Predicting hotel selection : an analysis of demand and supply perspectives

Abstract

Given the intensified competition to attract today's increasingly discerning hotel guests, it is vital for hotel managers to understand the set or bundle of hotel attributes which are demanded by the potential traveler. Unfortunately, marketing research in hotels is still mainly confined to the analysis of guest related data collected during registration or from guest comment cards. These data, however, has little merit as predictors of customer behaviour and cannot answer the "what if' questions. In recent and past research in hospitality and tourism, researchers have demonstrated the use of product attributes in predicting consumer preferences and segmenting markets (Bell and Morey 1997; Kim 1996). In undertaking this approach to behavioral study, many researchers in the Field have been attracted to use conjoint attribute part worths-the judgmentally quantified attribute values derived from conjoint modeling. This study attempts to utilize the compositional self-explication model as a reference model to assess the comparative superiority of the conjoint model in predicting clientele preferences among travelers to Hong Kong. It also seeks to reveal subgroup preference variations by using two segmentation approaches. In addition, the study of international travelers' hotel preferences is extended to the perspective of hotel managers, to examine their perceived hotel preferences of international travelers. The "gap(s)" between the hotel preferences sought by international travelers and those perceived to be sought by them from the perspective of hotel managers is estimated and discussed. Three hundred international travelers participated in the self-explication measurement and conjoint experiment in Hong Kong. Respondents were invited to complete a two-sage rating task on hotel attribute importance as well as to rank two sets of hypothetical hotel profiles - to obtain their preference structures for the above two kinds of preference modeling. A similar measurement approach and experiment was replicated with a sample of twenty hotel managers. The relative superiority of the two models was assessed by the percentage of correct choice predictions and Spearman rank order correlations. A series of market shares simulation were demonstrated. The "a priori" and "post hoc" segmentation were conducted to discover the variations of consumers' hotel preferences. The "gaps" of hotel attributes sought revealed by international travelers and those revealed by hotel managers were evaluated by direct attributes comparison, simulated market shares comparison and analysis of hotel managers' perceptions. The findings of this study revealed that the conjoint modeling technique has useful applications in hospitality and tourism, and suggested that variants of the basic model and alternative conditions be tested as well. It was evident from the results that the hotel choice decision of the traveler varies according to the benefits sought and the selection made is influenced by all situational, geographic and behavioral dimensions. A "gap" between hotel attributes sought by international travelers and those perceived to be sought by hotel managers in Hong Kong was also established. The policy implication of the Findings is that hotel managers in Hong Kong should make use of the preference variations in and influences on hotel choice decision along with the magnitude of the "gap" identified to formulate strategy for the provision of better product and services to international travelers.