With the current growth of e-commerce, couriers need to serve a bigger demand and it is often the case that receivers are not at the location when their parcel is delivered. The use of parcel lockers is seen as a good option for improving the problem of failed deliveries. However, implementing a parcel locker network has an impact on both couriers and receivers, which still prevents couriers from directly adopting this technology. The aim of this research is to develop a new tool that helps couriers analyse the impact of using a parcel locker network taking also into account receiver preferences to pick up parcels from lockers. An optimization model is combined with a choice behaviour model to determine the trade-offs between delivering to parcel lockers or to end-users, in terms of service level and total transportation costs. In order to solve this model for large instances, an Adaptive Large Neighbourhood Search algorithm has been adapted with operators and criteria specific for the problem presented. Computational experiments were carried in a set of the Solomon instances and show that the problem can be solved efficiently with the proposed algorithm. Moreover, the algorithm was applied in a small delivery area in Rotterdam, showing the sensitivity of the model to the preferences of receivers in the area.