Estimating the daily rate of tourist apartments in a cultural tourism city

Purpose This research aims to estimate daily rates of tourist apartments (TAs) and identify their main determinants by comparing the hedonic price methodology with a model based on artificial neural networks (ANNs). Design/methodology/approach The empirical study was carried out in Seville, one of the main Spanish destinations for cultural tourism, where the market has seen a notable increase in this type of accommodation. ANNs using Statistical Product and Service Solutions (SPSS) and hedonic pricing method (HPM) methodologies are employed. Findings Size, number of occupants, scores from previous users, location (measured as the time spent walking to an emblematic place) and, especially, factors linked to seasonality and holidays in the city are shown as the most influential variables for the daily rate. Artificial intelligence techniques, due to the flexibility that characterises them, show a greater predictive power for daily rates than traditional hedonic models. Practical implications The application of this study can be useful for both the hosts and guests of these new types of accommodations, as well as tax offices, to determine whether a listed price is higher or lower than that normally observed in the market in the predetermined conditions acquired by customising variables in the hedonic and ANNs models obtained. Originality/value ANNs and HPM methodologies are compared in a TAs context.