Why Surveys Lie · Orbital Brand Science

People do not know why they buy. They invent reasons after the fact, and the reasons sound great. That is the problem.

If you have ever sat behind the one-way glass at a focus group, you have probably watched the same thing happen. A facilitator asks why someone chose a brand, and the respondent pauses, looks at the ceiling, and produces a sentence that sounds like it came out of a marketing textbook. "I prefer the quality." "I trust them." "It feels premium." Heads nod around the table. Someone writes it down. And the brand walks away convinced it knows what its customers value.

The respondent is not lying. That is the strange part. They believe what they just said. The trouble is that the answer was not a memory of a real decision. It was a story their brain assembled in the half second between hearing the question and opening their mouth, and like all stories, it had to make sense. People are exquisite makers of meaning. Ask them why they did something and they will tell you a tidy reason, because the alternative, admitting they cannot fully say, is uncomfortable for everyone in the room.

This is the founding problem of market research, and it is older than market research itself. Psychologists have known about it for at least fifty years. Richard Nisbett and Timothy Wilson published a paper in 1977 called "Telling more than we can know," which has the kind of title that should give every brand manager pause. They ran a now-famous experiment in a department store. Four pairs of stockings were laid out on a table. Shoppers were asked to evaluate them and choose a favourite. The pairs were identical, but a striking preference emerged for the pair on the right. When asked why, almost nobody said "because it was on the right." They invented reasons about texture, sheen, and elasticity, and they defended those reasons under questioning. The position effect was real. Their explanations were not.

The brain explains itself badly

The human brain has two broad modes for handling information, and neuromarketing borrows the shorthand from Daniel Kahneman to call them System 1 and System 2. System 1 is fast, intuitive, emotional, and almost always running. It handles roughly ninety-five percent of what we do, including most purchase decisions. System 2 is slow, deliberate, and effortful. It is what you use to fill in a tax return or compare mortgage rates. Surveys talk to System 2. Markets are decided by System 1.

When a respondent answers a question, System 2 takes the floor. It produces fluent, articulate, defensible reasons. The reasons feel like the truth because they are spoken in the same voice as the truth. But the actual decision, the one made in the supermarket aisle in 2.5 seconds, was made by System 1 long before language got involved. The two systems do not even share the same vocabulary. One uses concepts. The other uses feelings, associations, and learned shortcuts.

This is not a moral failing of the consumer. It is a feature of how cognition is built. The brain is, among other things, an interpretation machine. It generates explanations for behaviour the way the liver produces bile, automatically and without consulting you. The neuroscientist Michael Gazzaniga has studied split-brain patients for decades and has shown that when the left hemisphere is shown a behaviour it did not initiate, it cheerfully makes up a reason and reports that reason with total confidence. We are doing this all day, every day, about everything we do.

The respondent is not lying. That is the strange part. They believe what they just said.

What surveys can and cannot do

Surveys remain useful. We use them. They are good for measuring stated awareness, declared usage, demographics, and shifts in claimed attitude over time. If you want to know whether more Jamaicans recognise your logo this quarter than last quarter, a survey will tell you. If you want to know whether the population believes obesity is rising, a survey will tell you.

What surveys cannot do is tell you why people choose one product over another. They cannot tell you which version of a thirty-second ad will drive incremental sales. They cannot tell you whether a redesigned pack will be more emotionally engaging on shelf. They cannot tell you what people will pay for. For those questions, the gap between what people say and what people do is the entire game, and that gap is where most marketing budgets quietly die.

There is a well-rehearsed example from the food industry. Decades ago, focus groups for ready meals consistently told manufacturers that they wanted healthy options, smaller portions, more vegetables. Companies launched products designed to those specifications. Sales were terrible. The same shoppers, asked in a different way, kept buying lasagne. The disconnect was not dishonesty. It was the difference between the person they imagined they were when answering a question and the person they actually were at six in the evening with two children to feed.

What the brain reveals that the mouth cannot

The interesting thing about neuromarketing is not that it bypasses the conscious mind. It is that it reads the parts of the response that the conscious mind never had access to in the first place. When you measure brain activity with EEG, you can see attention shifts and emotional responses in fractions of a second, long before a person could possibly articulate them. When you track eyes, you see exactly where the gaze landed, how long it lingered, and what was missed entirely. When you analyse micro-expressions on the face, you catch the involuntary half-smile or flicker of disgust that the social brain immediately corrects away.

None of these signals are conscious. None of them can be reported in a survey. All of them are predictive of actual behaviour in ways that stated preference is not. A 2015 meta-analysis from the Advertising Research Foundation found that biometric measures of advertising effectiveness predicted in-market sales lift roughly twice as well as the best survey methods. Other studies have put the multiplier higher. The signal is there. It is in the body.

This is not magic. The brain is doing what brains do, processing thousands of inputs per second and generating responses far faster than awareness can keep up with. Modern measurement just listens to the part of the conversation that has always been happening below the waterline.

Why this matters in the Caribbean

Caribbean consumers face an additional layer that makes survey data even less reliable. Politeness norms run deep across the region. Telling a researcher you dislike a product, especially a local product, feels rude. Saying you find an ad confusing implies the brand is at fault, which is uncomfortable. Acquiescence bias, the tendency to agree with the interviewer, is well documented in cross-cultural research, and it tends to be higher in collectivist cultures than in highly individualist ones. Caribbean cultures are not monolithic, but they tend toward the warmer end of that spectrum.

The result is that survey scores in the region often skew positive on dimensions that, when measured behaviourally, look much more mixed. Brands believe they are loved when they are merely tolerated. Ads test as engaging when they are scrolled past in two seconds. The brand health tracker keeps ticking up, and the share keeps drifting down, and nobody can quite explain why.

We have measured this gap directly. In one recent piece of work for a beverage client, stated brand preference suggested a comfortable lead over the nearest competitor. Implicit reaction times told a different story. When respondents were asked to associate the brand with positive attributes under time pressure, the lead vanished. The competitor was actually faster to "good" by a meaningful margin, which is the kind of signal that shows up in market share six to twelve months later. The survey was telling the client they were fine. The brain was telling them to move.

The honest version of the question

Some of the best neuromarketing studies start with the same question that fills focus groups. "What do you think of this ad?" The difference is that the answer is not the answer. It is the warm-up. While the respondent talks, we are measuring what their attention did during the ad, where their pupils dilated, where their face moved, which moments lit up the engagement signal and which moments killed it. The verbal answer becomes a useful cross-check, not the truth itself.

Sometimes the two line up. Sometimes they do not. The cases where they do not are usually the most valuable, because they reveal a real opportunity to fix something that the brand did not know was broken.

One frequent pattern: people remember the wrong moment of an ad. The marketer is sure the punchline lands at the end. The viewer remembers a visual from the middle. The brand asset that the agency spent two weeks designing into the closing frame is invisible because attention had already collapsed. Nobody in the focus group would have said this. The eye tracker says it within fifteen seconds of playback.

Stop asking. Start measuring.

None of this is an argument for abolishing the survey. Surveys are cheap, scalable, and good at the questions they are good at. The argument is for matching the method to the question. If you want to know whether someone has heard of your brand, ask them. If you want to know whether your brand makes them feel anything, measure them.

The brands that pull ahead over the next decade will be the ones that treat consumer truth as a multi-method problem. They will keep asking, because asking is sometimes the right move. They will also measure, because measuring catches what asking misses. And when the two disagree, they will trust the body over the words, because the body has no incentive to lie.

The shopper in front of your shelf is not the same person who answered your survey three weeks ago. The shopper is faster, less articulate, and more honest. The shopper does not know why they reached for the blue pack. They will not be able to tell you afterwards. But their brain knew. And the brain can be measured.