NPS…and Beyond

NPS certainly has its appeal, especially to CEOs and senior managers who want their information in direct and simple form. NPS is simple to calculate and therefore explain to others, has face validity and intuitive appeal, and can be the one number management seeks for its balanced scorecard and key measures.

But…is NPS really the best measure for all companies across all industries and sectors? Even its strongest proponents concede that Willingness to Recommend is ‘simply irrelevant’ in certain markets and sectors.

  • Cincinnati’s Burke, Inc. found that of 15 measures of loyalty and value, willingness to recommend was found to be among or one of the best predictors in some sectors and markets, but one of the worst in others.

Calculation of NPS has serious limitations which produce conflicting interpretations:

Respondents are asked to evaluate their likelihood to recommend the company/product on a 0 to 10 scale, with 10 being the highest. The NPS is calculated by subtracting the percentage of respondents giving a response from 0 through 6 (Net Detractors) from those giving a 9 or 10 response (Net promoters.)

Consider the following two scenarios where NPS equals 40%:

% Promoters % Detractors NPS
Company ‘A’ 40% 0% 40%
Company ‘B’ 70% 30% 40%

Do these two companies really have similar challenges?

There are better predictors of actual behavior. Other measures correlate more strongly with customer retention and growth.

  • Research conducted by Maritz found that NPS was explained by drivers (i.e., customer experiences) different from those that explain actual behavior. Other measures provided a better match. What this means is that using NPS to determine where a company should focus its improvement efforts to drive ultimate business success, could very well point a company in the wrong direction.
  • An article in the highly respected journal, Marketing Science, tested several satisfaction and loyalty metrics and found that NPS had no predictive value in determining future business performance.
  • A second article in the Journal of Marketing found a weak relationship between NPS and firm growth. Other metrics were found to be far stronger predictors.

Willingness to Recommend is not a universally acceptable or preferred measure. Not everyone or every case can be judged by recommendation.

  • ‘Hard earned’ knowledge, especially in a business context, may cause a person not to want to give away what they know.
  • If business colleagues follow your positive recommendation, the work load of the company recommended could increase such that support of your own account may suffer.
  • Some do not believe in recommending products to others especially in certain product categories, such as medical or pharmaceutical. It may not be in their makeup to make recommendations, but when they do (which is seldom) they take it very seriously and with great caution.
  • When there are several close competitors, NPS can overstate growth potential. A respondent may be highly likely to recommend several companies.

Try putting results from just one number into action within your company.

In order to improve, a company must know the specific actions it has to take. But, without additional information to guide, focus and prioritize, a company is left simply with one number and a great deal of frustration!

The Loyalty Research Center typically includes Likelihood to Recommend as part of its assessment of customer loyalty. We often compare measures and identify those that are most effective and predictive of loyalty.

A recent program compares the ability of NPS to LRC’s own loyalty measure to discriminate customer behavior – in this case share of business:

LRC Measure Share of Business NPS Measure Share of Business
Loyal 66% Promoter 49%
Neutral 45% Passive 47%
Vulnerable 28% Detractor 19%

Note that the LRC loyalty metric is far better able to discriminate the amount of business purchased from this company. The NPS has difficulty discriminating the buy behavior of the Promoter and Passive customers.

The greatest appeal of Net Promoter is as well its greatest downfall. Reichheld would have you believe that one number is all a company needs. If all the publicity surrounding the Net Promoter Score gets your management’s attention, know that it can be a great place to start. However, don’t be surprised if you learn there are better measures out there!

Posted in Financial Services, Information Technology, Insights, Professional Services, Retail, White Papers.