In many online retail settings, customers are exogenously divided into different classes based on their observable attributes. This information is available upon arrival via customers’ login information or internet cookies. Although advances in data analytics have paved the way for the implementation of promotions targeted at the individual level, retailers have limited information on customers’ product preferences and willingness-to-pay values. Personalized (or targeted) promotion strategies require the estimation of product preferences and willingness-to-pay values by observing the customers’ purchasing decisions. Therefore, the retailer faces a trade-off between earning revenue immediately and learning customers’ attributes for future revenues. This study focuses on the estimation of customer attributes by utilizing purchasing decisions and explores the revenue impact of dynamic learning along with different promotion strategies. The student is expected to conduct an extensive literature review of the related streams of research and implement efficient solution algorithms existing in the literature.
Faculty Department of Project Supervisor:
School of Management
Number of Students:
Related Areas of Project:
Computer Science and Engineering