This study proposes a marketing decision support system (DSS) for firms using case-based and peak-end approaches to model consumers. The proposed DSS is composed of model and estimation methods. We employ a similarity function used in case-based decision theory to examine the degree of similarity between the past and current products offered to a consumer. The contributions of this study are as follows: First, by extending the peak-end approach, the proposed model could be utilized to analyse not only the same product but also multiple similar products. The DSS could be applied to a broad range of decision problems. Second, by extending the case-based decision model, our DSS considerably reduces the number of computational operations needed. Third, the model demonstrated the best fit among the compared models and possessed high prediction accuracy when analysing the viewing data for Japanese television dramas. The DSS could increase the future purchase probability of a product. Our research bridges two significant concepts: case-based decision models in DSSs and peak-end rules in behavioural economics.