66

• Research objectivesThis research examines the determinants of internet-user engagement with tourist destinations on social media. • MethodologyA quantitative study was conducted on a sample of 27 tourist destinations, a topic model (LDA) is used to identify the attractions highlighted in the content of messages, and regression analysis is performed to assess the relative influence of each determinant on engagement. • ResultsThe results show that a destination is characterised by four types of attraction (cultural, heritage, gastronomic, and natural attractions) and that internet-user engagement increases when natural attractions are highlighted. Moreover, the results show that informativeness of the message has a negative effect on internet-user engagement, whereas valence, association of the message with a visual, and imitation effects increase it. • ImplicationsThese results help to determine the type of attraction that generates the strongest engagement and suggest that positive, visual, and short (<50 words) messages should be preferred. Consideration of imitation effects can also be used by destination management organisations to develop follow-up strategies. • OriginalityThis research shows that internet-user engagement is determined not only by a message’s characteristics, but also by the visibility of the engagement of other internet users, through imitation effects.