In many fields of the socio-economic analysis, the really interesting information consist of two parts: the absolute level and the deviation with respect to an “expected” value. Deviations are also called gaps. Dealing with one or few indicators, the gap analysis can be faced with classical statistical methods. However, in the case of complex phenomena (described by many variables) a descriptive multidimensional approach can offer better results. In this framework, the paper presents a multidimensional approach to look at the gap as a whole concept. In particular, the attention is focused on the Customer Satisfaction analysis, where gaps are defined by the perceived and the expected satisfaction levels and are coded as interval-valued variables. The paper proposes a distance measure between intervals and a suitable factorial treatment to analyze and visualize the gaps over factorial planes.