Article by ICG member Iona Carter, Tracer Insight Consultancy.
The phrase “say/do gap” has become one of the most common criticisms of consumer research. If you’re a consumer or shopper insight user (or provider), you’ll have heard it all before…
- Consumers say one thing and do another.
- Only behaviour reveals the truth.
- Self-reported data can’t be trusted.
But I think the industry has become so focused on proving the say/do gap exists that we’ve stopped asking a more interesting question: Are we focusing on the wrong problem?
Firstly, some so-called say/do gaps are actually research design gaps. Researchers are often asking consumers to perform surprisingly difficult tasks: predict future behaviour or explain past behaviour.
Secondly, what consumers say isn’t necessarily wrong simply because it doesn’t show up in behaviour. In fact, understanding what consumers say may be just as important as understanding what they do – the gap often becomes a goldmine of valuable insight.
Behaviour, Motivation and Barriers
I like to think of the say/do gap using three lenses:
Behaviour What happened?
Motivation What was the consumer trying to achieve?
Barriers What prevented those motivations translating into behaviour?
The mistake is assuming that if motivations don’t show up in behaviour, they must be irrelevant.
So let’s think of an example: consider sustainability.
A consumer tells us sustainability is important when buying laundry detergent. But, behavioural data then shows they bought the cheaper alternative.
The conventional interpretation is often:
“They said sustainability mattered, but their behaviour proves it didn’t.”
I’m not convinced that’s the right conclusion.
Perhaps sustainability genuinely mattered… but it simply lost the battle in the complex trade-offs between multiple goals that we all make when making decisions.
Behaviour tells us which of these goals or motivations won…. it doesn’t tell us which goals or motivations existed.
And, it certainly doesn’t tell us why one goal defeated another… and that is where understanding barriers becomes important.
Whereas the conventional interpretation of the ‘say/do gap here might be:
“They said sustainability mattered, but their behaviour proves it didn’t.”
An experienced researcher might conclude…
“Sustainability is important, but is currently constrained by something else”
Those constraints (or barriers) might be price, trust, availability, convenience or something else entirely.
Understanding those barriers may be more commercially valuable than simply observing the outcome.
Why Gaps Occur
There are two broad reasons why what consumers say and what they do don’t always align.
The first is prediction.
When consumers tell us what they will do in the future, they’re making forecasts: any behavioural scientist will confirm that humans are notoriously poor at forecasting their future behaviour. Not because they’re dishonest, but because future decisions are made in future contexts that do not yet exist.
The second is explanation.
When consumers tell us why they did something, they’re reconstructing a decision after the event.
But…
- memory is imperfect.
- some (a lot) of what influences behaviour is non- conscious
Yet, people naturally create explanations that are coherent, rational and (depending on the case in point) socially acceptable.
In short… consumers are often imperfect predictors of their future behaviours and imperfect witnesses to their past behaviour.
But imperfect doesn’t mean useless.
The Limitation of Behaviour Alone
Over recent times, there are those in the marketing community that have moved from
“Behaviour is important.”
to:
“Behaviour is the only truth.”
That’s a much harder claim to defend.
Behaviour tells us what happened, but it often struggles to tell us:
- What almost happened
- What alternatives were considered
- What the consumer was trying to achieve
- What tensions existed
- What barriers prevented a different outcome
- What opportunities might emerge if those barriers were removed
Behavioural insight is inherently backward looking – it looks at what happened. Yet, growth depends on understanding what could happen under different conditions.
And that’s why understanding motivations and barriers still matter… not because they always predict or explain observed behaviour, but because they often reveal future opportunity.
The Role of Research Design
If consumers struggle to answer some questions directly, the solution isn’t necessarily to stop asking them, as has unfortunately become the fashionable thing to claim in some circles.
The solution is to ask better question, and use more intelligent analysis and interpretation.
For example…
When exploring past behaviour, good qualitative researchers will avoid jumping straight to the “why?”
Instead, they help consumers reconstruct past events and behaviours. They reinstate context. They allow insights to emerge through conversation and observation not just questions. They explore rejected alternatives and uncover tensions and trade-offs. And, they use techniques that help consumers explore motivations, barriers and associations that may otherwise be difficult to articulate or access.
In short, the goal is not simply to collect explanations – it is to help consumers access explanations that may not be immediately available.
In the same vein, good quantitative research often avoids relying solely on direct questioning.
Key Drive Analysis, for example, infers what matters from the relationship between attitudes and outcomes rather than simply asking consumers what’s important.
Implicit and other behaviourally informed techniques can help uncover associations that are difficult to consciously articulate.
Similarly, when predicting likely future behaviour, the best researchers often improve prediction by introducing realistic constraints, trade-offs and competitive choices rather than asking consumers to evaluate ideas in isolation. And they don’t look at respondent answers in isolation, they benchmark against historical outcomes, combine multiple measures and so on.
Many of the most valuable research techniques exist for exactly the same reason:
To reduce reliance on consumers’ ability to perfectly explain or predict their own behaviour in response to an overly simplistic question.
What This Means for Modern Research
One implication of all this is that the quality of research design matters enormously.
If predicting future behaviour and explaining past behaviour are inherently difficult tasks for humans, then simplistic questioning becomes particularly risky.
The growth of DIY platforms, automated surveys and rapid insight tools has made research faster, cheaper and more accessible than ever before.
In many ways, that’s a positive development. But it has also become easier than ever to “collect answers” without fully considering whether we’re asking the right questions, let alone interpreting the answers correctly.
A question such as “How important is sustainability?” may appear straightforward. Yet experienced researchers immediately start considering issues such as social desirability, trade-offs, predictive validity and whether direct questioning is even the best way to answer the business problem.
A Better Way to Think About the Say/Do Gap
The most common interpretation of the say/do gap is that consumers are unreliable.
But rather than throwing the baby out with the bath water, and ignore what consumers say completely – I believe that an altogether better solution is to acknowledge that human behaviour is inherently complicated.
Behaviour tells us what happened.
Motivations tell us what consumers are trying to achieve.
Barriers explain why the two don’t always align.
Great insight comes from understanding all three.
The solution to the say/do gap isn’t choosing between what consumers say and what consumers do.
It’s understanding what each can tell us, and what insight gold dust lies in the very gap that we’ve come to fear so much.