Updated: September, 2026
Published: February, 2022
The Say-Do Gap: Why Consumers Don’t Do What They Say — and How to Trust Your Survey Data Anyway
Consumers rarely do exactly what they tell you they will. It is one of the oldest and most stubborn problems in consumer research — and, as behavioural and AI-generated data compete for attention, it is increasingly used to argue that surveys can no longer be trusted at all. But that overstates the case. Here is why the say-do gap is unavoidable, why it does not render your survey data worthless, and how the right research design lets you trust what your consumers tell you.
Why Bother Asking Consumers at All?
Consumers are unreliable narrators of their own behaviour. They report intentions they never act on, offer motives they are not consciously aware of, and express preferences that dissolve the moment a real decision is in front of them. This is not dishonesty; it is simply human.
It also raises a fair question. If people cannot be relied upon to do what they say, why ask them at all? Why is quantitative market research still a multi-billion-pound industry, and why not move straight to behavioural data?
The answer is that no single source holds the whole truth about consumers. There is no methodology that will reveal, once and for all, exactly what drives a purchase, and no shortcut that removes the need for judgement. That is the one certainty worth trusting, whichever way you set out to understand trends, opinions, behaviours or sentiment.
What genuinely advances your brand positioning, communication and product decisions is understanding consumers on two levels at once: the conscious and the unconscious. You need their considered reflections, and you need to account for the say-do gap in your survey-based market research.
What Is the Say-Do Gap?
The say-do gap is the difference between what consumers say they will do and what they actually do. A shopper insists that price is not their priority, then reaches for the cheapest option on the shelf. A respondent commits to switching to a more sustainable brand, then never changes their habits. The cause is rarely deception: we are all shaped by influences we do not consciously register, and we routinely act against our own stated intentions.
The important consequence is that every form of research carries some bias. Surveys capture stated attitudes but not always subsequent action. Behavioural data captures action but not the reasoning behind it. Because no method sees everything, the say-do gap cannot be removed. It can, however, be understood and managed, and that distinction is where good research earns its value.
The Real Problem Is Reality
There is, undeniably, a difference between what a person reports in a survey and how they behave afterwards. But that difference invalidates neither the survey nor the observation. More often, it is in identifying where the gap sits that the clearest opportunities for sharper marketing appear.
Consider the shopper who assures you that price is not their main concern, then reconsiders at the shelf and selects a competitor’s product, having found, despite your confidence in your organic credentials, that the alternative simply tastes better. That does not make your organic positioning irrelevant. It may, however, signal a genuine need to improve taste, pricing or packaging.
Or consider a sponsored digital advertisement that outperforms a carefully planned brand campaign because a competitor bid higher, wrote a sharper message, or produced stronger creative, and captured the purchase in the process.
This is reality, and it does not diminish the value of consumer insight. Improving your chances still depends on understanding consumers before the point of decision. The aim is to prepare for that moment, not to be caught out by it.
This is not merely a thought experiment; it surfaces in real research too. According to Opeepl’s Youth Pulse study (Wave 3), only around 17% of young consumers say they actively seek out and prioritise sustainable products when shopping. Nearly half — 48% — say they consider sustainability but that other factors such as price, brand preference or convenience usually take priority, and roughly a quarter say sustainability matters to them “in theory” yet admit they do not always act on it in their purchasing decisions.
The attitude is real; the behaviour it predicts is not. And note what this data is: not a record of what these consumers did at the checkout, but what they told us about themselves. Even in their own words, the gap is visible — which is exactly why well-designed survey research is one of the most honest ways to bring it into view.
Why Quantitative Data Is the Answer to the Say-Do Gap
Understanding the real path to purchase means accounting for more than consumers’ values and attitudes. It means recognising the influence of the unconscious — the way context, status and social convention can override what people believe they will do.
A perfect picture of this is unattainable, regardless of budget. Fortunately, perfection is not the requirement. The objective is to generate enough insight to understand the context in which consumers will evaluate your brand, and what is most likely to resonate with them.
This is where quantitative research proves its worth. At scale, robust numbers outweigh anecdote. Insight drawn from a small number of consumers may be interesting, but it carries a substantial risk of reflecting noise rather than reality.
Scale alone, however, does not guarantee quality. Three factors determine how effectively you handle the say-do gap in survey-based research. Address them well, and you produce quantitative data with the least possible bias — the foundation for insight you can confidently act on.
How to Close (or at Least Narrow) the Say-Do Gap
The say-do gap is addressed during the research-design phase — one more reason design deserves careful attention, and why investing in research-design capability is never wasted.
1. Ensure your questionnaire does not introduce bias. Questionnaire bias is more common than most teams expect, and among the leading reasons results fail to reflect reality. If you are uncertain, a capable research partner should advise on your questionnaire and, ideally, flag any questions likely to distort your data before fieldwork begins. When they do, revise the questions and answer options accordingly.
2. Include the right dimensions. Omitting an important dimension can conceal significant patterns — for example, that a meaningful share of your audience buys on promotion or personality rather than loyalty. Approach your research more exploratively than first instinct suggests, and consider deliberately investigating why consumers might choose not to buy from you.
3. Research the right people. When defining screening criteria, consider whether both your existing and your potential audiences have something valuable to contribute. Screening is too often insufficiently exploratory to reveal potential you have not yet identified.
Not sure whether your own questionnaire is quietly introducing bias?
A good research partner should catch that before fieldwork begins.
Why the Say-Do Gap Matters More than Ever
The say-do gap is not new, but the pressure surrounding it is. As behavioural, first-party and AI-generated data become easier to obtain, it is tempting to conclude that asking consumers anything is outdated. But that conclusion goes too far. Behavioural signals reveal what happened; they seldom explain why — and it is the “why” that allows you to influence the outcome next time. The brands that succeed are not those that abandon surveys, but those that design them well enough to trust and combine them with other evidence, rather than trading one blind spot for another.
Work with the Say-Do Gap, Not against It
The say-do gap is yours to manage, and manage it you must, if consumer insight is to become the competitive advantage your organisation is investing in. Consumers will always deviate to some degree from what they say. That does not render market research obsolete; the alternative is to make decisions with no visibility at all.
The priority, then, is sound research design and a clear-eyed acceptance that every data-collection method carries uncertainty. The best data comes from the best design; no amount of analysis afterwards can rescue insight from a flawed one.
In short: understand where uncertainty lies along your consumers’ path to purchase, and account for it in both your research design and your analysis. Do so, and the say-do gap ceases to be a reason to doubt your data, and becomes a reason to design it more carefully.
Frequently Asked Questions
What Does the Say-Do Gap Mean?
The say-do gap is the difference between what consumers say they will do and what they actually do. It reflects the ordinary human tendency to be influenced by factors we do not consciously notice, and to act differently from what we intended.
Can the Say-Do Gap Be Eliminated?
No. Because every research method carries some bias, the gap cannot be removed entirely. It can, however, be understood and reduced through careful research design — chiefly by avoiding questionnaire bias, including the right dimensions, and screening the right people.
Does the Say-Do Gap Mean Survey Data Is Unreliable?
Not on its own. Surveys measure stated attitudes, which remain valuable for understanding the reasoning behind behaviour. The gap becomes a problem only when research is poorly designed; well-designed surveys account for it and stay trustworthy.
How Do You Close the Say-Do Gap In a Survey?
You address it at the design stage: write questions that do not lead respondents, include the dimensions that actually drive choice (including reasons not to buy), and screen both existing and potential audiences so you capture the full picture.
Get Survey Data You Can Actually Trust
The say-do gap is manageable, but only with the right research design. Talk to Opeepl about designing a study, or reviewing your questionnaire for bias, so your next round of consumer insight holds up.