Repeated Customer Ratings Can Change the Experience They Measure. Marketing Should Check the Cadence.
Repeated ratings do not only record how an experience is going. Across sixteen studies, they changed how much people enjoyed it — a bounded finding with a practical question for every customer feedback program.
What happened
On 30 September 2026 the Journal of Consumer Research published online an accepted manuscript by Stephanie Hong (University of Chicago Booth) and Ed O'Brien (Ohio State University Fisher): Rating an Experience Makes People Notice How It Is Changing: Consequences for Hedonic Adaptation and Consumer Enjoyment.
Across sixteen studies with 16,311 participants, people who were asked to rate an experience repeatedly were compared with people asked less often, or not at all. Repeated ratings drew attention to how the experience was changing. When it was getting worse, the decline became more noticeable and boredom and disengagement came sooner. When it was getting better, the improvement became more noticeable and was enjoyed more.
The authors open with marketing itself: repeated measurement is "common in both marketing research and practice", from experiments where people rate several similar products to service providers who "regularly check in to see how customers are doing", on the assumption that asking more simply reveals more.
What the evidence actually says
Observed. Sixteen studies, N = 16,311, with hedonic stimuli — games, foods, songs, videos and artwork — and several methods, including mediation and moderation tests.
Reported. The direction of the change decides the effect: repeated ratings amplify a decline and amplify an improvement. The authors discuss how repeated measurement may distort understandings of hedonic decline and consumer preferences. They also discuss how the effect might be used to make enjoyment last longer; we do not recommend that without field tests.
Context. That measuring can change behaviour is not new in consumer research: the "mere-measurement" effect has been studied since at least the 1990s (Morwitz, Johnson and Schmittlein, Journal of Consumer Research, 1993). What this paper adds is a specific mechanism: repeated ratings direct attention to change, and the direction of the change decides whether that helps or hurts.
Unknown / limitation. We read the publisher's abstract and metadata; the full text was not accessible to us. The abstract does not give effect sizes. The stimuli are hedonic experiences, not NPS, CSAT or post-purchase programs, so nothing here shows the same effect in a customer feedback program. The accepted manuscript is not the final version of record.
Why this matters for marketing
Our earlier Signal on measurement was about teams switching rulers between a decision and its review. This one is harder to see: the ruler can stay exactly the same, and the act of reading it can still move the reading.
Many teams now ask for feedback at several points in a journey. If the effect transfers — which this paper does not test — the same experience could be reported differently depending on how often, and when, customers are asked.
Our inference, not the paper's finding: take a service improvement that arrives gradually, piloted with a weekly satisfaction pulse and then rolled out with a quarterly survey. The pilot's frequent ratings may have made the improvement more noticeable, and more enjoyed, than it will be at scale. The rollout then looks like a fade, when part of the gap is how often customers were asked. The reverse holds for a slowly declining experience: a team that pulses often may see the decline sooner and sharper than one that asks rarely.
Most survey platforms already log the prompt, channel and trigger. What usually gets lost is that context in the deck where the comparison is made and the decision is taken.
The operating implication
Treat asking cadence as part of the treatment. When you compare cohorts, a pilot with a rollout, or before with after, check that customers were asked as often and at the same moments. If not, say so next to the number: the two groups were not asked the same way.
Keep asked and observed evidence apart. A rating, a repeat purchase and a support contact are different kinds of evidence. When ratings and behaviour diverge, check how often customers were asked before trusting either.
The non-move: do not cut feedback programs on the strength of one paper. The question this research raises is when and how often you ask, not whether.
The Olymuse lens
This sits under Learning. In Measurement Is Changing Mid-Flight we argued that a decision should keep a record of the measures it relied on. This paper points to a harder case: even when the measure stays the same, how often and when customers are asked can be part of what moved it.
Our point of view: when a team learns from customer feedback, the way it asked belongs to the evidence, not to the settings of a survey tool. A number without its asking conditions is half a fact.
This is a practice a team can start with the survey settings it already has.
What we would watch next
First, field evidence from real customer feedback programs, where ratings are triggered repeatedly across a lifecycle.
Second, effect sizes for familiar marketing measures — CSAT, NPS, review prompts, post-purchase satisfaction — beyond the hedonic stimuli tested.
Third, whether timing matters more than frequency: after purchase, after a service recovery, during onboarding.
Fourth, the final version of record, and whether any method or boundary condition changes.
Read next: Your Marketing Team Should Not Have to Start From Zero Every Time · Olymuse is in founder-assisted private alpha. Apply for the Private Alpha →
Applications are open for the founder-assisted private alpha.
A marketing operating layer above the stack you already run. You stay the decision-maker.
Apply for private alpha