We measure the links between store attribute perceptions and customer satisfaction, and between customer satisfaction and sales performance, in the food retail sector. Our data set consists of six waves of customer satisfaction and sales information for about 250 stores over the period 1998- 2001 for a publicly held supermarket company. We construct a statistical model to address nonlinearities and asymmetries in the satisfaction-sales performance links, and we illustrate how food retailers can affect store revenues by managing customer satisfaction. Contributions of our study include the analysis of behavioral consequences of customer satisfaction in the food retail sector, the measurement of complexities of the satisfaction-sales performance links based on an empirical model of first differences, and a discussion of how managers can use such results for customer satisfaction policies.