Wine tourism quality perception and customer satisfaction reliability: the Italian Prosecco District

ABSTRACT

Recently, we witnessed a growing interest in those tourist experiences that focus on the enhancement and promotion of typical local cultures, in particular related to winemaking traditions. In Italy this evidence is also demonstrated by the emergence of events and tours linked to the promotion of Italian wine. It is therefore necessary to pay particular attention to the degree of satisfaction expressed by visitors with regard to the services currently offered within the framework of their wine tourism experience. The present research applies a study on customer satisfaction in the area of the Prosecco district, in the Veneto Region in Italy. The aim of the paper is to understand, by the use of a mixture model denoted as the Covariate Uniform Binomial Model, whether and to what extent the high degree of quality perceived by respondents depends on factors related to personal experience rather than on the circumstances in which the interviews were conducted. Results show that socio-demographic variables have a great impact on wine tourism behaviour. As a consequence, the possibility of creating segments of consumers based on socio-demographic characteristics and consumer behaviour shows the need to consider different marketing policies to satisfy the different needs of tourists that emerged.

KEYWORDS:

• Italy
• Prosecco District
• quality perception
• wine tourism
• cub model
• customer satisfaction reliability

ORCiD

Christine Mauracher http://orcid.org/0000-0002-8913-3429

Isabella Procidano http://orcid.org/0000-0001-9129-8807

Giovanna Sacchi http://orcid.org/0000-0002-9286-7810

Notes

1. Hereafter, according to the most recent literature on these issues, the acronym CUB (Covariate Uniform Binomial) will be used to indicate a MUB model without and with covariates.

2. If the covariate is a dummy variable, it must be said ‘moving from D = 0, the reference group, to D = 1’.

3. G has been used to compare two nested models, whereas AIC and BIC allow the general comparisons of models.

4. The sample size of the selected variables is of at least 150 respondents. There are not missing values.

5. As already stated, the number of income earners was considered as a proxy of income, in order to avoid the problem of missing data in Income variable.