Frederick F. Reichheld claims that you only have to ask customers one question to grow your business: “On a scale of 1 to 10, how likely is it that you would recommend our brand to a friend or colleague?” He argues that willingness to promote is a good indicator of an individual’s loyalty because in doing so they are aligning their own reputation with the brand.
Reichheld advocates analysing data collected in this way by constructing the now widespread metric of the Net Promoter Score (Promoters [scores 9 and 10] less Detractors [scores 1 to 6] = NPS). But what makes the recommend question superior to other customer experience metrics such as value for money, quality or satisfaction?
The premise is that as willingness to recommend increases then loyalty increases. However, we might also expect a similar relationship for satisfaction or indeed any other of the commonly-used customer experience metrics.
Whichever the chosen metric, there will inevitably be outliers: cases that buck the trend, such as individuals who recommend wholeheartedly but for whatever reason don’t happen to purchase again. For recommend to be the superior metric, there need to be fewer outliers than for, say, satisfaction. It is often not this straightforward – a personal recommendation of a brand is complicated by many other issues. Is the product or service relevant to your friends and family? Is it an appropriate product to recommend? Are you the type of person who is likely to recommend?
Considering these external factors, a graph depicting the recommend ratings against a business metric would be more prone to outliers than other customer experience metrics and not fewer as Reichheld argues. Further, we might expect considerable differences between industry sectors.
Despite the widespread use of the recommend question and NPS, Reichheld has not published his original data in a peer-reviewed journal, nor is there a peer-reviewed rebuttal to the many legitimate concerns that have been raised about the supposed superiority of this metric.
His 2003 Harvard Business Review article discusses it and a white paper is available on Satmetrix’s website. But nowhere are the numbers that matter presented. With so much now riding on NPS, surely it is time to see the original analysis in its entirety?
Analytics director, 2CV