Last-mile distribution in urban areas is challenged by congestion and restriction for motorized traffic. To support operations, this work investigate the impact of operating urban consolidation points and facilitating cargo-bikes for urban last-mile distribution. Motivated by sample setting originating from the food delivery industry, a decision support system combining agent-based simulation with heuristic optimization procedure is developed. It considers a logistics provider who performs the last-mile delivery for multiple competing restaurants in an urban area. Therefore, both demand and the availability of cargo-bikes, which are operated by freelancers, are subject to randomness. Computational experiments investigate the impact of the available amount of cargo-bike drivers as well as the number of operated consolidation points, highlighting the importance of facilitating simulation models to support operations in highly dynamic and uncertain settings.