An Approach for Optimizing Inventory for Quick Commerce Platforms

Quick commerce platforms face a dual challenge: firstly, they increasingly offier a wide assort- ment of items ranging from electronics to groceries. Secondly, while doing so, they guarantee fixed-time delivery for all products. This necessitates a sophisticated inventory management ap- proach as storage space is limited, and the platform is required to optimize both the item assortment and its quantity. Our study attempts to solve this problem using a multi-stage stochastic optimiza- tion model to determine the optimal SKU assortment for such a capacity-constrained dark store. The model first determines the optimal quantity for each SKU by balancing stock-out costs (lost revenue plus fulfillment penalties) against overstocking costs (wastage and storage). This analysis feeds into a higher-level assortment model that selects the most profitable set of SKUs to maximize total revenue. Such analysis is subject to constraints on storage, specialized capacity, and picking time. Finally, we demonstrate this approach using a simulation by formulating it as a Mixed- Integer Non-Linear problem (MINLP) and proposing an approximate solution using linearization techniques.