Segments are Convenient. They're Also Wrong.

Key takeaways

• Demographics (age, gender, income) explain only 4–6% of consumer behavior variance, making them poor targeting tools.
• Demographic profiles of competing brands' customers look nearly identical, proving they're weak proxies for what marketers actually care about. 
• Ad responses are driven by Knowledge, Preference, and Relevance (KPR), not by who someone is, but by what they know, feel, and find personally meaningful. 
• Unlike demographics, KPR dimensions are measurable, actionable, and can be shifted through smarter creative and messaging strategy.

The Uncomfortable Number in the Room

Here is a number that should unsettle every market researcher who has ever built a targeting brief around age, gender, income, or geography: demographic variables explain roughly 4–6% of the variance in consumer behavior.

Four to six percent. That is not a rounding error, but a structural problem.

And yet the industry continues to organize campaigns, allocate budgets, and evaluate advertising performance through exactly these lenses. We ask whether our message landed with women 25–54, whether men over 45 engaged, and whether the affluent suburban segment responded. We refine our demographic models and call it precision targeting.

What we are actually doing is optimizing a proxy. And the real signal — the one that governs how any given brain will process any given ad — is sitting somewhere else entirely.

Competing Brands Sell to the Same People

Let's start with what the data actually says, because it is far more radical than most marketers are willing to accept.

Kennedy and colleagues examined customer profiles of competing brands across more than 40 industries, including demographics, psychographics, attitudes, and media habits. Their finding: brand user profiles differed by an average of just 2–3 percentage points across over 110,000 individual profile comparisons. Ford owners and Chevrolet owners look the same. Nike buyers and Adidas buyers look the same. This finding has been replicated across 50 product categories and 25 years of data.

Byron Sharp, in How Brands Grow, drew the logical conclusion: you cannot reliably boost sales by tailoring your marketing to a specific demographic, because in most categories, your buyers and your competitors' buyers are essentially the same people.

The implications are jarring. Catalina Marketing's analysis of $415 million in TV ad spend found that 53% of brand sales fell entirely outside the conventional demographic target. For frozen dinners, the figure was 60%. Binet and Field, analyzing hundreds of IPA Effectiveness Award cases, found that campaigns targeting the whole market achieved three times as many large business effects as those targeting narrow segments.

We are not just leaving reach on the table. We are using a ruler to target and are wondering why our measurements don't match the territory.

The Brain Doesn’t Check Its Demographics

When a consumer encounters an ad, their brain doesn't start by retrieving their age or income bracket. It does something far more specific. It asks three questions, almost simultaneously and almost entirely below the level of conscious awareness:

Do I know this? Is there a stored memory structure — a schema, a network of associations — that this ad connects to? If yes, the brain processes it quickly and efficiently. If no, the message is routed through slower, more effortful, and often more skeptical circuits.

How do I feel about it? Existing brand attitudes function as automatic, pre-stored evaluations. Russell Fazio's research on attitude accessibility showed that strong prior attitudes are activated instantly upon brand exposure, shaping what we notice, how we interpret what we see, and whether we act. This isn't deliberation, but a filter running before deliberation begins.

Does this matter to me, right now? Personal relevance is perhaps the strongest moderator of all. When information connects to an active need or self-concept, the brain engages its self-referential processing network — the medial prefrontal cortex and default mode network — producing dramatically deeper encoding and better memory.

These three questions — knowledge, preference, relevance (KPR) — correspond to three fundamentally different neural states. And two consumers with identical demographic profiles but different answers to these questions will process the same ad as if they were different species.

The Pepsi vs Coke Experiment and What It Really Tells Us

The famous Pepsi-Coke fMRI neuroscience study by Sam McClure and colleagues back in 2004 is often cited as evidence that branding is powerful. But its deeper lesson is about the primacy of prior knowledge.

In blind taste tests, Coke and Pepsi produced no meaningful difference in neural response. But when brand identity was revealed, self-described Coke lovers showed dramatic activation in the hippocampus and dorsolateral prefrontal cortex — memory retrieval and cognitive control systems. Their prior knowledge of the Coke brand didn't just bias their judgment. It rewrote the neural representation of the sensory experience itself.

Esch and colleagues extended this with fMRI studies comparing strong, familiar, and unknown brands. Strong brands activate reward circuitry and memory systems. Unfamiliar brands activated the insula — associated with risk and aversion — and triggered effortful, skeptical processing. The same ad creative, encountered by two consumers with different levels of brand knowledge, is literally a different experience in the brain.

This is not a metaphor. It is a measurable, replicable neuropsychological fact. And it has nothing to do with the consumer's age or gender.

Why Relevance Is the Gatekeeper

The third variable in this framework is the one most poorly captured by any existing segmentation approach.

Rogers, Kuiper, and Kirker demonstrated the self-reference effect in 1977. They found that information encoded by relating it to oneself yields the best memory of any encoding strategy. Better than semantic depth and better than structural processing. Symons and Johnson's meta-analysis of 129 studies confirmed this robustly. The self acts as a master framework (a "superordinate schema") that both encourages deeper thought and organizes incoming information.

In practical terms, this means that two demographically identical consumers — say, two 40-year-old women — will have entirely different neural responses to the same ad for running shoes depending on whether running is currently part of their self-concept. A person who just signed up for a half-marathon will encode the ad deeply, emotionally, and memorably. Someone who has no interest in running, however, will barely notice it.

Judy Zaichkowsky's work on personal involvement, Celsi and Olson's research on felt involvement, and the entire Jobs to Be Done framework by Clayton Christensen all converge on the same insight: what matters is not who the consumer is, but what they are currently trying to do, become, or solve. As Christensen himself put it: "The fact that you're 18 to 35 years old with a college degree does not cause you to buy a product. It may be correlated with the decision, but it doesn't cause it."

Demographic variables are, at best, distal correlates of relevance. They tell you something about the probability that the right need state might be present. But they cannot tell you whether it is — and they certainly cannot tell you with the precision needed to allocate creative strategy or media spend.

The 3D Framework: What to Measure Instead

If demographic segments are proxies, what are they proxies for? Drawing on converging evidence from marketing science, consumer psychology, and neuroscience, I propose three fundamental variables that explain the vast majority of variance in consumer response to any ad:

Knowledge — Does the consumer know this brand, and how richly? Knowledge determines schema availability, processing fluency, and consideration set membership. Consumers with rich brand schemas process advertising faster, remember it better, and are more likely to recall the brand in buying situations. Low-knowledge consumers need to expend more mental energy (cognitive effort) to process the same message, often perceiving the brand as risky or unfamiliar. Advertising strategy should therefore differ radically depending on where a consumer sits on the knowledge spectrum, not on which age bracket they occupy.

Preference — Does the consumer hold a pre-existing positive attitude toward the brand, and how accessible is that attitude? Fazio's research shows that highly accessible attitudes function as automatic orienting filters — they direct attention, color perception, and ultimately behavior, all before conscious deliberation begins. Hilke Plassmann and colleagues demonstrated that brand favorability modulates predicted-value signals in the ventromedial prefrontal cortex, thereby directly influencing experienced pleasantness. An ad reaching a consumer with strong positive brand preference is operating in an entirely different psychological environment than the same ad reaching a neutral or unfamiliar consumer. Again, this has nothing to do with age or gender.

Relevance — Is the product or category currently relevant to this consumer's active needs, life situation, or self-concept? This is the gating variable. Without relevance, even perfect knowledge and strong preference will not produce a response. With relevance, even a relatively unknown brand can win attention and engagement. Relevance is situational and dynamic — it shifts with life events, seasons, contexts, and states of mind. This is precisely why it resists demographic capture: a 55-year-old and a 25-year-old can have identical relevance to a given category in the same moment, while two demographically identical 35-year-olds can have completely different relevance profiles depending on what is happening in their lives.

What This Means for Research Practice

The argument here is not that demographic data is worthless. It has legitimate uses in media planning, market sizing, and regulatory contexts. The problem arises when demographics are used as explanatory variables for ad response — when we assume that "women 35–54" tells us something meaningful about how those individuals will process a given campaign.

For behavioral researchers, this framework suggests a different set of questions to ask and a different set of signals to measure: 

• Rather than asking "Who is our target?", ask "What do our targets already know and believe about us?" Measure brand familiarity and depth of knowledge, not just demographics. 
• Rather than asking "Which segment shows the strongest purchase intent?", ask "Among whom is our category currently relevant, and what situational triggers are activating that relevance right now?" This is the insight that Category Entry Point research and need-state segmentation are beginning to address. 
• Rather than assuming that demographic targeting improves message relevance, test whether it actually does. In many categories, a contextual or behavioral signal will outperform a demographic one. 

The good news is that predictive tools — including AI-driven attention and emotion measurement — will soon be able to operationalize these variables at scale. These solutions will be able to predict, with increasing accuracy, how consumers at different levels of brand knowledge will process a given creative. They will be able to identify category entry points that signal active relevance and measure whether an ad is generating the neural engagement associated with deep encoding and attitude formation, rather than passive exposure.

The infrastructure for a better kind of segmentation exists. What we need is the willingness to stop organizing our work around variables that are easy to measure but weakly connected to what we actually care about.

The Real Question behind Your Targeting Brief

There is nothing wrong with knowing that your typical buyer is female, middle-income, and between 30 and 50 years old. The problem is what you do with that knowledge.

If you use it as a proxy for psychological state — assuming that "women 30–50" are likely to be in the right knowledge, preference, and relevance space for your brand — you are making a leap that the evidence does not support. Your research should stress-test that assumption, not embed it.

Demographics describe who people are. The brain responds to what people know, what they want, and what matters to them right now. Until we start building segmentation frameworks around those variables, we will keep optimizing for the wrong thing — and wondering why the numbers don't add up.

References

1. Kennedy R., Ehrenberg A., & Long S. (2000) Competitive Brands’ User-Profiles Hardly Differ. Market Research Society Conference (UK), Brighton, London: Market Research Society. https://www.researchgate.net/publication/260555051_Brand_user_profiles_seldom_differ
2. Uncles, M., Kennedy, R., Nenycz-Thiel, M., Singh, J., & Kwok, S. (2012). In 25 years, across 50 categories, user profiles for directly competing brands seldom differ—Journal of Advertising Research. https://www.tandfonline.com/doi/full/10.2501/JAR-52-2-252-261
3. Sharp, B. (2010). How Brands Grow. Oxford University Press. https://global.oup.com/academic/product/how-brands-grow-9780195573565?cc=dk&lang =en&
4. Fennell, G., Allenby, G., Yang, S., & Edwards, Y. (2003). The effectiveness of demographic and psychographic variables for explaining brand and product category use. Quantitative Marketing and Economics, 1(2). https://faculty.marshall.usc.edu/Sha-Yang/SMM_QME.pdf
5. Binet, L. & Field, P. (2013). The Long and the Short of It. IPA. https://maynardpaton.com/wp-content/uploads/2019/12/SYS1-long-and-short-of-it.pdf
6. McClure, S.M., Li, J., Tomlin, D., Cypert, K.S., Montague, L.M., & Montague, P.R. (2004). Neural correlates of behavioral preference for culturally familiar drinks. Neuron, 44(2), 379–387. https://pubmed.ncbi.nlm.nih.gov/15473974/
7. Esch, F.R., Möll, T., Schmitt, B., Elger, C., Neuhaus, C., & Weber, B. (2012). Brands on the brain: Do consumers use declarative information or experienced emotions to evaluate brands? Journal of Consumer Psychology, 22(1). https://psycnet.apa.org/record/2012-06094-008
8. Plassmann, H., Ramsøy, T.Z., & Milosavljevic, M. (2012). Branding the brain: A critical review and outlook. Journal of Consumer Psychology, 22(1), 18–36. https://www.sciencedirect.com/science/article/abs/pii/S1057740811001136
9. Fazio, R.H., Powell, M.C., & Williams, C.J. (1989). The role of attitude accessibility in the attitude-to-behavior process. Journal of Consumer Research, 16(3). https://psycnet.apa.org/record/1990-18609-001
10. Symons, C.S., & Johnson, B.T. (1997). The self-reference effect in memory: A meta-analysis. Psychological Bulletin, 121(3), 371–394. https://psycnet.apa.org/record/1997-03609-003
11. Celsi, R.L., & Olson, J.C. (1988). The role of involvement in attention and comprehension processes. Journal of Consumer Research, 15(2). https://psycnet.apa.org/record/1989-10455-001
12. Christensen, C. (2016). Milkshake marketing. Harvard Business School Working Knowledge. https://www.library.hbs.edu/working-knowledge/clay-christensens-milkshake-marketing
13. Ramsøy, T.Z. (2024). How to Make People Buy: The Art & Science of Enabling, Engaging, and Empowering Your Customers. Amazon