This study aims to examine the comparative effectiveness of digital destination image sources on Instagram. Analysing content from destination management organisations (DMOs), influencers and user-generated content (UGC) across three island destinations, it also explores the use of artificial intelligence (AI) virtual influencers in tourism communication.
A mixed-methods design combines qualitative content analysis with econometric modelling of 189 Instagram posts from the Canary Islands (Spain), Hawaii (United States) and the Seychelles. Multiple regressions (OLS with clustered standard errors and Poisson models) ensure robustness. Analysis of over 6,000 comments provides qualitative depth.
Empirically derived patterns challenge conventional assumptions about harmonious co-creation in destination image formation. UGC is associated with engagement rates 28 times higher than DMO content, despite visual similarity (Jaccard = 0.735), suggesting that perceived authenticity may operate at the sender rather than the content level. Influencer posts are associated with the highest proportion of comments expressing visit interest (42.1%), with micro-influencers outperforming larger influencers in engagement. Convergence on primary destination attributes coexists with divergence on secondary themes.
Findings derive from island tourism destinations and Instagram; replication across other destination types and platforms is needed before broader generalisation. Behavioural intention is inferred from comment content rather than measured through validated scales. The virtual influencer component is exploratory (n = 5 cases).
To the best of the authors’ knowledge, the study provides the first systematic multi-actor comparative analysis of digital destination image effectiveness using consistent engagement metrics across island destinations. It proposes the concept of sender-dependent authenticity, in which audience engagement may respond to perceived sender identity rather than content attributes, and proposes a tension-aware dynamic model that positions DMOs, influencers and UGC along the control–authenticity and reach–engagement dimensions. Although empirically grounded in island tourism, the framework identifies mechanisms that may extend to other destination types, although empirical validation beyond island contexts remains necessary. The exploratory integration of AI virtual influencers into destination image theory anticipates shifts that existing models do not accommodate.
