Tourist overnight stays are difficult to estimate properly due to the substantial gaps in the official data and the multiple types of accommodation. This trend has been accentuated with the emergence of new forms of accommodation resulting from the development of information and communication technologies and collaborative economies. Several studies have highlighted the importance of quantifying these accommodations because of their social, economic and environmental consequences for tourism destinations. This article explores the use of mobile data analysis as a complementary approach to official statistics to quantify the tourist overnight stays in the Walloon region. The mobile data highlight several aspects of tourism that are ignored by official statistics and raise questions about the underestimation and the current definition of a tourist. However, important limitations to this approach have been identified and are discussed. • Official statistics greatly underestimate tourist overnight stays. • The traditional definition of a tourist is no longer adapted to new forms of tourism. • Mobile data analysis provides a complementary approach for measuring tourism activity. • The paper proposes a methodology for locating visitors' overnight stays using mobile data analysis. • Knowledge of the local context is necessary for monitoring tourism activity with mobile positioning data.