Consumer purchasing behaviour is often influenced by numerous factors, including the visibility of the products and the influence of other customers through their own purchases or their recommendations. Motivated by trial-offer and freemium markets and a number of online markets for cultural products, leisure services, and retail, this paper studies the dynamics of a marketplace run by a single firm and visited by heterogeneous consumers whose choice preferences can be modelled using a Mixed Multinomial Logit. In this marketplace, consumers are influenced by past purchases, the inherent appeal of the products, and by how visible each product is, which depends on the firm ranking policy. The resulting market generalises recent models verified by social experiments. Professor Berbeglia, jointly with Franco Berbeglia and Pascal Van Hentenryck, examine various ranking policies for this market and analyse their long-term dynamics and the potential benefits of social influence. In particular, the team shows that the heterogeneity of the customers complicates the market significantly. They found that, in markets where consumers have very different product preferences, the display of past purchases can be detrimental to rate of purchases as consumers could become confused as to which products to try. The team then explores a market segmentation strategy and quantifies its benefits in the long run. They show that the market segmentation strategy always benefits from social influence and that it significantly outperforms other ranking policies when the consumers have a large preference diversity.
— Joint work with Franco Berbeglia and Pascal Van Hentenryck —

Speaker Bio

Gerardo Berbeglia is an Assistant Professor of Operations at Melbourne Business School, University of Melbourne. He teaches in the MBA and in the Master in Business Analytics programs. After completing his PhD at the Université de Montréal, Gerardo was a Senior Scientist at ExPretio Technologies Inc. doing pricing optimisation. He later joined the Department of Mathematics at McGill University as a NSERC postdoctoral fellow. His current interests include revenue management, transportation problems, pricing optimisation, and more recently, the analysis of quantitative models of consumer behaviour under social influence.

For more information about the ESD Seminars Series, please contact Ying Xu at xu_ying@sutd.edu.sg.