A mathematical model is developed to calculate the costs of alternative distribution set-ups for last mile transportation in a supply chain with small and fragmented volumes. The model is based on input from logistics cost models for urban areas combined with cost variables related to logistics processes, receiver attributes and local city characteristics. The cost variables for each aspect (logistics, receiver, city) influence the cost-effectiveness and applicability of an alternative distribution set-up. The model is applied on the delivery of fast moving consumer goods (FMCG) towards small independent retailers in a megacity. The current supply of these stores is characterized by high costs, inefficiency and unsustainability. Four different set-ups are modelled. The model shows the effects of different city and store request parameters. When drop sizes are low and distances are short, direct shipments with smaller vehicles outperform the current direct set-up. When drop sizes are low and distances are long, collaborating in an urban consolidation centre (UCC) shows a saving. The model can be further validated with data from other cities and other distribution set-ups.