congestion. As part of commercial VRP solutions, they analyze the distribution of past deliveries and create regions of delivery (or clusters), facilitating their route optimization process. In this work, we are proposing to segment these clusters of delivery in micro regions within the cluster, by estimating the cluster density. We use contours to classify central versus peripheral areas within the cluster, and run the route optimization separately in each of these micro regions. Finally, LoggiBUD was used to benchmark the proposed method against traditional clustering segmentation, comparing which of them has resulted in the shortest distance to deliver a set of parcels in a given day. Results have shown that the method proposed in this article can potentially reduce transportation costs.