Do Preferences Come First? The Impact of Smart Recommendations on Tourists’ Purchase Intentions

This study investigates the impacts of smart recommendations adopted by online travel agency platforms on tourists’ purchase intentions. Prior research has advocated recommendations based on consumers’ own preferences (i.e., preference-oriented), whereas in the tourism context, a diverse range of popular destination services may be more appealing to tourists (i.e., destination-oriented). On the basis of the stimulus–organism–response framework and valence theory, we developed a perceived value–risk analytical model to reveal the effects of different recommendation algorithm logics on tourists’ purchasing decisions. One pre-study and three experimental studies were conducted to test the proposed hypotheses. This study revealed that approaches of smart recommendation algorithms (i.e., preference-oriented versus destination-oriented) influence tourists’ purchase intentions such that tourists exhibit greater willingness to purchase tourism services recommended by destination-oriented recommendation algorithms than by preference-oriented recommendations; in this influence process, perceived privacy risk and information richness are key factors. In addition, this study revealed that tourists’ privacy concerns negatively moderate the effect of approaches of smart recommendation algorithms on perceived privacy risk and that brand familiarity negatively moderates the relationship between information richness and purchase intention. This investigation contributes to the research on smart recommendations and provides theoretical and practical implications for tourism service marketing and technology management.