Accurate prediction is a key basis for making the proper decisions within the management of any accommodation facility. The prediction of the quantity demanded thus represents one of the basic knowledge of a solid management system. The paper focuses on discussing and presenting the possibilities of using individual prediction techniques based on time series analysis. The SARIMA models, which in their modified form provide an appropriate basis for the application, seem to be a suitable element for the needs of accommodation services. The correct setting of the input values (p, d, q) (P, D, Q) can then be solved by using the presented algorithm based on the Akaike information criterion.