The Robotic Mobile Fulfillment System (RMFS) is widely used in e-commerce warehouses and includes pods, storage locations, workstations for order picking or replenishment, and mobile robots. Decision-making in warehouse systems can be strategic, tactical, or operational. Product assignment, a tactical decision, significantly impacts picking efficiency. This research concentrates on SKU-to-pod allocation, the preliminary step before simulation, involving two phases: product grouping and product combination. Effective product grouping can enhance picking efficiency, while product combination in pod allocation aims to optimize the units picked per pod, termed pile-on, thereby reducing the reliance on Automated Mobile Vehicles (AMVs). Three scenarios were examined: Random Baseline, Class Combination, and Cluster Combination. Class Combination employs ABC classification to sort SKUs into classes using Pareto’s principle, correlating SKU percentages with order frequency percentages. In contrast, Cluster Combination considers product dimensions for pod placement. Simulations determine the optimal pile-on by evaluating the units picked per pod visit to the pick station, with a higher unit count per visit indicating reduced mobile robot transport and increased efficiency. The simulations revealed that Cluster Combination, the final scenario, achieved the best pile-on, with improvements of 30.82% and 8.19% over the first two scenarios, respectively. These results were validated using one-way ANOVA.