Customer satisfaction in the presence of imperfect knowledge of data

Ensuring customer satisfaction is at the top of any manager's agenda wishing to keep or raise profits. To do so, conducting regular surveys is quintessential. However, provided the large number of questions that typically constitute those surveys, resulting data are plagued with imperfect knowledge. Data from these surveys are typically expressed in ordinal scales. No mathematical operation like addition or multiplication is allowed in these cases. Recognizing the limitations of existing alternatives in the literature to model imperfect knowledge, this work proposes an alternative model for the study of satisfaction and its determinants. Our approach is based on merging the Stochastic Multicriteria Acceptability Analysis and the Multicriteria Satisfaction Analysis (model). We explore the potential of this approach and apply it to an empirical case related to hospitality.