Forecasting tourism demand with Google trends: Accuracy comparison of countries versus cities

Abstract Previously, Google Trends indices have been found to be useful in improving the tourism demand forecast accuracy relative to a purely autoregressive baseline model. The purpose of this study is to extend previous research in terms of comparing the forecasting accuracy of cities and countries using Google Trends Web and image indices. The study compares forecasting models with Web and/or image search indices regarding 2 cities (Vienna and Barcelona) and 2 countries (Austria and Belgium). Overall, the forecast accuracy of Vienna with the Web and/or image indices was the best among the 4 destinations, followed by Belgium, Barcelona, and Austria.