Tourism companies such as restaurants and hotels are a crucial part of the economy of many countries, including Switzerland. Therefore, the investigation of factors indicating success or failure of tourism companies is of interest for many stakeholders. In this study, we research the role of online ratings to predict business performance. Applying a random forest model to a large sample of hotels and restaurants, we find a high feature importance of attributes such as number of rooms, occupancy and online rating. We show that information extracted from online reviews, which are publically available, have a high prediction power in estimating revenue and revenue growth of small and medium sized tourism companies.


Time: 10:45 - 11:15
Track: Track 3
Speaker: Daniel Müller