Abstract: Firms in many industries (including telecommunications, financial services, newspaper and magazine subscriptions, and Internet-based commerce) can recognize their previous customers and can thus price discriminate on the basis of purchase history. Advances of information technologies ease collection of customer data, and data-driven marketing further promotes adoption of BPD. Not surprisingly, such behavior-based price discrimination (BPD) has drawn growing research interests over the past decade or so. Consumers' valuation uncertainty is introduced, and it is assumed that consumers do not observe their values of each product before purchase when the firms practice BPD.Once the true values are determined, their purchase decisions made under such uncertainty may lead to post-purchase regret. So they will consciously anticipate the possible regrets before decision making, i.e., anticipated regret. To investigate the effects of customer anticipated regret on dynamic price competition, behavior-based price discrimination (BPD) in a two-period experience good duopoly with a discrete value distribution is considered. First, adynamic game model between the consumer and both firms is established. Furthermore, the equilibrium configuration when the firms practice BPD is summarized in two types of regret aversion:(a)r_S>r_R=0 and (b)r_R>r_S=0,which can provide firms with guidelines about how should the firm use behavior-based pricing (BBP) to discriminate between its own and competitors' customers in a competitive market and how to most profitably manage the consumer's anticipated regret. The results of the study show that:(1) the consumer's anticipated regret can significantly impact his purchase decision byaffectingconsumer'sprice sensitivity; (2) the firms should reward repeat customers when the segment of consumers who realized a high value in period 1 is relatively small and when the high-low value difference is high and when the consumersswitching regret aversion is very weak or repeat-purchase regret aversion is relatively strong and reward switchers otherwise; (3) the consumer's anticipated regret has non-monotonic effects on firms' profits. Finally, some numerical examples are present to investigate the impacts of consumers' anticipated regret on the retailer'sprofit and prices to repeat customers and switchers. The results of this study can help managers develop better price discrimination strategy and marketing policies.

Key words: anticipated regret, behavior-based price discrimination, valuation uncertainty, price competition