The distribution of Net Promoter Score in socio-economic surveys

Inmarketingstudiesdevotedtocustomersatisfactionandconsumers’loyalty analysis, Reichheld (2003, 2006) proposed the Net Promoter Score (NPS) to synthesize by means of an excess index the distribution of the sample responses to a question as: “How likely is that you would recommend our Company/Institution to a friend or colleague?”, on an ordinal scale from 1 to 10. More specifically, this measureisobtainedbythedifferencebetweentheproportionof“enthusiastics”minus that of “passives”. This index may be fruitfully exploited in different research fields, of course, where the whole set of information is meant to be summarized by comparing the relative frequencies of “supporters” and “detractors” with respect to products, services, items, etc. Although the literature remarks critical and positive aspects of such an index, only recently Rocks (2016) are faced with inferential procedures with regard to NPS. In this study, we search for the distribution of NPS based on a convenient structure of the response patterns. Indeed, we assume a parametricmixturefortheresponsesandverifythebehaviourofNPSovertheparameter space.