Reichheld (2003) introduced the net promoter score with a specific empirical claim attached, namely that the single recommendation question predicted growth better than the satisfaction measures it was meant to replace. The claim invited testing, and the tests were not kind to it. Keiningham et al. (2007a) examined it longitudinally in the industries Reichheld had cited as exemplars and could not reproduce any clear superiority of the recommendation measure over established loyalty measures. Morgan and Rego (2006) found that average satisfaction carried the greatest predictive value for business performance while recommendation-based metrics showed little or none, and van Doorn et al. (2013) reported that the competing metrics performed equally well on current performance and equally poorly on future sales growth. Keiningham et al. (2007b) added that multivariate models using several metrics together outperformed the best single-metric models by 20 to 25 percent in adjusted R squared in banking and retail.
The intention ceiling
A more basic limit sits underneath the score. Sheeran (2002), reviewing the intention-behavior literature, reported an average correlation of .53, which accounts for roughly 28% of the variance in behavior, and a metric built entirely from a stated intention inherits that ceiling before any of its own problems are considered. Schmitt et al. (2012) tested the link directly and found that intention to recommend did not enhance actual customer lifetime value or actual loyalty. Much of this literature, however, compares one self-report with another or with firm-level financial results, and comparatively little of it reaches recorded customer behavior at the level of the individual destination or supplier.
The lift gate evidence
Matzler et al. (2026) closed part of that gap. The study combined biannual customer satisfaction survey data from 2011/12 through 2017/18, more than 120,000 responses in total, with electronic lift gate records from 38 Alpine ski resorts across seven seasons, so that the score could be set against visits that actually occurred. The net promoter score accounted for about 4% of the variance in visits, a real but thin signal, and plain customer satisfaction outperformed it on the same data. The finding of greater consequence was an interaction with experience, reported by the authors in these terms: “if the sample exceeds destination experience levels of 2.3, the net promoter score does not significantly predict destination skier visits anymore” (Matzler et al., 2026). The interaction was reported at a weaker significance level (p < .10) and warrants caution in proportion to that.
A reading through credence qualities
One interpretation of the experience effect draws on Darby and Karni (1973), whose treatment of credence qualities describes attributes a buyer cannot verify before the purchase and sometimes cannot verify after it. A customer without experience of a supplier has little to judge by except reputation, and a recommendation question is well suited to capturing reputation. A customer with experience judges from episodes, and the episodes, being specific and remembered, appear to displace the reputational judgment the score was built to measure. The transfer of this reading from leisure travel to industrial buying remains an inference, and it should be held as one, although the mechanism it describes has nothing in particular to do with skiing.
In finality
Industrial distribution sits almost entirely above such a threshold. Plants with long purchasing histories, distributors that have carried a line across several product generations, and repair shops that reorder on a steady cycle are experienced customers by any definition. If the Matzler et al. (2026) result generalizes, the whole customer base of a typical industrial brand sits above the level at which, in the one study that tested the score against recorded behavior, the score stopped predicting anything. A proportionate response retains the survey and narrows what is asked of it. Following Keiningham et al. (2007b), several metrics read together, with satisfaction, pricing, support, and ease of doing business reported separately, can be expected to outperform any single figure, and reading those figures beside order behavior over time moves channel measurement toward the kind of evidence the lift gate records provided.
The full argument is in the guide on what happens to the score on your best customers. It informs the design of the Channel Health Index.
References
Darby, M. R., & Karni, E. (1973). Free competition and the optimal amount of fraud. The Journal of Law and Economics, 16(1), 67-88. https://doi.org/10.1086/466756
Keiningham, T. L., Cooil, B., Andreassen, T. W., & Aksoy, L. (2007a). A longitudinal examination of net promoter and firm revenue growth. Journal of Marketing, 71(3), 39-51. https://doi.org/10.1509/jmkg.71.3.039
Keiningham, T. L., Cooil, B., Aksoy, L., Andreassen, T. W., & Weiner, J. (2007b). The value of different customer satisfaction and loyalty metrics in predicting customer retention, recommendation, and share-of-wallet. Managing Service Quality, 17(4), 361-384. https://doi.org/10.1108/09604520710760526
Matzler, K., Strobl, A., Teichmann, K., & Aigner, G. (2026). Testing the predictive validity of the net promoter score in ski resorts: A longitudinal analysis. Review of Managerial Science, 20(9), 3265-3288. https://doi.org/10.1007/s11846-026-00986-2
Morgan, N. A., & Rego, L. L. (2006). The value of different customer satisfaction and loyalty metrics in predicting business performance. Marketing Science, 25(5), 426-439. https://doi.org/10.1287/mksc.1050.0180
Reichheld, F. F. (2003). The one number you need to grow. Harvard Business Review, 81(12), 46-54.
Schmitt, P., Meyer, S., & Skiera, B. (2012). An analysis of the link between customers’ intention to recommend a firm and the lifetime value of its customers. Recherche et Applications en Marketing (English Edition), 27(4), 121-142. https://doi.org/10.1177/205157071202700405
Sheeran, P. (2002). Intention-behavior relations: A conceptual and empirical review. European Review of Social Psychology, 12(1), 1-36. https://doi.org/10.1080/14792772143000003
van Doorn, J., Leeflang, P. S. H., & Tijs, M. (2013). Satisfaction as a predictor of future performance: A replication. International Journal of Research in Marketing, 30(3), 314-318. https://doi.org/10.1016/j.ijresmar.2013.04.002