With the development of tourism promotional campaigns based on new artificial intelligence techniques, it is important to know how to identify a visual that provokes a positive emotion from a visual that provokes a negative emotion. These emotions can be seen through the facial expressions of the tourist. In the field of tourism, several researches support the link between the positive emotion of a tourist and his degree of satisfaction in his experience. We propose in our study to confirm the potential of using facial expressions to reliably evaluate the performance of a web visual. From a managerial point of view, the interest of the study is for experts in the tourism sector to use facial expressions before launching their promotional campaign and to verify the relevance of the chosen visuals.