This paper examines customer satisfaction with a hypothetically new broadband service using a two-stage stated choice experiment. The first stage of the experiment introduces the new broadband service. After indicating a choice between the two services respondents are given a hypothetical product experience with one of the services and satisfaction is measured. The satisfaction response is analysed using binary logit to determine the impacts of expected performance, actual performance and disconfirmation on customer satisfaction. Results suggest disconfirmation of expectations is important in determining customer satisfaction /dissatisfaction as well as the price of the experienced alternative. Interactions between disconfirmed attributes and other attributes also are significant overall.