Bayesian inference for the Net Promoter Score

TL;DR

This paper introduces a Bayesian method for estimating the Net Promoter Score, including sample size determination, with practical tools and an example from financial services, addressing a gap in statistical analysis of this measure.

Contribution

It develops a Bayesian framework for NPS estimation and sample size calculation, filling a gap in existing statistical methods for this metric.

Findings

Bayesian estimators provide accurate NPS estimates.

Sample size determination method improves study planning.

Practical computational tools facilitate implementation.

Abstract

The Net Promoter Score is a simple measure used by several companies as indicator of customer loyalty. Studies that address the statistical properties of this measure are still scarce and none of them considered the sample size determination problem. We adopt a Bayesian approach to provide point and interval estimators for the Net Promoter Score and discuss the determination of the sample size. Computational tools were implemented to use this methodology in practice. An illustrative example with data from financial services is also presented.