A score tells you the weather, not the cause

You cannot fix a number. You can only fix the thing the number is reacting to.

Satisfaction scores are popular because they compress everything into one movable figure, which is exactly what makes them useless for action. A score can drift for reasons that have nothing to do with your service: a price change, a news cycle, a seasonal shift in who is buying. It moves late, long after the causes have set, and it cannot tell you which of the fifteen things you changed this quarter was responsible. Managers then argue about the number rather than about the experience, which is a conversation nobody can win.

Worse, a score invites management of the score itself. Once a figure is attached to a bonus, energy flows towards when and how the question is asked, which customers are invited to answer, and how staff frame the request. None of that improves anything. It simply degrades the measure until it stops carrying information, at which point everyone still trusts it because it has been on the report for years.

The free text next to the score is worth more than the score. So is the count of people who contacted you twice about the same thing, the number of transfers per case, the age of unresolved matters, and the proportion of contacts that exist only because someone was left uninformed. These are operational facts rather than opinions, they point at a specific decision, and nobody can game them without actually improving something.