To Each Their Own: Personalized Product Positioning and Competition

This paper studies a model of competition between two firms that offer personalized product positioning to consumers with horizontally distributed tastes. Firms have private, imperfect signals of each consumer's ideal location and offer each consumer a personalized positioning and price depending on that signal, without observing the competing firm's personalized offering. We characterize the equilibrium personalization strategy and examine how the accuracies of firms' signals affect equilibrium strategy, profits, and consumer welfare. We find that a competing firm charges a higher price and earns a higher profit for a more niche positioning unless the firm's prediction accuracy is sufficiently low. When both firms have access to the same industry-level prediction accuracy, we find that the average price does not depend on accuracy, that firms charge a higher price when offering a more niche positioning, and that the average level of differentiation first increases and then decreases in the prediction accuracy. Interestingly, equilibrium profits also have an inverse-U shape in the prediction accuracy. A higher accuracy can decrease welfare for consumers with very mainstream tastes. When firms can endogenously invest in prediction accuracy, firms over-invest in equilibrium which results in a prisoner's dilemma. In such a case, firms can benefit from industry-level self-regulations that restrict their ability to predict individual consumer preferences. The paper also discusses what happens if firms charge subscription pricing or if consumers' ideal locations are distributed on the Salop Circle, highlighting price discrimination between mainstream and niche consumers as the key driver of our model.