Randomized trial assesses digital features predicting tourism readiness in wineries, suggesting a roadmap for improvement.
Key Points
This study aims to develop a predictive framework linking wineries' online features to their tourism orientation.
89 winery websites audited for 66 features to create a Tourism Readiness Index (TRI) and Proximity Index.
Predictive models, including bias-reduced Firth logistic regression and Random Forest, were utilized for analysis.
Rigorous robustness testing was conducted with FDR correction and validation against an importance-weighted TRI.
Experiential features like activity duration (OR = 14.36) and winery photography (OR = 9.52) strongly predict tourism readiness (p < 0.001).
The Random Forest model achieved an AUC of 0.96, demonstrating high predictive accuracy.
Digitally advanced but remote wineries performed comparably to centrally located wineries, indicating digital execution can offset spatial disadvantages.