Date: 22^nd November, 2019

Time: 10:00 – 11:30 a.m.

Venue: Room1102, Administration Building, Zijingang Campus, Zhejiang University.

【Speaker Introduction】：Prof. Xiaojing Dong is a tenured Associate Professor of Marketing and Business Analytics, and the Faculty Director of Master of Science in Business Analytics at Santa Clara University. Her research area applies Data Analytics techniques (Bayesian and classical) to develop model examining how company marketing actions influence customer decisions, and therefore provides suggestions on improving Marketing and Business decisions. Her studies span multiple industries, including social network, online retail, pharmaceutical, travel, among many others. Her papers have appeared in top academic journals in business and top computer science conferences in Big Data and AI. Some of her research papers have attracted media attentions and been widely cited. She has three patents pending in the US, in the area of modeling, forecasting and field experiments. 

【Lecture Introduction】：In online shopping platform, not only the purchase information, but also the clicked pages can be conveniently recorded. It is a common practice to examine only those purchased when inferring the consumer’s preference and tastes. In this study, we proposed a structural model approach to dive deeper into such question leveraging the click stream data. A structural model approach allows us to examine the decision process at individual consumer level, and obtain inferences for each consumer, for his/her preference and search cost. 

A consumer’s preference and search cost would play critical roles for an online retailer when selecting a product to recommend and when choosing targeting strategies. A consumer with higher search cost will have a lower probability to explore the products that are not recommended. 

Using a unique data-set from an online retailer that contains panel information on consumers' search and purchase behavior, our proposed approach demonstrated that when ignoring search costs, we overestimate consumers’ preference heterogeneity by 40%. Using the example of personalized pricing, we show that this bias has important consequences for targeted marketing. Finally, we show that panel dimension of the data is crucial for identifying preference heterogeneity separately from search costs.