This paper examines the link between a supplier’s marketing and service operations and its business customers’ subsequent repatronage behavior. We develop a dynamic model of service contract renewal for an individual firm, at the contract level, recognizing interdependencies among service contract renewal decisions due to the firm’s purchase of multiple contracts from the same supplier. The decision to renew a service contract is modeled as a function of service quality and price, where service quality is measured by the supplier’s service operations metrics over time. By incorporating longitudinal data about the supplier’s service operations, this study investigates how average service levels, variability in service levels (especially extreme outcomes), and timing of service delivery influence firms’ service contract renewal decisions. The study context is support services for high-technology systems in business markets in Germany and the United Kingdom, where service operations metrics over time typically have skewed distributions. Firm behavior is represented by a binary choice model at the contract level, estimated as a binary response model with a complementary log-log link function incorporating random intercepts. The study shows that a firm that has a few extremely favorable experiences for a given service contract is more likely to subsequently renew that service contract, after controlling for average service levels. Firms weigh recent experiences (i.e., within the past year)—rather than earlier experiences—when deciding whether or not to renew, so the timing of service experiences may be critical to the survival of buyer-seller relationships. Overall, the study suggests that models of customer retention should incorporate the extent, variability, and timing of a supplier’s service delivery over time at the contract/product level.
