Aggregated models, such as spatial interaction (SIM) models are widely used in location analysis. Despite their popularity, there are certain limitations to their use. In particular, the method struggles to account for the passing-by population and multi-purpose trips of retail clients, temporal changes in accessibility and some bottom-up processes potentially important for services. Agent-based ... [Show full abstract] modelling (ABM) is a promising technique that attempts to address all these problems. However, it still lacks examples of real-world applications. This article aims to provide an example of how hybrid ABM (H-ABM) can be built on a SIM foundation, by incorporating most of its ideas, such as distance-decay function, facility attractiveness parameters and demand elasticity. The author aligns the two models as close as possible and compares their input data, calibration procedures and results. In the final analysis, the hybrid agent-based model proved to be more realistic because it incorporated the time-space variability of supply (i.e., limited numbers of available places in swimming pools), demand (the popularity of certain entry hours) and transport (traffic jams during rush hours). The spatial interaction model was much faster to execute and turned out to be more convenient for more straightforward applications, which do not require detailed data concerning individuals.