Forecasting tourism arrivals with an online search engine data: A study of the Balearic Islands

Abstract

[eng] This study explores issues related to the forecasting in revenue management in the Mechanism of predicting tourism arrivals for the Balearic Islands. Specifically, the study uses queries from a web search data (Google Trends) in order to demonstrate the forecasting power of such measures compared to traditional methods. I developed a database formed by the two main tourist volumes, namely, Germany and UK and then, compare each model with its corresponding baseline to figure out whether the Google Trends indicator can increase accuracy of the prediction. After estimating the four different models, I selected the baseline model for Germany, and the alternative for UK. Consequently, Granger causality test indicated a positive causality between variables suggesting good estimating results. Besides, I calculated the Mean Absolute Percentage Errors (MAPE) for each model and the results showed a considerable improvement of the Google Trends models compared to baseline models. The results provide some hints for increasing company efficiency and enhance policy maker decision making.