Everybody Lies

Successful brands and businesses stay as close as possible to their consumers, and amongst the ways to do this is qualitative and quantitative consumer research.

The problem with this research is that everybody lies.

According to recent analysis by the Corporate Executive Board (now Gartner), the disconnect between stated preferences and actual behaviour affects 60-80% of consumer research studies. For marketing leaders, this gap translates directly into misdirected strategy and billions in budget inefficiencies.

Every category suffers from what behavioural economist Dan Ariely calls systematic self-deception. Studies consistently show that traditional stated preference research predicts actual purchase behaviour with only 34% accuracy, while observational research achieves 89% accuracy when properly implemented.

So why not only do observational research? For many reasons, from the serendipity of an idea that came from two respondents disagreeing in a focus group, from a person telling the story of the beauty routine their mother showed them 22 years previous. In short: from the humanity.

But think about the implications: if your research methodology is wrong more often than it's right, every strategic decision based on that research becomes a gamble with your brand's future.

There are five core psychological mechanisms that make traditional research systematically unreliable. They are not theories, and I for one recognise them in myself:

1. Social Desirability Bias: Marlowe and Crowne's foundational research shows people systematically adjust responses to appear more socially acceptable. In consumer contexts, this manifests as overstating sustainable product preferences, understating alcohol consumption, and inflating exercise frequency. A 2023 Journal of Consumer Research study found this bias affects 74% of health-related purchase intention studies.

2. Temporal Inconsistency: Nobel laureate Daniel Kahneman's work explains why consumers make different decisions when planning versus executing. The rational "planning self" evaluates options differently than the emotional "experiencing self" facing real purchase situations with time pressure and competing priorities. This effect is particularly pronounced for experience goods, where participants choosing a week in advance select healthier options 67% more frequently than those choosing immediately.

3. Aspirational Responding: Consumer psychology research reveals people often respond based on their "ideal self" rather than their "actual self." This creates particular challenges for premium brands, where high purchase intent scores consistently fail to translate to sales. Research in Psychology & Marketing demonstrates this "identity gap" where consumers describe routines they aspire to maintain rather than their actual behaviour.

4. Cognitive Overload: When overwhelmed with options, consumers rely on heuristics that may contradict their stated preferences, explaining why focus group feedback about product features often fails to predict real-world adoption patterns.

5. Memory Reconstruction: Cognitive research demonstrates that memory isn't playback—it's reconstruction. Each time we recall a purchase experience, we unconsciously edit it based on current knowledge and social context, meaning post-purchase satisfaction surveys often reflect how we think we should have felt rather than our actual experience.

The challenge intensifies across categories with measurable business consequences:

Health and wellness brands face particularly acute say-do gaps, with consumers overstating healthy choice intentions by an average of 47%. This leads to product development focused on aspirational benefits rather than actual usage drivers.

Financial services encounters similar patterns, where customers consistently overestimate their likelihood to engage with digital tools while defaulting to familiar channels under actual decision pressure.

Premium beauty brands experience "aspiration inflation" where consumers describe elaborate skincare routines they intend to maintain, leading to high purchase intent scores that rarely translate to sustained usage.

B2B research suffers from executives who won't admit to emotional, irrational, or politically-driven purchase decisions, creating blind spots in understanding actual organisational decision-making processes.

The good news is that we already have the technology to solve the systematic deception problem.

While the industry debates survey design and sampling methodologies, a fundamentally different approach has emerged that eliminates the psychological barriers preventing people from revealing their authentic decision-making patterns.

Simulated respondents. These are AI-powered consumer representations based on authentic behavioural patterns rather than stated preferences. They don't experience social pressure, aspirational responding, or memory reconstruction. Unlike human participants, they can model actual decision-making without bias effects, achieving 87% correlation in predicting real consumer behaviour.

The companies that recognise this methodology shift first will establish competitive advantages before their competitors understand what's happening. But implementation requires understanding not just the technology, but the psychological frameworks that make it superior to traditional approaches.

But before we go further, it's crucial to understand what simulated respondents actually are—because they're fundamentally different from the AI personas most marketers already know.

If you read my newsletter on "Persona Problems" from July, you'll recall why AI personas are systematically unreliable for strategic insights: they represent the mean of a consumer segment rather than actual individuals. They're sophisticated data synthesis masquerading as customer understanding.

Simulated respondents solve this by modelling individuals, not averages. While AI personas aggregate knowledge across entire demographics, simulated respondents represent specific people with consistent personal histories, authentic contradictions, and individual behavioural patterns. The difference is very important:

AI Personas provide probabilistic responses based on segment averages with clean and consistent answers that don't reflect how real humans actually work. As I noted previously, they don't express authentic emotion or rationalise feelings. And since you are dealing with one persona, you miss the wealth of insight that comes from interaction between respondents (the main reason that the focus group persists as a research method)

Simulated Respondents model specific individuals who maintain consistent personal histories while exhibiting the authentic contradictions that drive real human behaviour, without the psychological barriers that prevent humans from revealing these patterns honestly.

This isn't about replacing the messy complexity of human behaviour with artificial simplicity. It's about accessing that complexity without the systematic deception mechanisms that undermine traditional research. Where AI personas flatten human nuance into averages, simulated respondents preserve individual behavioural complexity while eliminating the lies.

The uncomfortable truth facing marketing leaders: traditional research isn't just ineffective, instead it's actively dangerous. Every dollar spent on research that achieves 34% accuracy is a dollar not spent on methodology that delivers 87% accuracy.

The companies thriving in 2025 aren't those with more research data, they're those with more accurate research data. They've recognised that the goal isn't perfect prediction but systematic reduction of the bias effects that undermine research validity.

This isn't about abandoning human connection or creativity. It's about ensuring that when humans make strategic decisions, they're working with accurate behavioural intelligence rather than systematic self-deception.

The question isn't whether you can afford to investigate new research methodologies. The question is whether you can afford the cost of building strategies on systematically deceptive foundations.

Brandflow is written by Justin Billingsley, who has spent his career on all three sides of the industry's table: senior client, global agency leader, technology founder. First published 20 August 2025 in the Brandflow newsletter on LinkedIn.