In this paper, a new multi-criteria procedure is devised for new product devel-opment decision-making made from survey data. Groups of panelists evaluateseveral product categories regarding different criteria, each one through a spe-cific qualitative scale, which ultimately will guide decision-makers to develop anew product in a specific category. These qualitative scales are equipped withordinal proximity measures that collect the perceptions about the proximitiesbetween the terms of the scales by means of ordinal degrees of proximity. Thelinguistic assessments provided by panelists are compared with the highest termsof the corresponding qualitative scales. In order to aggregate the obtained or-dinal degrees of proximity, a homogenization process is provided. It avoids anycardinalization procedure in the ordinal proximity measures associated with theordered qualitative scales used for assessing the alternatives regarding different criteria. Products categories are ranked taking into account the medians of thehomogenized ordinal degrees of proximity.
