The Pepsi Challenge — When Taste Doesn't Equal Choice
Covers lectures
F2-02 · F2-07 · F2-09
The Pepsi Challenge — When Taste Doesn't Equal Choice
Module: F2 — Consumer Behaviour Type: Neuromarketing Case Cross-references: F2-02 (dual-process theory), F2-07 (brand associations and memory structures), F2-09 (the role of emotion in decision-making)
The Situation
In 1975, Pepsi launched what would become one of the most famous marketing experiments in history. The concept was disarmingly simple: set up tables in shopping centres, pour Coca-Cola and Pepsi into unmarked cups, and ask consumers to taste both and choose which they preferred. No branding. No labels. Just liquid and preference.
The results were consistent and, for Pepsi, exhilarating. In blind taste tests across the United States, a majority of participants — typically in the range of 53-58% — chose Pepsi over Coca-Cola. The sweeter flavour profile of Pepsi, it appeared, was objectively preferred by the American palate. Pepsi had the better product. The data proved it.
And yet Coca-Cola dominated the market. At the time of the Pepsi Challenge's launch, Coca-Cola held approximately twice Pepsi's market share in the US cola category. Through the 1980s and 1990s, despite the Challenge being run continuously, Coca-Cola maintained its commanding lead. By the early 2000s, Coca-Cola's global brand value was estimated at over $67 billion by Interbrand — the world's most valuable brand — while Pepsi, despite its taste test victories, remained a distant second.
This is the paradox at the heart of consumer behaviour theory. If consumers prefer the taste of Pepsi, why do they buy Coca-Cola? If explicit preference — the kind you can measure with a blind test — predicted market behaviour, Pepsi should have been the market leader. It was not. Not then, not now.
The answer to this paradox lies not in the mouth but in the brain. And in 2004, a neuroscience study provided the first direct neurological evidence of why taste and choice are fundamentally different cognitive processes — and why brand equity is not a metaphor but a measurable force operating in the neural architecture of decision-making.
The Data
The Original Pepsi Challenge
The Pepsi Challenge, as designed by PepsiCo's marketing team, was a straightforward paired comparison test. Consumers were presented with two unmarked cups — one containing Pepsi, one containing Coca-Cola — and asked to taste both and indicate which they preferred. The test was conducted in public spaces (shopping centres, high streets, sports events) and filmed for use in advertising.
The methodology, while compelling as marketing theatre, had known limitations from a sensory science perspective.
The sip test problem. Sensory researchers have long noted that single-sip tests favour sweeter products. Pepsi has a slightly higher sugar content and a more immediately detectable sweetness than Coca-Cola, which has a more complex, slightly spicier flavour profile. In a single sip, sweetness wins. Over a full can or bottle — the actual consumption context — the preference picture may differ. Malcolm Gladwell explored this distinction in Blink (2005), noting that what people prefer in a sip is not necessarily what they prefer over a full serving.
Demand characteristics. As a public marketing event, the Pepsi Challenge was not a controlled experiment. Participants knew they were being filmed, knew it was a Pepsi-sponsored event, and may have been influenced by the social context. However, independent replications of blind taste tests — conducted under more controlled conditions by academic researchers — have broadly confirmed that Pepsi does tend to be preferred in unlabelled paired comparisons, though the margins vary.
The critical finding. Regardless of methodological quibbles, the core observation remained robust across decades of testing: when consumers could not see the brand, a majority preferred Pepsi. When they could see the brand, a majority chose Coca-Cola. The brand was not just a label — it was changing the experience itself.
The McClure et al. fMRI Study (2004)
In 2004, Samuel McClure, Jian Li, Damon Tomlin, Kim Cypert, Latane Montague, and Read Montague published a landmark study in the journal Neuron that provided the first neuroimaging evidence of how brand knowledge alters sensory perception at the neural level.
The study design was elegant. Sixty-seven participants underwent functional magnetic resonance imaging (fMRI) while tasting Coca-Cola and Pepsi under two conditions: blind (no brand information) and branded (participants told which drink they were tasting).
Blind condition results. When participants did not know which drink they were consuming, the brain activation patterns were consistent with a straightforward sensory preference task. The ventromedial prefrontal cortex (VMPFC) — a brain region associated with reward processing and the subjective experience of pleasantness — showed activation that correlated with stated taste preference. Participants who preferred Pepsi showed stronger VMPFC activation when drinking Pepsi, and vice versa. The neural data matched the behavioural data: in blind tasting, preference was driven by the sensory experience of the liquid itself.
Critically, in the blind condition, roughly half the participants preferred each drink — a result consistent with the commercial Pepsi Challenge data showing a slight Pepsi advantage but no overwhelming dominance.
Branded condition results. When participants knew they were drinking Coca-Cola, the neural picture changed dramatically. Knowledge of the Coca-Cola brand activated a different set of brain regions: the hippocampus (memory), the dorsolateral prefrontal cortex (DLPFC — cognitive control and higher-order processing), and the midbrain (dopaminergic reward pathways). These regions are not associated with taste perception. They are associated with memory retrieval, emotional processing, and the integration of prior knowledge with current experience.
The crucial finding: when participants knew they were drinking Coca-Cola, their stated preference shifted decisively toward Coke — and the VMPFC activation pattern shifted accordingly. The brand knowledge was not merely influencing the cognitive report ("I know I should prefer Coke"). It was altering the subjective sensory experience itself. The drink tasted different — better — when participants knew it was Coca-Cola.
The Pepsi asymmetry. Fascinatingly, the brand effect was asymmetric. Knowledge of the Pepsi brand did not produce the same neural shift. When participants knew they were drinking Pepsi, the activation patterns were similar to the blind condition. Pepsi's brand did not override the sensory signal in the way Coca-Cola's did. This asymmetry is consistent with the market data: Coca-Cola's brand equity was not just stronger in marketing terms — it was stronger in neurological terms. It had a greater capacity to modulate the actual experience of consumption.
The Dual-Process Interpretation
The McClure et al. findings map directly onto the dual-process framework that underpins much of modern consumer behaviour theory.
System 1 and System 2. Daniel Kahneman's (2011) distinction between System 1 (fast, automatic, intuitive) and System 2 (slow, deliberate, analytical) provides the interpretive framework. In the blind condition, participants were engaging primarily in System 2 processing: consciously evaluating a sensory experience and making a deliberate judgement. System 2 processed the taste and, in many cases, concluded that Pepsi was preferable.
In the branded condition, System 1 took over. The Coca-Cola brand activated a dense network of associations — memories, emotions, cultural meanings, identity signals — that operated below conscious deliberation. These associations did not argue with the taste data. They overwhelmed it. The System 1 response — "This is Coca-Cola, and Coca-Cola is good" — was faster, more emotionally charged, and more influential than the System 2 taste evaluation.
The speed advantage. System 1 processing is not merely an alternative to System 2. It is faster. In real-world purchase decisions — standing in a shop, scanning a shelf, choosing between dozens of options in seconds — System 1 wins by default because System 2 does not have time to engage. The blind taste test is an artificial environment that gives System 2 its best possible chance. Remove the brand, remove the time pressure, remove the competing stimuli, and System 2 can process taste. In the real world, none of these conditions hold.
Broader Replication and Context
The McClure et al. study has been cited over 2,500 times and has catalysed a substantial body of neuromarketing research examining how brand knowledge modulates perception.
Koenigs and Tranel (2008) studied patients with damage to the ventromedial prefrontal cortex — the brain region implicated in integrating emotion with decision-making. These patients showed no brand preference effect: they chose based on taste alone, even in the branded condition. This finding provided causal evidence that the VMPFC is necessary for the brand effect to operate, not merely correlated with it. Without the neural machinery for emotional integration, brand equity has no purchase on choice.
Plassman et al. (2008) demonstrated a similar effect with wine. When participants were told a wine cost $90 versus $10 (regardless of the actual wine), the medial orbitofrontal cortex — a region overlapping with the VMPFC — showed increased activation for the "expensive" wine. The wine literally tasted better when participants believed it was expensive. Price, like brand, is not just information — it is a perceptual modifier.
De Araujo et al. (2005) showed that verbal labels ("rich and delicious" versus "body odour") applied to identical odours changed neural activation patterns in the orbitofrontal cortex. The label changed the smell. The principle is the same: contextual information — brand, price, description — does not merely influence the report of an experience. It alters the experience itself at the neural level.
The Analysis
The Gap Between Preference and Choice
The Pepsi Challenge paradox — preferring one product but choosing another — is not an anomaly. It is the normal condition of consumer behaviour, and it reveals a fundamental error in how many marketers think about their category.
The error is this: the assumption that product superiority drives market share. If your product tastes better, cleans better, performs better, consumers will choose it. This is the rational model of consumer choice — what economists call "utility maximisation" — and the Pepsi data demolishes it.
Consumers do not choose the product they prefer. They choose the product they reach for. And what determines reaching behaviour is not the outcome of a careful sensory evaluation. It is the speed and strength of the associations activated at the moment of choice. Coca-Cola's brand equity — built through decades of consistent advertising, distinctive assets (the contour bottle, the Spencerian script, the red disc), and cultural embedding — creates a set of System 1 associations so powerful that they override a taste advantage.
This is not irrationality. It is a different kind of rationality. The brain has evolved to use shortcuts because full evaluation of every choice would be cognitively paralysing. Brand associations are not errors — they are efficient heuristics. The problem is not that consumers are wrong to choose Coca-Cola. The problem is that marketers are wrong to think the taste test measures what matters.
Brand Equity as Neural Architecture
The McClure et al. study reframes brand equity from a marketing abstraction to a neurological reality. Brand equity is not "what consumers think about your brand." It is the density, accessibility, and emotional valence of the neural pathways that your brand activates.
Coca-Cola's brand equity, in this framework, is a physical structure — a network of synaptic connections linking the brand to memories (childhood, summer, celebration), emotions (happiness, nostalgia, belonging), and sensory cues (the fizz, the colour red, the bottle shape). This network was built over more than a century of consistent brand communication, and it is activated automatically — without conscious effort — whenever a consumer encounters the brand.
Pepsi's brand, by contrast, activated a thinner neural network. This is not because Pepsi has done poor marketing. It is because Coca-Cola has done extraordinary marketing — so extraordinary that it has, quite literally, rewired the brains of billions of consumers to experience its product as better than it tastes.
This has profound implications for how we think about brand building. Brand equity is not a soft metric, a nice-to-have, or a luxury for companies with large budgets. It is the mechanism by which markets are won and held. A brand with strong neural architecture will be chosen over a product that is objectively superior — because at the moment of choice, the brain does not conduct an objective evaluation. It activates a network.
The Marketing Implications
Product quality is necessary but not sufficient. The Pepsi Challenge does not show that product quality is irrelevant. Pepsi must be a competent product — if it tasted terrible, no amount of branding would save it. But competence is the entry fee. Beyond competence, brand associations determine market outcomes far more than marginal product differences.
Blind tests mislead. Any marketer who relies on blind product tests to predict market performance is making the same error Pepsi made. Blind tests strip away the very factor — brand associations — that most determines choice. They measure System 2 preference in a world where System 1 decides. This has implications for product development, market research methodology, and competitive strategy. The question is not "Is our product better in a blind test?" The question is "Is our brand more mentally available at the moment of choice?"
Consistency builds neural pathways. Coca-Cola's brand advantage is not the result of a single campaign or a single year of marketing. It is the cumulative result of more than a century of consistent visual identity, consistent emotional territory, and consistent presence. Neural pathways strengthen through repetition. Every exposure to the Coca-Cola brand — every advertisement, every vending machine, every restaurant glass — reinforces the network. This is why Byron Sharp's prescription for brand building emphasises consistency and reach: you are not just communicating a message. You are building architecture in the brains of every person in the market.
Emotional associations beat rational claims. The Coca-Cola brand activates emotional and memory regions, not analytical ones. This is not an accident — it is the result of decades of advertising that has emphasised feelings (happiness, togetherness, refreshment) over functional claims (taste, ingredients, calories). Pepsi, by contrast, has historically oscillated between emotional campaigns and rational product claims — including the Pepsi Challenge itself, which is fundamentally a rational argument ("We taste better"). The irony is acute: Pepsi's most famous marketing campaign was an argument for System 2 evaluation in a market where System 1 decides.
The Both/And Resolution
The Pepsi Challenge case is not a simple story of emotion beating rationality. It is a Both/And case.
The product must be good AND the brand must be strong. Coca-Cola's formula is not irrelevant — it provides the sensory foundation upon which the brand associations rest. If Coca-Cola tasted like cough syrup, no amount of brand equity would sustain it. The product provides the rational basis; the brand provides the emotional shortcut.
Similarly, the Pepsi Challenge was not a failure — it was an incomplete strategy. It proved product competence (System 2) but did not build the emotional architecture (System 1) needed to convert that competence into choice. The lesson is not that Pepsi should have ignored taste. The lesson is that taste alone was never going to be enough.
The most effective marketing strategies do both: they ensure the product is genuinely good, AND they invest in building the dense, emotionally rich brand associations that determine choice at the moment of decision. Product quality without brand equity leaves you winning tests and losing markets. Brand equity without product quality leaves you vulnerable to any competitor who delivers a better experience. The Both/And is the only sustainable position.
The Questions
F2-02 Application. Apply the dual-process framework (System 1 and System 2) to the Pepsi Challenge paradox. Why does System 2 preference (blind taste) fail to predict System 1 choice (branded purchase)? What does this tell us about the limits of rational consumer research methodologies?
F2-07 Application. Using the concept of memory structures and brand associations, explain how Coca-Cola's brand equity operates at the neurological level. How do consistent distinctive assets and emotional advertising build the neural pathways that determine choice? What would a brand need to do to build similar neural architecture from scratch?
F2-09 Application. The McClure et al. study demonstrates that brand knowledge alters the subjective experience of consumption — the drink literally tastes different when you know the brand. What are the ethical implications of this finding? If marketing can change how a product is experienced, does the concept of "objective product quality" have any meaning in consumer markets?
Sources
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.
Kahneman, D. (2011). Thinking, Fast and Slow. Penguin.
Gladwell, M. (2005). Blink: The Power of Thinking Without Thinking. Penguin.
Koenigs, M. & Tranel, D. (2008). "Prefrontal Cortex Damage Abolishes Brand-Cued Changes in Cola Preference." Social Cognitive and Affective Neuroscience, 3(1), 1-6.
Plassman, H., O'Doherty, J., Shiv, B. & Rangel, A. (2008). "Marketing Actions Can Modulate Neural Representations of Experienced Pleasantness." Proceedings of the National Academy of Sciences, 105(3), 1050-1054.
De Araujo, I.E., Rolls, E.T., Velazco, M.I., Margot, C. & Cayeux, I. (2005). "Cognitive Modulation of Olfactory Processing." Neuron, 46(4), 671-679.
Sharp, B. (2010). How Brands Grow: What Marketers Don't Know. Oxford University Press.
Binet, L. & Field, P. (2013). The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies. IPA.