Current study aims to prioritize SQDs for hotels through a two-step procedure: firstly, forming congruent groups of customers based on their level of expectations as well as perceptions of service; and secondly prioritizing SQDs for each group and analysing the results. To reach that aim, the customers of three 4-star hotels were targeted and asked to report on the quality of the service they expected to receive prior to its delivery as well as their feelings toward the service after they received it. Then, using Data Envelopment Analysis we discriminated the majority of customers, who were satisfied easier (ordinary customers), from the ones who were harder to please (fastidious customers). Finally, we exploited TOPSIS and Shannon Entropy to prioritize SQDs for each group. The approach taken in this paper, which first distinguishes two groups of fastidious from ordinary based on their expectations and perceptions, is rather noble. Findings revealed that the two groups do not value different dimensions of service quality in a similar way. The results of this study can provide insightful guidance for the hoteliers to customize their service and exploit their resources more efficiently. Also from theoretical perspective, the reported difference of the results for the two groups may partly explain the reason why there exists inconsistency in the literature and the findings are mixed.