Purpose Online reviews (OR) play an important role in consumer decision-making in the hospitality sector. OR research has expanded significantly over the past two decades and continues to evolve with new conceptual and technological developments, including artificial intelligence (AI)-generated and multimodal reviews. This study provides an updated and comprehensive literature review that maps major themes and captures recent progress in OR research. Design/methodology/approach This study conducts an integrative review of over 190 peer-reviewed articles published between 2000 and 2024 across more than 50 journals. Thematic mapping and conceptual integration are used to organise the literature and develop a coherent understanding of online review research in hospitality. Findings This review highlights three dominant themes in hospitality OR research: review authenticity, review helpfulness and consumer purchase intention, aligned with the stimulus-organism-response framework. The synthesis provides the conceptual basis for the empirically testable framework proposed in this study. Research limitations/implications The synthesis offers a strong basis for theoretical refinement and future empirical testing. For practitioners, the review highlights how authenticity cues, multimodal signals and emerging AI-generated content shape consumer decisions, helping hospitality managers design more trustworthy and effective review environments. However, the findings are limited by reliance on the Scopus database and exclusion of non-English language studies, which may constrain their generalisability. Originality/value This review provides a timely update to hospitality-focused online review research by consolidating two decades of work into a coherent and empirically testable framework with clearly defined constructs and propositions. It advances theoretical understanding by clarifying the roles of authenticity, helpfulness and purchase intention and by incorporating contemporary developments that influence how consumers interpret and rely on OR.
Singh et al. (Thu,) studied this question.