The article presents a client loyalty model which may be used in the process of managing an optimum portfolio of clients by means of identifying and selecting key clients of a tourist enterprise. There are also described results of the author's own research allowing us to identify the key factors determining the occurrence of behavioral loyalty of travel agency clients. The analysis covered real data from the booking system of one of the Polish nationwide franchise networks consolidating 44

more »

... el agencies, referring to 3 923 tourist services reservations. In the study of the occurring relationships a non-parametric method of MLP (Multi-Layer Perceptron) neural networks was applied. The following variables turned out to be the most important determinants of behavioral loyalty: duration of the tourist event, number of purchased offers and time between the booking date and the date of tourist event commencement. The customer loyalty prediction model was used to carry out in-depth customer segmentation which, in turn, allowed the development of a loyal customer portrait. A loyal client of travel services is a family-oriented client, travelling for a period of two weeks or longer, accompanied by the family, very often with children under 14, planning and booking holidays well in advance, attached to the place of making the booking (i.e., the travel agent).