Smart retail is an important concept in smart cities, due to its potential to deliver competitive advantage to companies. As the retail sector is moving beyond locating products, in-store tracking technologies are used to deliver better services to consumers and acquire insights regarding their behavioral patterns. Retail stores are challenging environments with multiple sources of noise (e.g., shoppers' moving) that impede indoor localization and there are settings with new requirements that should be further examined. This paper has developed and deployed a low-energy Bluetooth Low Energy (BLE) beacon-based indoor positioning system that utilizes machine learning to improve its accuracy, enabling retailers to offer more effective customer location-based services and improve decision making in retailing.