A Customer Survey Can Change the Sale It Claims to Measure.

In a 1993 Journal of Consumer Research study, Vicki Morwitz, Eric Johnson and David Schmittlein found that people asked once whether they intended to buy a car or personal computer were later more likely to make the purchase.

That means a standard research question can act like a marketing touchpoint, altering the behavior a company thinks it is merely observing. The decision is no longer only which survey to send, but whether the business is measuring demand, nudging it or accidentally mixing both.

The Question Enters the Experiment

Marketing teams often treat surveys as windows: look through one and the customer’s existing opinion becomes visible. An intention question is different. To answer “How likely are you to buy?” a person has to retrieve feelings about the product, picture the action and translate uncertainty into a stated position. The measurement creates a brief rehearsal of the decision.

Researchers call this the mere-measurement or question–behavior effect. The question can make an attitude, intention or action easier to retrieve later. It can also create pressure for consistency between what someone said and what that person subsequently does. The survey is therefore part of the customer journey, even when the company labels it research.

The original study also blocks the most tempting misuse. Repeated questions did not simply produce more buying. Among people with low initial intent, repeated measurement reduced purchase propensity, and prior product experience weakened the effects. Asking can clarify resistance as easily as it strengthens interest. This is not a magic conversion phrase; it is evidence that questions can sharpen whatever is already psychologically available.

A Survey Can Move Behavior Beyond the Survey

The finding did not remain inside a laboratory. In a financial-services field experiment published in the Journal of Consumer Research in 2002, Utpal Dholakia and Vicki Morwitz reported that measuring customer satisfaction changed one-time purchases, defection, aggregate product use and profitability. In that setting, some effects increased for months and persisted a year later.

Another Journal of Consumer Research study, by Pierre Chandon, Vicki Morwitz and Werner Reinartz, compared online-grocery customers whose repurchase intentions were measured with similar customers who were not surveyed. Measurement increased repeat-purchase incidence and shortened the wait to the first repeat purchase. Those direct effects decayed rapidly after three months, yet earlier buying accelerated later purchases enough to leave a longer profitability effect.

These studies do not mean every satisfaction form creates loyalty. They show why a surveyed customer and an untouched customer are not automatically interchangeable observations. The invitation, the wording and the act of answering can become exposures of their own.

The Mechanism Is Useful—and Dangerous

A 2016 Journal of Consumer Psychology meta-analysis synthesized 104 question–behavior studies across 51 published and unpublished papers. It found substantial variation in questions, behaviors and contexts, with evidence supporting several possible mechanisms: attitudes become more accessible, people seek consistency, the action becomes easier to imagine and motivation changes.

That heterogeneity matters. A company should not assume a survey will create a positive effect, or that the same question will behave identically across products, customers and moments. A delighted subscriber may use the prompt to rehearse renewal. A frustrated subscriber may use it to organize reasons for leaving. A prospective buyer may interpret a purchase-intention question as commercial attention rather than neutral research.

The measurement risk is managerial as much as statistical. If a brand surveys only high-value customers and then sees stronger retention, it may credit its service when the research contact contributed to the outcome. If it compares respondents with nonrespondents, motivation to answer may explain more than the score itself. And if “research” is designed mainly to push a sale, the company has replaced learning with disguised persuasion.

Build a Cleaner Learning Loop

A small business can start by treating each questionnaire as a campaign exposure. If possible, hold back a comparable group and compare completed bookings, repeat visits or purchases—not merely survey scores. A midsize company can record who received which question, when it arrived and whether the person answered, then separate invitation effects from response effects. An enterprise can randomize invitation, wording and timing while monitoring long-term outcomes, consent and customer fatigue.

OrionPilot’s published approach connects business context, strategy, creative production, execution, connected data, analytics and continuous learning. The question–behavior effect adds an important discipline to that chain: preserve not only what the customer answered, but also the fact that asking may have influenced the next action. When feasible, an uncontacted control group keeps the next strategy from treating stimulated behavior as untouched demand.

Every research question therefore has two jobs. It gathers evidence, and it enters the customer’s memory. Good marketing measurement acknowledges both—using questions thoughtfully while protecting the comparison needed to learn what actually changed the sale.