Radio frequency identification (RFID) is an information facilitator that can directly improve decision‐making; thus many retailers and suppliers have adopted it. A vast amount of RFID data streams has been gathered, however, it remains unutilized, or it is been exploited solely for inventory count purposes. This research proposes a way to analyze the immense volume of RFID data reflecting the behavior of products in retail stores, in order to produce information for inventory availability and inventory flows at different stages of the supply chain. We propose an RFID data analytics artifact that transforms RFID data captured in retail stores to the flows of the inventory/ products between locations in the stores. By mining the RFID data streams, we reveal the flow patterns of the products; these patterns correspond to the frequent product paths in the stores, and we provide them to the retailers in a visual manner. This unprecedented knowledge is valuable, because it can enable decisions ranging from shelves space allocation, dynamic pricing programs for slow-moving fresh products to product assortment. Furthermore, to testify artifacts’ correctness and usefulness, we have put it in practice, using real data provided by an Italian fashion retailer, in order to show how it can really support such decisions.