Improving the predictive validity of NPS in customer satisfaction surveys

Abstract

Though widely used, NPS has been challenged for its doubtful validity in predicting loyalty behaviors and business growth. Its arbitrary assignment of promoters/passives/detractors is one of the factors impacting its predictability. This paper addressed the validity issue by taking into account the scenarios NPS is used: (1) NPS for an organization in developing stage with more of new customers vs. a well established business with more of regular customers; (2) NPS in online survey vs. offline survey. By using the data of an online NPS survey for an online supermarket and tracking the purchase behaviors of all respondents before and after the survey, the authors found that NPS works better for new customers than regular ones in predicting repeat purchase behaviors and changes of purchase volume and value. In addition, in online surveys, a polarized segmentation of promoters/passives/detractors could effectively improve the predictive validity of NPS for new customers. © 2014 Springer International Publishing.