It turns out there’s a reason customer complaints never end.

One of the great benefits of working with long-standing clients is getting to watch the story unfold in longitudinal research. You put a strategy in place, make some changes, and eventually the data tells you whether it worked. Scores improve. Old pain points recede. The wins are real and, for a consultant who recommended the changes in the first place, deeply gratifying.

Then something irritating happens.

A different score slips. A new complaint appears. The thing customers were furious about last year barely registers, but something else has taken its place. You fix that, satisfaction improves, and before long another source of irritation pops up.

At first this can feel like strategic whack-a-mole. Over enough waves of research, though, you start to recognise the pattern. The complaining may not be a sign that the business has failed to improve. It may simply mean that customers have adapted to the improvement.

Psychologists Philip Brickman and Donald Campbell gave us a memorable way of thinking about this in the 1970s: the hedonic treadmill. People adapt remarkably quickly to improvements in their circumstances. What once delighted us becomes normal, the baseline resets, and our attention moves to the next thing that is not quite right.

That translates rather neatly to customer experience.

Fix the delivery problem and customers stop thinking about delivery. Improve the app and the better experience quickly becomes the minimum expected standard. Solve one irritant and satisfaction rises, but the gain is rarely banked forever. Expectations move with you.

This matters because the wrong response is to treat every new complaint as fresh evidence that something is broken. Sometimes it is. Sometimes you are simply watching the treadmill turn.

Over the years, I’ve developed a simple way of separating meaningful customer signals from the inevitable resetting of expectations. I call it The Treadmill Test, and it asks four questions.

First, the Loyalty Filter: does the complaint actually show that the customer cares enough to stay? Some complaints are a form of engagement. Indifference is often considerably more dangerous. A customer who is still telling you exactly what they want fixed may be signalling loyalty rather than abandonment.

Second, the Expectation Filter: have we broken a promise? If the organisation has explicitly or implicitly told customers to expect something and is failing to deliver it, this is not the hedonic treadmill. It is an execution problem. Fix it.

Third, the Impact Filter: what happens if we ignore the complaint? If the likely result is churn, loss of share, declining usage or some other commercially meaningful consequence, pay attention. Volume of complaints is less important than the behaviour sitting behind them.

Finally, the Hedonic Filter: is this simply expectations resetting? If customers have absorbed the last improvement, moved the goalposts and found a new imperfection to focus on, the right answer may be to hold the line rather than launch another improvement programme.

That last one is the hardest, particularly in organisations that pride themselves on being customer-centric. Listening to customers is important. Doing everything customers ask for is not the same thing.

Longitudinal research is valuable precisely because it gives you perspective that a single wave cannot. It lets you distinguish a deteriorating experience from a rising standard, a broken promise from a newly discovered preference, and a genuine retention risk from the background hum of human dissatisfaction.

The goal isn’t to eliminate complaints. That may be impossible. The goal is to know which complaints deserve action, and which are simply evidence that yesterday’s improvement has become today’s expectation.