The Say-Do Gap: Why Consumer Research Lies
Conventional research asks. Causal AI observes. Here's why the difference matters for your next decision.
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. First documented by Richard LaPiere in 1934, it shows up everywhere consumer research relies on stated intent: 80% of new products fail despite positive concept-test results. 65% of consumers say they'd pay more for sustainable products; 9% do.
The gap isn't a measurement problem. It's a methodology problem.
Surveys ask people to forecast their own behavior. That requires conscious access to motivations that — as 30 years of cognitive science have shown — are mostly unconscious. The result: well-intentioned, articulate, completely wrong answers.
Three reasons surveys lie
Social desirability bias
People say what they think they should say. Anyone who's worked in market research has watched a focus group consensus form around the most virtuous-sounding answer. The phenomenon is well-studied. It's also unsolvable through better question wording.
Memory reconstruction
"Why did you buy this?" assumes a coherent reason existed and got stored. Neither is reliably true. Most purchase decisions get rationalized after the fact — and the rationale is built from whatever cognitive frame the question creates.
Conscious vs. unconscious
About 95% of consumer decisions happen below conscious awareness. Surveys, focus groups, and conjoint analyses access only the 5% that consumers can articulate. The most predictive drivers of behavior are systematically missing from stated-preference research.
Where the say-do gap costs companies the most
Brand tracking
Stated brand preference: 70% prefer your brand. Actual share: 15%. Reconciling that gap — usually by hiring more research — produces more reports, not better strategy.
Concept testing
80% of new product launches fail. The pre-launch concept tests almost all looked positive. The gap is the failure rate.
Pricing research
Conjoint analysis says customers will pay €X. Market launches at €0.7X. The gap is margin left on the table — or pricing courage destroyed by overly conservative methodology.
Sustainability claims
65% say they'd pay a premium for sustainability. 9% actually do. The largest documented say-do gap in modern consumer research.
How to measure what consumers actually do
Implicit Association Testing (IAT)
Reaction-time-based methodology developed at Harvard in 1998. Measures associations the conscious mind can't deliberately mask. Strong predictor of behavior in domains where social desirability biases stated answers.
Behavioral observation
Where possible: measure actual purchase data, panel data, in-market A/B tests. Better than asking. Usually slower.
Reaction-time measurement
A family of techniques (IAT, AAT, Single-Category IAT) that measure speed of cognitive association. Faster response = stronger underlying preference.
Deep Implicit Framework — SUPRA's methodology
Combines multiple implicit techniques with Causal AI to identify not just what drives behavior, but the causal chain. The output isn't a survey result — it's a decision-grade map of consumer behavior.
"Correlation tells you what happened. Causation tells you what to do."
Frank's take: Why the say-do gap is killing research budgets
CMOs aren't cutting research budgets because they don't believe in research. They're cutting them because too many research projects don't survive contact with reality.
That's the say-do gap doing its work — quietly, project after project, eroding the credibility of the insights function until leadership stops listening.
The way out isn't more research. It's better methodology. Implicit measurement plus causal inference plus advisory that connects findings to decisions. Anything less is theater.
Frequently asked questions
What is the say-do gap?
The say-do gap is the difference between what consumers state in research (intent, preference, willingness to pay) and what they actually do in market. First documented by LaPiere in 1934, it remains the largest source of error in consumer research.
What causes the say-do gap?
Three main causes: social desirability (people say what they think they should), memory reconstruction (purchase reasons get rationalized after the fact), and conscious-vs-unconscious cognition (95% of decisions happen below awareness).
How do you measure the say-do gap?
By comparing stated-preference data (surveys, focus groups, conjoint) against revealed-preference data (actual purchases, behavioral A/B tests) or implicit measurement (IAT, reaction-time tests).
Is the say-do gap the same as intent-action gap?
The terms are often used interchangeably. "Intent-action gap" tends to be used in sustainability and behavior-change contexts; "say-do gap" in consumer research and marketing.
Can the say-do gap be eliminated?
Not entirely — humans don't have full conscious access to their motivations. But it can be measured and accounted for. Implicit methods plus causal modeling can reduce decision risk dramatically.
How big is the say-do gap typically?
It varies by category, but ranges from approximately 20% in low-involvement purchase categories to 80%+ in sustainability claims, sensitive social topics, and aspirational positioning.
Test your strategy against the say-do gap
Most strategic decisions get made with stated-preference data. The ones that survive market contact almost always had implicit and causal evidence behind them.
Get my AI Diagnostic →