Importance–performance analysis (IPA) is very useful for practitioners to improve service quality and increase customer satisfaction by identifying the major strengths and weaknesses. A survey is typically conducted to gather the needed information about importance and performance. If the survey does not include the information about the importance, this study introduces two information-based methods, namely entropy and mutual information methods. These two methods can be applied in either ordinal or metric data without the assumption of data to be Gaussian or symmetric distributed. Moreover, mutual information method uses the objective function to reduce the uncertainties when the outcome of the dependent variable is predicted where the dependent variable is the overall satisfaction. In this study, integrated frameworks of these two methods are proposed, and each framework consists of four steps to describe how the information of importance can be derived when the matrix of importance–performance analysis is to be constructed. Finally, a case study of Hi-Life convenient store in the university is illustrated to examine the service quality by these two proposed frameworks when the importance is unavailable directly from the survey.