Considering the importance of market segmentation as a marketing tool to determine promotional policies, this paper aims to contribute to the tourism literature using the two-level approach proposed by Vesanto and Alhoniemi (2000) as an alternative and effective method to conduct cluster analyses.

For this purpose, an empirical study was conducted interviewing tourists who visited three different Christmas Markets in Northern Italy. The two-level approach is based on two clustering techniques used in sequence: a Self-Organizing Map (SOM) followed by a clustering algorithm. The Silhouette index (Rousseeuw, 1987) is used as a guideline during the second level in the selection process of both the best clustering techniques (between hierarchical and non-hierarchical) and the best partition.

The analysis identified three cluster segments and this paper demonstrates the suitability of the clustering method adopted. In the discussion of the results, marketing and managerial implications are also highlighted.