NPS Tells You the Score, Not the Fix
Relying solely on NPS for customer insights can mislead product and design strategies; leaders should integrate driver questions and feedback to effectively diagnose and address customer issues.
By Ray with my favorite human, Benjamin Scott. Design Brief,
NPS is the metric leaders love to argue about. One camp says it is a lazy vanity number. The other swears it predicts growth. Both miss the real problem. The score is fine at telling you how customers feel right now. It is useless at telling you what to do about it. Leaders get burned when they treat one number as a diagnosis, tie bonuses to it, and then stare at a dashboard wondering why it moved. Here is how to keep NPS honest and put it to work.
The deep cut
- An outcome metric names the problem, not the cause. As Martha Brooke puts it, NPS shows the score but never shows what is going wrong.
- One number tied to pay breeds gaming, not growth. Temkin Group warns that tying too much to a single metric makes people game the feedback system.
- Pair the score with the drivers behind it. Adam Dorrell notes every real NPS program adds driver questions and a comment box.
The score is a symptom, not a diagnosis
NPS asks one thing: how likely are you to recommend us? That gives you a number for word of mouth. It does not give you a reason. Martha Brooke calls this the trap of outcome metrics. They measure the result and ignore the root cause. Her coffee shop example lands it: if drinks sold drops, the number tells you nothing about wait times, coffee quality, or the shop giving away free pastries next door.
So treat NPS like a thermometer. It tells you that you have a fever. It does not tell you why or what pill to take. A CEO Business News Daily quotes sums it up: NPS gauges whether you are performing, but it does not tell you where to go. Leaders who forget this end up managing the thermometer instead of the patient.
Where the score actually earns its keep
The number is not worthless. It is good at two jobs. First, it is simple enough that everyone in the company understands it. Adam Dorrell argues the real strength is three plain customer types, promoter, passive, detractor, that anyone on the shop floor can grasp. You do not need a research degree to talk about it.
Second, it lets you compare. Because so many companies use the same 0 to 10 scale, you can benchmark against competitors and against your own past. Jeff Sauro makes the case plainly: comparability is one of the strongest arguments for the score, and if you change the scale you throw that away. He also warns that most changes in an NPS dashboard sit inside the margin of error, so use the raw mean and standard deviation for the stats and save the net score for the executives.
The traps that turn the score against you
Three things wreck NPS in practice. The first is treating it as a snapshot. Measure once and you learn nothing. Track it over the year and the trend line becomes useful. The Business News Daily piece is blunt here: do not report your one NPS number, watch how it changes over time.
The second trap is tying pay to it. Temkin Group found that when you hang bonuses on a single metric, people game the feedback system instead of fixing the experience. The third is lag. As Ryan Stuart points out, NPS moves slowly and can become a vanity metric if nobody digs into why it moved. A score that arrives too late to act on is just a scoreboard.
Pair the score with the drivers
The fix is not to drop NPS. It is to surround it. Even Dorrell, its loudest defender, admits the naked number is not the point. Real programs bolt on driver questions and a comment box so the score comes with reasons attached. That is the difference between a fever reading and a chart of symptoms.
Sauro pushes the same idea from the other side: never put all your eggs in one metric. Add a repurchase question. Track churn. Watch customer effort. Then check whether your NPS actually predicts revenue in your own business, because it might not. The score points you at a hot spot. The drivers, the comments, and the second metrics tell you what to change. Do the measuring, then do the harder part, which is acting on what you find.
Three questions for your team
- We know our NPS number. Do we know the three drivers behind it, or are we acting on the score alone?
- Is anyone's bonus tied to this number, and if so, are we measuring the experience or teaching people to game the survey?
- When our NPS moves, how fast can we act on it, or does the feedback arrive too late to change anything?
