A novel classification method for customer experience survey analysis

Abstract

Monitoring of Customer Experience (CX) is based on surveys which include metrics like the Net Promoter Score (NPS) and the customer satisfaction on experience attributes (e.g., call center, website, network coverage, service quality, tariff plan, etc.). The objective of mobile network operators is to maximize NPS through the improvement of the most important CX attributes. However, there is often a lack of clear and accurate association between NPS and the CX attributes. In this paper, we address the aforementioned deficiency based on grouped regression model of increased statistical fit, which in turn enables the employment of state-of-the-art machine learning algorithms in order to define the most important features that maximize the NPS index.