Establishing warehouse policy parameters is a challenging task that can affect, among others, the retailer’s profit. Very few scientific publications deal with the impact of warehouse policy on customer trust in the retailer. The paper focusses on two warehouse policy parameters: rbs (retailer’s base stock level size) and dbs (delivery batch size). The experiments described in the paper ask the question of what values of the parameters provide the retailer with maximum net profit and, at the same time, allow to achieve the assumed level of customer trust in the retailer. To obtain the answer, a multi-agent simulation was used, implemented in the NetLogo environment, in which an appropriate software tool was developed. This tool allows to create simulation models and analyse parameters related to deliveries to the retailer and warehouse policy. The research conducted for the analysed case study determines the optimal values of rbs and dbs, which ensure maximum net profit and maintain a high level of customer trust. The results also show nonlinear relationships between dbs and the retailer’s profit and customer trust. The results obtained and the developed tool are important for managers and decision makers in the retail sector, who can use them to adjust their inventory management policy. Research can be continued in many directions. One of them is to consider variable delivery time.