When to "Fire" Customers: Customer Cost-Based Pricing

The widespread adoption of activity based costing enables rms to allocate common service costs to each customer allowing for precise measurement of both the cost to serve a particular customer and the customer's pro tability. In this paper, we investigate how the use of such customer cost information a ects a rm's customer acquisition and retention strategies, and ultimately its pro t using a two period monopoly model with high and low cost customer segments. While past purchase and cost information helps rms to increase pro ts through di erential prices for good and bad customers in the second period ( price discrimination e ect ), it can hurt rms because strategic forward looking consumers may delay purchases to avoid higher future prices ( ratchet e ect ). We nd that when the customer cost heterogeneity is su ciently large, it is optimal for rms to re some of its high cost customers, and CCP is pro table. Surprisingly, it is optimal to re even some pro table customers. This result is robust even when the cost to serve is endogenous and determined by the consumer's choice of service level. Our results shed insight on retention-acquisition dynamics, on when rms can improve their pro tability by selectively ring known old bad customers, and on replacing them with a mix of new good and bad customers.