Disclaimer: This is going to be unpopular with about 60% of the consultants I know. That's fine. I don’t really care about consultants. Let me explain.

NPS was a useful framework when Fred Reichheld introduced it in 2003. The world was different. Customer data was sparse. The internet was not yet the universal feedback channel it is now. The single-question survey was an elegant shortcut for executives who needed a directional read on customer sentiment without an entire research apparatus. For its moment, it was a real contribution.

The problem is that we are not in that moment anymore. And the way most brands use NPS today does three actively destructive things.

First, it gets averaged. A score of 40 made up of half passionate promoters and half quiet passives is treated the same as a score of 40 made up of mild positive sentiment across the board. These are completely different businesses with completely different futures, and the metric flattens them into the same line on a slide. The first business has a movement on its hands. The second has a slow-moving liability. NPS cannot tell you which one you are running.

Second, it optimizes for the wrong end of the distribution. Most NPS programs I have seen end up obsessing over detractors. Customer success teams get assigned to "close the loop" on 6s and below. Executives demand action plans on every negative comment. Entire org charts get bent around minimizing the bottom of the curve. The problem is that detractors are not what drive growth. Promoters do. And the energy spent dragging a 5 up to a 7 is energy not spent figuring out how to create more 10s. I have watched brands spend three years lowering their detractor rate while their promoter rate stayed flat, and then wondered why revenue didn't move.

The only thing that actually drives loyalty is the extreme positive. Not the removal of negatives. Not the conversion of passives. The creation of people who would, without being asked, tell their best friend about you tomorrow. Everything else is noise.

Third, and this is the one that should bother every CMO reading this, NPS measures stated intent in a survey, not actual behavior in the wild. The promoter rating is a number a customer typed into a box after you emailed them. It is not whether they actually came back. It is not whether they actually told a friend. It is not whether they renewed, upgraded, or showed up to your next event. The correlation between stated NPS and actual behavior is weaker than the industry pretends. I have watched brands celebrate a 12-point NPS jump in a quarter while repeat visit rates and membership renewals were quietly going the other direction.

A few years ago I was sitting with the marketing leadership of one of the biggest luxury brands in the world. They had just run a global NPS study and were proud of the numbers. I asked one question. I asked whether the people who rated them a 9 or a 10 were any more likely to actually buy the product again than the people who rated them a 6. There was a long pause. Nobody knew. They had not connected the survey data to any behavioral data. They were managing to a number that was floating, unanchored, telling them a story disconnected from what their customers were actually doing.

Here is what works instead.

Measure behavior, not sentiment. Did the customer come back. Did they refer someone whose visit you can attribute. Did they upgrade their membership. Did they buy at full price the second time. These are the questions that matter, and they are answerable for any brand willing to instrument the relationship properly.

Measure the top of the distribution explicitly. Track promoter behavior separately. What did your top 10 percent of customers do this quarter that the rest didn't. The strategy for growth is sitting in that delta.

Measure at the moment. Survey at the moment of experience, not three weeks later when the customer has gone cold. The best brand teams I know are capturing sentiment in real time, at the tasting room, at the event, at the pop-up, when the emotion is hot and the data is honest.

And stop treating NPS as the headline number. Put it on page three of the deck if you have to put it anywhere. Make the headline a behavioral metric tied to a dollar outcome. Repeat purchase rate. Member retention. Referral-driven revenue. Anything tethered to reality.

NPS is not evil. We actually use it extensively at AnyRoad! It’s actually a very good predictor of all the good engagement data I just ranted about. It is just misunderstood and over-relied upon. In the world that existed when it was invented, an elegant proxy made sense. In a world where every consumer brand has the ability to measure actual behavior at actual moments, continuing to manage only to a survey number is a choice to fly with the instruments turned off.

The brands that are going to win the next decade are the ones building loyalty around the extreme positive and measuring it in behavior, not in self-reports. Everything else is theater.