Market Research: Foundations to Practice · Measuring Customer Sentiment
NPS's Real Methodological Criticisms
NPS's popularity and its methodological soundness are two separate questions, and a genuine academic literature has raised serious challenges to the second one. This chapter covers the substantive criticisms directly, not dismissively.
A team of researchers led by Timothy Keiningham set out to independently replicate Reichheld's original methodology and, according to their published findings, could not reproduce the claim that NPS is 'the single most reliable indicator of a company's ability to grow,' directly challenging the strength of the original 2003 claim rather than merely offering a milder methodological quibble around its edges [31].
Key Takeaways
- Keiningham's team could not independently replicate Reichheld's original claim that NPS is the single most reliable predictor of company growth.
- The calculation itself is critiqued for excluding Passives entirely, discarding real information about a distinct, meaningful customer segment.
- Academic critiques have questioned whether an 11-point scale is optimal and argued that multi-metric regression models predict outcomes more accurately.
- NPS's massive popularity should not be mistaken for settled methodological proof; even its architect acknowledges real problems with how it gets abused in practice.