NPS: No Panacea Sadly

Consumer-facing industries are increasingly adopting Net Promoter Score (or NPS) as a Board-level metric to measure the performance of the business and management. This trend is supported by regulatory drive, especially in Financial Services, to improve customer outcomes. While the NPS metric is widely used, it is not necessarily well understood. We believe adopting NPS must be done carefully to avoid value-destroying outcomes—and that in some circumstances there may be better alternative metrics to apply.

Adopting NPS must be done carefully to avoid value-destroying outcomes

NPS measures customers’ response to the question: “How likely are you to recommend [Brand] to a friend or colleague?” While many studies have indicated that NPS is a good predictor of loyalty, there are a range of issues with the metric. A few of the most common:

  1. Measurements are strongly influenced by the overall brand strength. Weaker brands can lower NPS significantly for an identical service. In multi-brand companies, for example, two identical offerings can have widely divergent NPS ratings because of relative brand strength.
  2. NPS is also strongly influenced by how, where, and when the survey is done. For example, at point of sale, NPS reflects recent transactions; if the survey is done at a later date, the consumer may be too removed from the transaction for it to be memorable, and the rating is thus more akin to a general brand perception than a predictor of loyalty. We have also found that phone answers tend to be more positive than answers delivered online.
  3. Infrequent experiences—such as efficiently handling a customer’s written correspondence—can be very important to customers, but may be missed in some surveys.
  4. In some cases, advertising that customers dislike can increase the number of a company’s detractors—when in practice nothing has changed with the quality of the customer experience.
  5. In some industries, we have found a difference between what people will recommend, and what they will choose for themselves. Recommendation introduces a social filter, about whether I want to be seen to recommend something, which may bias it up or down depending on the brand association.

In the end, if your business is focusing specifically on NPS as a metric, then you need to understand in detail what drives it—both at a high level (such as call-center performance vs. billing) and at a detailed level (such as how long a call-center wait time is needed to drive down NPS ratings). For example, in our own research for consumer service industries, we have found that the lifetime value of moving a “detractor” to “neutral” is more than 2x the value of moving a “passive” to a “promoter.”

NPS is not the straight-forward panacea metric that many companies think it is

More broadly, we are cautious about over-reliance on NPS as the one brand and customer performance metric, as you risk bypassing your business and brand strategy and driving your people to follow a lowest-common-denominator approach that kills differentiation. This is because the organization will end up taking the direct route to an NPS score, and neglect investing in the image attributes your brand strategy wants you to stand for. Some industries, like telecom, are suffering from putting their effort solely into reclaiming detractors, rather than building powerful new advocates. Only to find they don’t have the customer relationships to compete with firms that have strong loyalties, like Google and Amazon.

To summarize, NPS is not the straight-forward panacea metric that many companies think it is. In some cases, simpler metrics may be a better reflection of value and loyalty. In other cases, a more sophisticated understanding of NPS mechanics is needed to drive long-term value. Even then, if you take the direct route to an NPS score, and ignore investment in the image attributes your brand strategy requires, you will fail to differentiate yourself, and you will see short-lived value creation.

The author would like to thank his fellow Partner, Simon Glynn, for the suggestions around this RevenueRx article.