Your Customers Can Tell You Exactly Why They Bought. Almost None of It Is True.

In 1985, Coca-Cola ran 200,000 taste tests, surveys, and focus groups. The data was overwhelming: consumers preferred the new formula. So they launched New Coke.

Within weeks, the company was drowning in 400,000 angry phone calls and letters. They reversed course within 79 days.

The research wasn't flawed. The methodology was sound. The sample size was enormous. The problem was more fundamental: the part of the brain that chose Coca-Cola in a supermarket had nothing to do with the part that answered the survey.

This wasn't a one-off failure. It's a pattern that quietly undermines marketing decisions in businesses of every size, every day. When you ask customers why they bought from you, they give you a clear, logical answer. It sounds credible because they believe it themselves. But the neuroscience of the last four decades tells us something uncomfortable: that answer is, at best, a partial truth. At worst, it's a story their brain invented after the fact.

Your Brain Has a Spin Doctor. You've Never Noticed It Working.

In the 1970s, neuroscientist Michael Gazzaniga was studying split-brain patients: people whose left and right brain hemispheres had been surgically disconnected to treat severe epilepsy. What he found changed our understanding of human decision-making.

In one experiment, Gazzaniga flashed the word "walk" to a patient's right hemisphere (which controls action but can't produce speech). The patient stood up and started walking. When asked why, their left hemisphere, the speaking side, which had zero access to the actual instruction, immediately generated a plausible explanation: "I wanted to get a Coke."

The left hemisphere didn't say "I don't know." It confabulated. It invented a coherent, believable story to explain behaviour it didn't cause. And the patient genuinely believed the explanation they gave.

This is not a quirk of damaged brains. Gazzaniga's decades of subsequent research demonstrated that all human brains do this, constantly. The left hemisphere acts as what he called an "interpreter": a narrative engine that takes whatever information is available and constructs a story that feels rational and intentional.

Daniel Kahneman's Nobel Prize-winning research built on this foundation. His System 1 and System 2 framework showed that fast, automatic, emotional processing (System 1) drives the vast majority of decisions. The slower, deliberate, rational processing (System 2) mostly serves to justify what System 1 already decided.

When you ask a customer "Why did you choose us?", you're asking System 2 to explain a System 1 decision. System 2 doesn't say "I have no idea." It does what Gazzaniga's interpreter does. It constructs a plausible, post-hoc story.

"Your prices were competitive." "I liked your website." "You came recommended."

These aren't lies. They're confabulations: the brain's best guess at why it did what it did. The customer believes them sincerely. That sincerity is what makes them so dangerously misleading.

The Say-Do Gap Costs Billions. We Measured It.

Rory Sutherland, Vice Chairman of Ogilvy UK, captures the problem in a single line: "People don't think what they feel, don't say what they think, and don't do what they say."

This isn't a throwaway observation. In January 2026, NielsenIQ launched its Say-Do Gap Measurement Framework after quantifying the disconnect between what consumers report and what they actually purchase. The finding: the say-do gap has cost the consumer goods industry more than 13 billion unit sales over the past five years.

Thirteen billion units. Not dollars. Units.

NielsenIQ's conclusion deserves quoting directly: "Consumers are not lying when they tell us what they intend to do. Most are answering honestly based on available information, emotions, and context at that moment. But real-world purchasing rarely happens at that same moment, or under those same conditions."

Traditional stated-preference research predicts actual purchase behaviour with only 34% accuracy. Two-thirds of what customers tell you in surveys does not translate into real-world action.

The implication: most customer research captures the 5-20% of decision-making that happens consciously. The rest, the part that actually drove the purchase, never makes it into your survey responses. This connects to something we've explored before: you know what's working in your marketing, but you have no idea why. The same blind spot applies to your customers.

What Byron Sharp Saw That Surveys Can't Show You

This is where it gets really interesting for marketers.

Byron Sharp and the Ehrenberg-Bass Institute have spent 50+ years tracking actual purchase behaviour across 130+ brands in 30+ countries. Their findings directly explain why customer self-reporting fails.

Sharp's central concept is mental availability: a brand's propensity to be noticed or come to mind in buying situations. This is not the same as brand awareness. Awareness is binary (have you heard of us?). Mental availability is contextual: do you think of us at the moment the need arises?

Here's the critical insight for customer research: mental availability operates below conscious awareness. When a plumber needs a new van wrap and calls the first sign company that comes to mind, they can't articulate the advertising exposure, the Google reviews they scrolled past, or the branded vehicle they saw in traffic three weeks ago that primed that association. All of those touchpoints built mental availability. None of them will appear in a "Why did you choose us?" response.

Instead, the customer will say something like "I Googled it and you came up." Which is technically true. But it skips the entire chain of prior brand exposure that made them click your listing instead of the five others on the same page. We've written about this dynamic before in why most buyers don't think about your business, and the implications run deeper than most businesses realise.

Sharp's data shows that when brands double their mental availability while keeping physical availability constant, market share grows by 30-50%. This growth happens through mechanisms that customers cannot self-report because they're not consciously aware of them.

The Advertising Data That Proves the Point

If customers truly made purchases through the rational process they describe in surveys, then advertising that appeals to rational evaluation should outperform advertising that appeals to emotion. It doesn't. Not even close.

Les Binet and Peter Field's analysis of 30 years of IPA effectiveness data found that emotional campaigns outperformed rational campaigns on every single brand metric measured: awareness, trust, differentiation, quality, fame, image, and commitment. Emotional campaigns produced on average 1.7 brand effects compared to 1.0 for rational campaigns.

Emotional campaigns were also almost twice as likely to result in top-box profit growth over the longer term.

As we covered in why logical ads fail, the gap between what sounds right in a boardroom and what actually drives action is enormous. Your customer research reinforces that gap by giving you rational explanations for emotional decisions.

Stop Asking "Why." Start Asking "What Happened."

"Why did you buy?" "Why did you choose us?" "Why not the competitor?"

"Why" puts people on the defensive. It forces them to justify their behaviour, which triggers the exact rationalisation mechanism Kahneman and Gazzaniga identified. You don't get what they actually felt or experienced. You get a post-hoc story constructed to make their decision sound sensible.

Tomlinson's recommendation: lead with open-ended, non-judgmental questions that invite stories, not justifications:

• "Walk me through how you ended up choosing [solution]."
• "What was going on when you started looking for something like this?"
• "What did you try before? What happened?"
• "What almost stopped you from moving forward?"

But the bigger shift isn't in how you ask. It's in what you're looking for.

Instead of trying to extract rational purchase reasons, look for three things:

What This Means for Your Business

Customer feedback isn't useless. But treating it as ground truth is actively dangerous. Here's what changes:

Rory Sutherland puts it best: the most valuable business insights hide in the things people dismiss as irrational, trivial, or impossible to justify in a spreadsheet. The same is true of customer research. The real reasons people buy from you are the ones they'll never write on a survey. Your job is to design research that catches them anyway.

Further Reading

• Why Consumers Say One Thing and Do Another - NielsenIQ's 2026 analysis of the say-do gap and its cost to the industry
• The Subconscious Mind of the Consumer - Harvard Business School's overview of Gerald Zaltman's research on unconscious purchase decisions
• Who's in Charge of Our Minds? The Interpreter - Farnam Street's exploration of Gazzaniga's split-brain research and the left-hemisphere interpreter
• The Long and the Short of It: Binet & Field's Framework Explained - How emotional campaigns outperform rational ones across 30 years of IPA effectiveness data
• Navigating the Say-Do Gap in Consumer Research - Practical methods for bridging stated preferences and actual behaviour