Reichheld claimed that the net promoter score (NPS) could predict the growth and health of a company. As it turns out, net promoter scores may not predict growth at all, and the net promoter question is no more accurate than any other customer satisfaction survey question.

The first time that a researcher outside of Reichheld and his Satmetrix team rigorously tested and publically refuted Reichheld’s net promoter score theory was in 2007. Timothy Keiningham, Senior Vice President of Ipsos Loyalty, examined a wide range of sectors and companies and found that the relationship between net promoter score and revenue was inconsistent at best. The relationship varied widely across industries, and for many, NPS and revenue seemed to have nothing in common. Some sectors, such as retail gasoline, actually demonstrated negative correlations, meaning that some companies with high net promoter scores made less than those with low ones! Even within industries, the relationship between NPS and revenue fluctuated dramatically.

Keiningham published his study in the Journal of Marketing concluding that “net promoter in no way fulfills Reichheld’s claim for being the ‘single most reliable indicator’ of a company’s ability to grow.” In the same year, D. Randall Brandt, Ph.D., Vice President of Maritz Inc., agreed that the net promoter score “isn't always the best predictor of customer behavior or business results.”

After the publication of Keiningham’s research, others proposed that the question itself is suspect. And from the beginning, Reichheld admitted that the net promoter score is “simply irrelevant” in certain markets and sectors.

For example, we find that in business-to-business contexts, where products and services are specialized, many customers don't have anyone to mention the product to and so the question doesn't lead to useful answers. Additionally, even for certain consumer-directed businesses, like rental cars, the question doesn't apply because unless something dreadful happened, only the odd bird mentions their rental car experience to a friend or colleague.

Mostly, we find the question overused. Therefore, as with an offhanded “how are you doing today?” it is unlikely to get anything other than a superficial answer.

Digging deeper into why the question may not work, all researchers who have done fieldwork know that there is a wide divide between what we say about ourselves and how we actually act. For example, while people may say they regularly watch PBS, when the ethnographer visits their home, they might very well find American Idol on the TV. The bottom line is this: when someone says they would be extremely likely to recommend your company or product, that’s nice, but it doesn’t mean they will refer you, nor does it even mean they were so impressed that they will return to buy from you again. And while we find tremendous value in Reichheld’s work, we encourage companies to think through how the question will work for their specific company before investing in the generic net promoter question.

Thinking about using net promoter? Get a free Net Promoter Consultation.