Hotel customer segmentation using the integrated entropy-CRITIC method and the 2T-RFMB model

Abstract

Customer segmentation helps the company better understand its target audience, which is vital to optimizing marketing strategies and maximizing the customer value for the company. This paper improves the original RFM model by including the potential loss to the hotel from a customer canceling their reservation in the indicator “Monetary” and adding a new indicator “Bonding” to indicate the degree of customer bonding with the hotel. The proposed model also includes the 2-tuple linguistic model to give hotel managers or decision-makers more easily understandable customer segmentation results. The aggregation of the four indicators (recency, frequency, monetary, and bonding) into a unique value is a Multi-Criteria Decision-Making (MCDM) problem. To generate the weights that can consider the relationship between various indicators and the level of data diversification contained in each indicator, the Entropy method and the CRiteria Importance Through Intercriteria Correlation (CRITIC) method have been integrated. Customer overall values are generated based on the 2T-RFMB model and the integrated Entropy-CRITIC method. Finally, various customer segments are obtained with K-means clustering. This proposal has been evaluated by a real dataset from a hotel in Lisbon. The results show that the proposed model can increase the linguistic interpretability of clustering results. It also demonstrates that the proposed model can provide hotel managers with more realistic customer values to assist them in allocating their Customer Relationship Management (CRM) resources efficiently.