Re-evaluating the Net Promoter Score

The ultimate question is simply, "How likely are you to recommend company X to friends and colleagues?" Respondents are asked to rate the likelihood of recommending on a scale from 0, meaning extremely unlikely, to 10, meaning very likely. Promoters are those who give a rating of 9 or 10. Promoters are important not only because they are more likely to promote your brand to friends and family, they also tend to spend more than non-promoters (Reichheld, 2003).

According to Reichheld, you need to take it a step further if you want to relate the rating to future corporate growth, by calculating a Net Promoter Score. This is simply the percentage of respondents who give your brand a 9 or 10 rating on a 0 to 10 recommendation scale, minus the percentage who give you a rating of 0 to 6 on the scale (Reichheld, 2003).

Popularity

On the back of Reichheld's finding that the Net Promoter Score correlates with future corporate growth across a number of sectors (Satmetrix, 2004), the measure has gained broad support from executives across a variety of industries. Some, such as GE's CEO Jef Immelt, have taken the unusual step of publicly praising the approach (General Electric, 2005).

Besides the claimed relationship with corporate growth and customer behaviour, an additional attraction is simplicity, something very appealing to time pressured executives.

You don't need a degree in statistics to figure out that there is something intuitively wrong with the claim that whether a customer says he will recommend your company or not is the best predictor of whether he will stay with your company.

Differences in the tendency to recommend don't automatically mean there will be differences in loyalty behaviour - whether retention or increased spend. Results from an analysis I conducted of a number of different markets, ranging from short term insurance to cars, made this clear. Clearly the Net Promoter Score does not relate to the percentage of customers switching away from each institution.

A survey of 8000 customers in the banking, retail and ISP industries, conducted by Keiningham et al. (2007), lends support to this.

Doesn't identify those who are truly loyal

Reichheld (2003) states that some clients buy out of habit, which means that past retention (ie did they stay or did they go) is not a good indicator of true loyalty. He suggests that the Net Promoter approach is able to identify true loyalty. This suggests that Net Promoters should be more resistant to competing offers than those who just buy out of 'habit'.

However, using cellphone data collected by Ask Afrika, I found hidden differences amongst consumers who were identified as 'Promoters'. Far from being uniformly loyal, some were far more prone to being tempted by competitor offers than others.

One of the foundational claims which undergirds the value of the Net Promoter Score is the claim that the Net Promoter Score relates to corporate growth. However, in a study of 80 companies over a seven year period, the researchers Morgan & Rego (2006) found that the Net Promoter Score was not predictive of company growth rates.

In summary

Reichheld has brought up some valid points about loyalty measurement. Simplicity, brevity and standardisation are worthwhile objectives.

That is one side of the story. The other is that the 'Ultimate Question' is far from being the most predictive measure of loyalty, and so will more often mislead than help.

In addition, calls for the elimination of service satisfaction measurement make little sense. Loyalty must come from somewhere, and in order to understand its sources it is necessary to measure potential drivers such as service satisfaction.