The measurement of service quality and customer satisfaction has received considerable academic attention over the past twenty-five years. Companies spend hundreds of millions of dollars and thousands of people hours trying to understand and influence customers, and often assume (incorrectly) that satisfied customers are loyal ones. Simply understanding customer satisfaction is not enough – researchers and managers must also examine how satisfaction influences customer behavior. This paper presents a new methodology for understanding customer satisfaction and its relationship to customer behavior through the use of discrete choice modeling. Using an empirical example of customer satisfaction, the post-paid cellular service in Western Canada, the results of this new model are compared to two traditional methods of modeling customer satisfaction data – exploring the differences between attribute performance and attribute expectation, and uncovering the relationship between attribute satisfaction and overall satisfaction. The results of the choice model are compared and contrasted against traditional results, and are found to be strategically and tactically useful.