Neuromarketing: the Hope and Hype of Neuroimaging in Business

Nature Reviews Neuroscience | AOP, published online 3 March 2010; doi:10.1038/nrn2795 PersPecTives

we aim to distinguish the legitimate hopes science and society from the marketing hype. As such, we hope that this article serves the dual purpose of rec- Neuromarketing: the hope and hype ognizing the real potential of neuro imaging in business and providing a guide for potential of neuroimaging in business buyers and sellers of such services. Why use brain imaging for marketing? Dan Ariely and Gregory S. Berns Marketers are excited about brain imaging Abstract | The application of neuroimaging methods to product marketing — for two main reasons. First, marketers hope that neuroimaging will provide a more effi- neuromarketing — has recently gained considerable popularity. We propose that cient trade-off between costs and benefits. there are two main reasons for this trend. First, the possibility that neuroimaging This hope is based on the assumptions that will become cheaper and faster than other marketing methods; and second, the people cannot fully articulate their prefer- hope that neuroimaging will provide marketers with information that is not ences when asked to express them explicitly, obtainable through conventional marketing methods. Although neuroimaging is and that consumers’ brains contain hidden information about their true preferences. unlikely to be cheaper than other tools in the near future, there is growing evidence Such hidden information could, in theory, that it may provide hidden information about the consumer experience. The most be used to influence their buying behaviour, promising application of neuroimaging methods to marketing may come before a so that the cost of performing neuroimaging product is even released — when it is just an idea being developed. studies would be outweighed by the benefit of improved product design and increased sales. In theory, at least, brain imaging could Despite many common beliefs about the and the product, production and distribu- illuminate not only what people like, but also inherently evil nature of marketing, the main tion systems have to be in place for market what they will buy. objective of marketing is to help match prod- tests to be conducted. There are some Thus far, this approach to neuromarketing ucts with people. Marketing serves the dual compromise approaches between these two has focused on this post-design application, goals of guiding the design and presentation extremes, which include simulated markets, in particular on measuring the effective- of products such that they are more compat- conjoint analyses, markets for information ness of advertising campaigns. The general ible with consumer preferences and facili- and incentive-compatible pricing studies approach has been to show participants a tating the choice process for the consumer. (see Supplementary information S1 (box)). product advertisement, either in the form Marketers achieve these goals by providing As in all compromises, these approaches of a print advertisement or commercial, and product designers with information about provide solutions with intermediate levels measure the brain’s response in the form of a what consumers value and want before a of cost, simplicity, realism and quality of blood oxygenation level-dependent (BOLD) product is created. After a product emerges data (TABLE 1). measurement, which is taken as a proxy for on the marketplace, marketers attempt to The incorporation of neuroimaging into neural activation. maximize sales by guiding the menu of the decision-making sciences — for example, The second reason why marketers are offerings, choices, pricing, advertising and neuroeconomics — has spread to the realm excited about brain imaging is that they promotions. of marketing. As a result, there are high hope it will provide an accurate marketing In their attempts to provide these types hopes that neuroimaging technology could research method that can be implemented of inputs, marketers use a range of market solve some of the problems that market- even before a product exists (FIG. 1). The research techniques, from focus groups ers face. A prominent hope is that neuro- assumption is that neuroimaging data and individual surveys to actual market imaging will both streamline marketing would give a more accurate indication of tests — with many approaches in between processes and save money. Another hope is the underlying preferences than data from (see Supplementary information S1 (box)). that neuroimaging will reveal information standard market research studies and would In general, the simpler approaches (focus about consumer preferences that is unob- remain insensitive to the types of biases groups and surveys) are easy and cheap to tainable through conventional methods. Of that are often a hallmark of subjective implement but they provide data that can course, with such high expectations, there approaches to valuations. If this is indeed include biases, and are therefore seen as is the accompanying hype. Several popular the case, product concepts could be tested not very accurate1–4. The approaches that books and articles have been published that rapidly, and those that are not promising are more complex and therefore harder to push a neuromarketing agenda, and there eliminated early in the process. This would implement, such as market tests, provide are now a handful of companies that market allow more efficient allocation of resources more accurate data but incur a higher cost, neuromarketing itself 5. In this Perspective, to develop only promising products.

Table 1 | comparison of selected marketing research approaches Focus groups Preference simulated choice Market tests questionnaires methods What is measured Open-ended answers, importance weighting for choices among products Decision to buy and body language and various product attributes choice among products behaviour; not suitable for statistical analysis Type of response process speculative, except when The respondent must try A hypothetical choice, An actual choice, with used to assess prototypes to determine his decision so the same process as customers’ own money, weightings through the actual purchase — and therefore fully introspection, then map but without monetary consequential those weightings into the consequences response scale Typical use in new-product early on to aid general Design phase, when Design phase, when end of process, to forecast development processes product design; at user determining customer determining customer sales and measure interface design for trade-offs is important trade-offs is important; the response to other usability studies may also be used as a elements of marketing, forecasting tool such as price cost and competitive risk Low cost; risk comes only Moderate cost and Moderate cost (higher High cost and high risk of from misuse of data by the some risk of alerting if using prototypes alerting competitors, plus seller competitors instead of descriptions) the risk of the product and some risk of alerting being reverse engineered competitors before launch Technical skill required Moderation skills for Questionnaire design and experiment design running an instrumented inside the group and statistical analysis and statistical analysis market and forecasting ethnographic skills for (including choice (highly specialized) observers and analysts modelling)

Thus, the issue of whether neuroimaging Similarly, we suspect (and economists are they were willing to pay (a measure of deci- can play a useful part in any aspect of market- certain) that methods that are incentive sion utility) correlated with activity levels in ing depends on three fundamental questions, compatible are better than methods that the medial orbitofrontal cortex (OFc) and which we will address in this paper. First, can are not. Incentive-compatible elicitation prefrontal cortex (PFc)9,10. Interestingly, neuromarketing reveal hidden information methods are methods that encourage the similar activation in the OFc has been that is not apparent in other approaches? participant to truthfully reveal what is observed when subjects anticipate a pleasant Second, can neuromarketing provide a more being asked of him because to do so would taste11, look at pretty faces12, hear pleasant efficient cost–benefit trade-off than other maximize the participant’s satisfaction (for music13, receive money14,15 and experience a marketing research approaches? Third, can example, he would earn the most money social reward16,17. Such generally close corre- neuromarketing provide early information or receive the product he likes the best). In spondence in regional brain activity between about product design? other words, it is in the participant’s interest the anticipation of rewarding events, the to answer product-related questions truth- consumption of enjoyable goods and the Revealing hidden information fully. However, using such methods is not willingness to pay for them suggests that Brain activity and preference measurement. always possible. the representation of expected utility may Allowing for the assumption in neuro- One important question for the rely, in part, on the systems that evaluate the marketing that the brain contains hidden potential of neuromarketing is whether quality of the consumption experience. The information about preferences, it is reason- the neural signal at the time of, or slightly theme of common systems for expectation able to set aside, for the moment, the issue before, the decision (assumed to be a and experience also applies to things that are of ‘hidden’ and ask what relationships are measure of decision utility) can be a good unpleasant or even painful (although this known to exist between brain activity and predictor of the pleasure or reward at the involves a different network including the expressed (that is, not hidden) preference. time of consumption (the ‘experienced insula)18–21. Such similarities suggest that As it turns out, different methods of utility’)8. A second question is whether the neuroimaging can become a useful tool in eliciting a person’s preference often result in link between these two signals holds even measuring preferences, particularly when different estimations of that preference3,4,6,7. when the preference elicitation methods incentive compatibility is important but there This makes it difficult to know which are not incentive compatible. If the answer is no easy way to achieve it (for example, method provides the truest measure of to both of these questions is positive, when the products have not been created). ‘decision utility’ (that is, the expected utility, neuromarketing could become useful for However, such similarities do not necessar- which would ultimately drive choice in the measuring preferences. ily mean that brain activation is the same marketplace). It is clear that market tests give Measurements such as willingness to across different elicitation methods, and there the most accurate answer, but having to run pay (wTP) have only recently come under are differences between the neural activa- a market test on every product would defeat functional MrI (fMrI) examination. In one tion representing decision utility and that the purpose of market research — namely, experiment, subjects bid on the right to eat representing experienced utility14,22,23. to provide early and cheap information. snacks during the experiment. The amount This caveat aside, the generally close

correspondence does suggest that neural activity might be used as a proxy for wTP fMRI in situations in which wTP cannot easily be determined — although this has yet to be demonstrated.

Reverse inference and reward. The practice Concept Design Testing of measuring an increase in BOLD activity Market analysis • Development in a region such as the ventral striatum or feasibility • Prototyping OFc and then concluding that a ‘reward- related’ process was active has become Delivery increasingly common. This form of deduc- tive reasoning is known as ‘reverse infer- Release ence’24,25. Given the readiness of many to interpret brain activation as evidence of a Support specific mental process, it is worth examining this type of inference. using a Bayesian analysis, it is possible to estimate the spe- Feedback cificity of activation in a particular region of the brain for a specific cognitive process. fMRI For example, Poldrack used the BrainMap Advertising database to analyse the frequency of activation of Broca’s area in language studies24. He Figure 1 | Product development cycle. Neuromarketing applications of functional Mri (fMri) can found that activation of Broca’s area implied potentially enter into the product development cycle in two places. inNa thetur first,e Re viefMwsri can| Neur beoscienc used ase a Bayes factor of 2.3 for language involve- part of the design process itself. Here, neural responses could be used to refine the product before it ment, which means that taking brain activity is released. in the second, fMri can be used after the product is fully designed, typically to measure neural responses as part of an advertising campaign to increase sales. into account can make a small but significant improvement to one’s prior estimate of whether a language process was involved. Many studies have shown that striatal correlations between nAc activity and others? Because the field of neuroeconomics activity correlates with hedonic rating product preferences in college students28. grew out of early brain-imaging studies of scales26. neuromarketers have been quick However, in logistic regression (r2) calcula- the neurobiology of reward31,32, most of the to invert this finding and use ventral striatal tions aimed at predicting consumer choice, neuroeconomic data are about valuation activity as an indication that an individual self-reported preferences outperformed mechanisms and the associated responses likes something; but what is the evidence for brain activation alone. Adding brain activa- of dopamine-rich brain regions. The OFc this? using Poldrack’s method to analyse the tion to a logistic model improved predic- and striatum have been consistently impli- BrainMap database, we estimated the pos- tions, but only slightly (increasing r2 from cated in goal-directed action9,22,33–35. It is also terior probability for a reward process given 0.528 to 0.533). re-analysis with more generally accepted that the insula has a key the observation of nucleus accumbens (nAc) sophisticated machine-learning algorithms role in physiological arousal, which is typi- activation27. The prior probability of engag- further improved the predictive value of cally, although not exclusively, aversive in ing a reward-related process was assumed to brain activation29. nature21. But because of the reverse inference be 0.5 (1:1 odds). According to this estima- Although some have argued for the exist- problem, using striatal and OFc activity tion, based on the number of fMrI papers ence of a “buy button” in the brain5, current as a read-out of ‘liking’ and the insula as a reported in the BrainMap database with evidence suggests that the cognitive proc- ‘disgust-meter’ is probably too simplistic to and without ‘reward’, and with and without esses associated with purchase decisions be of use in a real-life setting. In the context nAc activation, nAc activation increases are multi factorial and cannot be reduced of neuromarketing, the statistical power of the probability of a reward-related process to a single area of activation. conversely, a these single-region correlations may be too taking place to 0.90 (odds 9:1). This yields given brain region may be involved in mul- low for the correlations to be of use as pre- a Bayes factor of 9, which is considered tiple cognitive processes. A recent review dictors of consumption unless, perhaps, the moderate to strong evidence for a causal of value-based decision making divided the neuroimaging data is combined with other relationship (BOX 1). Although meaningful process of making a choice into five catego- measures of preference. in a statistical sense, the assumptions behind ries: representation of the decision; assign- such a calculation are rather liberal and ment of value to different actions; action fMRI as a brain decoder. Given the limited may suffer from a publication bias for posi- selection; outcome evaluation; and learning30. power of reverse inference from single- tive results as well as differing definitions even within this simplified framework, cur- region brain activations, more data-driven of reward. In real-world settings, the ability rent data suggest that responses to market- methods for interpreting brain imaging to infer whether an individual likes some- ing efforts and consumer choices depend on data have been at the forefront of analysis thing based on nAc activation alone may be an array of neurobiological processes, and techniques. These techniques treat sites of substantially less. that no single brain region is responsible brain activity agnostically — that is, without In the context of a product likeability for a consumer choice. But is it possible that reference to prior hypotheses. The primary experiment, Knutson et al. found significant some brain regions are more involved than assumption is that, regardless of how an

Box 1 | NAc activation in studies of tasks with and without reward representations of the product that may lead to future consumption. Only certain types The BrainMap database was searched for functional MRI studies with and without a reward task of products can be consumed in an MrI and with and without nucleus accumbens (NAc) activation. The NAc was defined as a bilateral scanner. Therefore, much of the post-design region of interest with vertices from MNI (Montreal Neurological Institute) coordinates (–12, 0, –12) neuromarketing literature has focused on to (12, 12, 0). The frequencies that were obtained are shown in the table below. brain responses to visual representations of Assuming that the prior probability of engaging in a reward-related process is 0.5, calculations 28,45 showed that NAc activation increases the probability of a reward-related process taking place to products, such as pictures or advertise- 46–48 0.90, yielding a Bayes factor of 9: ments for the product ; however, these advertisement studies, which used magneto- Probability of NAc activation given a reward task = 27/68 = 0.397 encephalography and electroencephal- Probability of NAc activation given no reward task = 59/1283 = 0.046 ography (BOX 2), did not link imaging data to Assuming the prior probability of reward = 0.5, then actual purchase decisions or other ratings, so it is not yet possible to determine the value Probability of a reward task given NAc activation = of this approach.

reward task No reward task The role of expectations. It has long been known that the manner in which choices NAc activated 27 59 are presented can have a dramatic effect on NAc not activated 41 1,224 decisions49. This is where advertisements and product placement come into play. To date, experiments have examined fairly simple choices and responses to things that individual’s brain represents information, genetic polymorphism associated with a par- can be presented in an MrI scanner. Before it does so consistently. The representa- ticular cancer without understanding what neuroimaging can be used to predict con- tions may be spatially dispersed, and they that gene does — which is likely to yield sumer choice, a greater understanding of may be distributed differently in different specific but not general insights. the interplay between the decision maker, the individuals, but they can still be reliably elicitation method and the decision context detected through multi-voxel pattern analy- costs and benefits is needed. sis (MvPA). Because MvPA methods are As noted above, it is not yet clear whether BOLD responses are influenced by so- not reliant on the activation of a small subset neuroimaging provides better data than called ‘expectation’ effects, which include of brain regions, they have substantially other marketing methods (TABLE 1), but pricing effects, biases in the way the choice increased sensitivity to detect activation36. through the use of MvPA methods it might is presented50 and placebo responses. This A crucial advantage of MvPA techniques be possible to reveal the ‘holy grail’ of hid- suggests that neuromarketing could be over approaches in which activation in a den information. Assuming that this is the helpful in identifying individual differences particular brain region of interest is meas- case, will using expensive neuroimaging in consumer reactions to different types of ured is that MvPA has the statistical power ultimately be more efficient than using inputs. In a study of neural responses to sips to predict the individual choices of a subject. cheaper methods? Typical charges for scan- of wine, medial OFc responses were higher Because MvPA involves statistical associa- ning in a university research setting average when subjects were told that the wine was tions of complex activation patterns that about uS$500 per hour. In a commercial expensive ($90 per bottle) versus inexpen- occur when an individual choice is being setting, these will be higher. However, sive ($5 per bottle)23. Activity in this region made, it does not depend on the vagaries actual scan charges account for a small por- also correlated with self-report ratings of of an experimenter interpreting the mean- tion of the total cost, with personnel and how much participants liked the wine, even ing of an activation map. Some of the most overhead expenses accounting for at least though all wines were actually the same. impressive demonstrations of MvPA have 75% of the costs of an imaging project. If These results suggest that the instantaneous been in decoding visual responses to simple neuromarketing is to compete with conven- experience of pleasure from a product — stimuli37–39 and subsequently, to watching tional marketing approaches on the basis that is, experienced utility — is influenced films40, the meanings of nouns41, event of efficiency, then the costs of labour and by pricing, and that this effect may be medi- boundaries of written narratives42 and city overheads will have to be reduced. ated by the medial OFc9. This result paral- navigation43,44. One area in which the cost of neuroim- lels a similar, behavioural finding that the It is possible, even likely, that such meth- aging can be compared with conventional strength of the placebo effect for analgesia is ods will soon be able to handle almost any marketing approaches is in the post-design greater for more expensive ‘medications’51. circumstance that can be created in an MrI phase, the goal of which is to increase sales Subjects’ expectations also play an impor- environment. with increasing stimulus com- of an existing product — for example, tant part in how the experimenter should plexity, simple interpretations of brain acti- through advertisements and other types interpret striatal responses. Many studies vation will become more difficult. However, of framing effects. early neuromarketing have shown that the reward-related signals for real-world marketing applications, it studies therefore used imaging approaches in the ventral striatum and nAc can be may be more important to predict future to evaluate consumer responses to adver- more accurately linked to prediction errors behaviour than to understand the ‘why’ tisements. At this point, it is important to for reward than to reward itself 22,52,53. of behaviour. Such a data-driven application distinguish between neural responses to the Placebo responses are an interesting of imaging (perhaps even lacking an under- consumption of a product (that is, expe- aspect of neuromarketing. The mecha- lying theory) is analogous to identifying a rienced utility) and neural responses to nism of the placebo response has been

Box 2 | Neuromarketing technologies but do not directly suggest that maximizing activity in a particular brain region results Functional Mri (fMri) in more sales. The technique uses an MRI scanner to measure the blood oxygenation level-dependent (BOLD) signal. The BOLD changes are generally correlated with the underlying synaptic activity. Spatial resolution is 1–10 mm, and temporal resolution is 1–10 s. In general, the higher the spatial Culture and advertising. neuroimaging resolution, the lower the temporal resolution. Of the three imaging technologies described in this is often hyped as an exciting new tool Box, fMRI has a substantial advantage in resolving small structures and those that are deep in the for advertisers. Despite its enormous brain. However, some important brain regions, especially the orbitofrontal cortex, are affected by cost, advertising effectiveness is a poorly signal artefacts that may reduce the ability to obtain useful information. State of the art MRI understood area of marketing. Although scanners cost approximately US$1 million per Tesla and have annual operating costs of advertising has been investigated in a few $100,000–$300,000. neuroimaging studies57,58, it is still unknown electroencephalography (eeG) whether neuroimaging can prospectively EEG uses electrodes applied to the scalp and measures changes in the electrical field in the brain reveal whether an advertisement will be region underneath. EEG has very high temporal resolution (milliseconds) and can therefore detect effective. In a famous coke–Pepsi study, par- brief neuronal events. Because the skull disperses the electrical field, EEG has low spatial resolution ticipants who described themselves as coke (~1 cm) that depends on how many electrodes are used. The number of electrodes can be as few as drinkers showed significant activation in the two or range up to hundreds in high-density arrays. The greater the number of electrodes, the hippocampus and right dorsolateral PFc better the spatial resolution. Apart from the low spatial resolution, EEG has poor sensitivity for deep brain structures. Equipment costs can be low (<$10,000) but increase with high-density arrays and when they were cued about the upcoming 45 the concomitant resources needed to process the data. A common technique is to measure the drink of coke . Self-described Pepsi drink- left–right asymmetry of the frontal EEG78. This is typically measured by the power in the alpha band ers did not have this response. In the absence (8–13 Hz). This research has suggested that relatively greater activity in the left frontal region is of brand information, there was no signifi- associated with either positive emotional states or the motivational drive to approach an object79. cant difference in preference during a taste Although there are strong correlations between frontal EEG asymmetry and personality traits, the test. The study suggested that any differences degree to which the asymmetry changes from moment to moment is still debated. Some have in the response (behavioural and neural) to 80 suggested a minimum of 60 s to reliably estimate power asymmetry , in which case the temporal the two brands must be culturally derived. advantage of EEG over fMRI is lost. Although some have used this approach to measure momentary One possibility is that brands achieve a life fluctuations in emotion in response to advertisements81, without accounting for autocorrelations in of their own by becoming animate objects, time or multiple statistical comparisons, the validity of such approaches is dubious. sometimes with human attributes, in the Magnetoencephalography (MeG) minds of consumers. However, one fMrI An expensive cousin of EEG, MEG measures changes in the magnetic fields induced by neuronal study that compared brain responses to activity. Thus, MEG has the same advantage of high temporal resolution and, because the magnetic field is less distorted by the skull than is the electrical field, it has better spatial resolution than EEG. persons and brands found that activation Like EEG, MEG is most sensitive to superficial cortical signals (primarily in the sulci). MEG requires a patterns for brands differed from those magnetically shielded room and superconducting quantum interference detectors to measure the for people — even for brands with which weak magnetic signals in the brain. An MEG set-up costs approximately $2 million. subjects are identified — suggesting that Transcranial magnetic stimulation (TMs) brands are not perceived in the same way 59 TMS uses an iron core, often in the shape of a toroid wrapped in electrical wire, to create a as people . Another possibility is that spe- magnetic field strong enough to induce electrical currents in underlying neurons when placed on cific emotions can be elicited in response the head82. TMS can be used as a single pulse, paired pulse or repetitive stimulation, and the to advertisements, although whether neuro- neuronal effects range from facilitation to inhibition of synaptic transmission. As a research tool, imaging will help to reveal these emo- TMS has been used to study the causal role of specific brain regions in particular tasks by tions may ultimately be limited by reverse temporarily taking them ‘offline’. inference constraints, especially if tied to specific regions. The issue of how culturally derived iden- debated for decades54, but ultimately it can The aforementioned manipulations of tities become embedded in the brain is of be considered an effect of marketing (that expectations are simple and direct. For great interest, not only from a marketing is, the actions of a doctor, pharmaceutical example, the experimenter can manipulate perspective. Although neoclassical economic company or experimenter). The neural a single dimension of expectation, such as theory describes a framework in which indi- correlates of the analgesic placebo effect price or descriptive words (for example, viduals assess costs and benefits during their are widespread but generally point to a “ultra” and “new and improved”), and decision-making processes, it is clear that modulation of the cortical pain matrix measure the effect on the consumer behav- people base many decisions on sociocultural in the brain55,56. Because consumers can- iourally and neurally. More cognitively com- rules and identities. Some are in a commer- not consciously report placebo effects, plex forms of expectations can be created cial context (for example, “I am a Pc” or “I the demonstration of neural correlates of through advertisements and commercials. am a Mac”) but many are not (for example, these effects suggests that having access to Post-design applications of neuroimaging “I am a Democrat” or “I am a republican”). hidden brain information could enable a have, for the most part, confirmed what These issues extend beyond the mundane marketer to measure the effectiveness of was known about the behavioural effects of questions of advertisement effectiveness and a placebo marketing strategy in a particular product placement, which bypass the coun- raise the more profound question of how the individual. How well this type of informa- ter-arguments in which people naturally marketing of ideas affects decision making. tion generalizes to a larger population will engage when facing advertisements. The But whether neuroimaging provides an determine the cost–benefit ratio of doing imaging studies confirm that there are neu- efficient tool to answer this question has yet neuroimaging. ral correlates of exposure to advertisements to be shown.

Box 3 | The ethics of neuromarketing of flavour is a multisensory integration process and thus provides several oppor- The introduction of neuroimaging into an environment in which the ultimate goal is to sell more tunities for neuroimaging to disentangle a product to the consumer may raise ethical issues. complex perception that subjects might not • Businesses will be able to read the minds of consumers. This concern is about the privacy of be able to articulate; taste, odour, texture, thoughts. Can neuroimaging be used to gauge a person’s preferences outside of the specific task appearance and even sound all contribute being performed? Possibly. This concern may be mitigated through transparency of purpose: to the gustatory experience. These differ- subjects must know what kind of endeavour they are helping, and their data should only be used ent dimensions have been mapped onto for that purpose. distinct brain regions but with substantial • Private versus public information about preferences. Individuals need to be able to exercise control overlap60,61. The OFc is consistently linked over what they choose to reveal about their personal preferences. A privacy breach occurs if to perceived pleasantness, whereas viscos- neuroimaging reveals a private preference that is outside the scope of the neuromarketer’s research question. ity and fat content seem to be represented in the insula62. The use of neuroimaging • Information will be used to discriminate against individuals or exploit particular neurological traits found in a subgroup of individuals. Neuroimaging data could potentially target marketing to by commercial manufacturers to design a specific people or groups. Many people would find this tactic repugnant because it exploits a more appealing food product is both feasible biological ‘weakness’ that only exists in some people. Similarly, this information could be used and likely. For this to work, however, one to time pricing moves to capitalize on individual weaknesses that are known to coincide with would need to identify which dimension of particular biological states (for example, raising beverage prices when someone is known to gustation is to be studied (for example, taste, be thirsty). odour or texture) and maximize a brain • Central versus peripheral routes of influence. A central route aims to influence consumers’ response to variations in that dimension. preferences about the functional aspects of the product (for example, fewer calories in a beer). The drawback to such an approach is the A peripheral route attempts to manipulate preferences through things that are peripherally related possibility of creating food products that are to the product (for example, sex appeal of people in advertisements). Neuroimaging could so highly tuned to neural responses that potentially be used to enhance both types of influence, but some consider the attempts to individuals may over-eat and become obese optimize the peripheral route more ethically dubious. (see BOX 3 for a discussion of some ethical • Brain responses obtained from a small group of subjects will be used to generalize to a large issues related to neuromarketing). Is it pos- population. Of course, this is done all the time in the scientific literature. If neuromarketing data sible that such a neuroimaging approach are used in product design and the product injures someone, neuroimaging will be partly to blame. could create a ‘super-heroin of food’ — a 83 • Abnormal findings. Approximately 1% of the population will have an abnormality on their MRI . In product so delicious that all but the most a population without clinical symptoms, the clinical significance of an MRI abnormality is unknown. ascetic individuals would find it irresistible? Many will be false positives; others will be real and require referral. Currently, there is no standard for how to handle these situations. However, it is standard practice to have a written policy in It is an extreme but real possibility. However, place for abnormal findings. Failure to do so opens both the neuromarketing firm and their clients that does not mean that neuroimaging is to medical liability. necessarily problematic for food product • A lack of regulation. Traditional marketing methods, because they are not typically viewed as development. Indeed, the same techniques experimentation, have not been subject to institutional review board (IRB) oversight. MRI scans are could be applied to making nutritious foods approved by the US Food and Drug Administration (FDA) for clinical use but, because no diagnosis more appealing. is being made in a marketing setting, there is the potential to circumvent both FDA and IRB requirements. The burgeoning neuromarketing industry would be well advised to adopt an Entertainment. As a typical big-budget industry standard of independent review. Clients should demand it. Hollywood film costs over $100 million, • Management of perceptions. How will the public react when they discover that neuroimaging has with almost as much spent on marketing, it been used to design or market a product? The public’s response to genetically modified food could would be surprising if film producers were provide an indication. not interested in using neuroimaging to • Companies might not be primarily concerned with the best interests of the consumer. Companies improve their product. After static images, and consumers maintain complex relationships in which some of their goals are compatible while films are probably the easiest product to others are in conflict. On the one hand, companies seek to design, manufacture and sell products present in the scanner. Moreover, an fMrI that consumers seek to buy, resulting in compatible goals that benefit both parties. On the other measurement is time locked to the film hand, companies also aim to maximize their short- or long-term profits, sometimes to the timeline. A film presents the same basic vis- detriment of their consumers. Much like marketing itself, understanding consumer preferences ual and auditory stimuli to everyone viewing can be used for goals that are in the best interests of both the company and their consumers or for objectives that are in the interests of the company and to the detriment of their consumers. Which it and thus should serve as a cognitive syn- approaches neuromarketers choose is an open question. chronizer. Indeed, an fMrI study of subjects viewing a segment of the classic western The Good, the Bad and the Ugly 40 showed early product design Food products. various food products and that large extents of the cortex responded As the ability of neuroimaging to predict or beverages have been administered in the similarly in time across subjects, suggesting influence post-design purchase decisions MrI scanner, from simple sugar solutions that much of the cortical response is essen- seems to be limited (see above), neuro- to chocolate, wine, sports drinks and colas. tially stereotypical. In another study, the imaging may be better suited to gauging Beverages are particularly easy to administer, ability to recall narrative content of the Tv responses before products are marketed. with the usual route through a computer- sitcom Curb Your Enthusiasm three weeks The primary reason is that neuroimag- controlled pump attached to a tube that later was correlated with the strength of ing may yield insights into the product delivers controlled amounts of fluid into hippocampal and temporal lobe responses experience itself. the participant’s mouth. The perception during viewing63.

Box 4 | What to look for when hiring a neuromarketing firm associated with seeing specific aspects of a building; and second, the use of neural We provide a list, which is by no means exhaustive, of what could be considered standard responses to guide the architectural design practice in the application of neuroimaging methods in cognitive neuroscience and related process. clearly, one would need to identify fields. It is based on standard criteria for reviewing research proposals and adapted to a these neural responses before attempting business setting. to use them in architectural design, but it is • What is to be gained from neuroimaging? Good neuromarketers will begin by discussing the pros precisely the application in design that places and cons of the proposal in detail. For example: what will neuroimaging yield over traditional methods? Ask for data about the predictive value of neuroimaging findings in a real-world setting. neuroimaging within the neuromarketing framework. • What are the dependent and independent measurements? Assessing brain activation is not virtual reality can provide a surprisingly generally useful without correlating it with some other measurement. It is necessary to have another behavioural measurement to anchor the interpretation of the brain activation. Be wary if accurate simulation of an architectural experi- someone claims to know what a person thinks based solely on brain activation. ence and can be used in an MrI scanner. It has already been used to understand neural • How many subjects are needed? Apart from the simplest of tasks, any task invoking a response 65,66 that is expected to vary across individuals demands a sample size of at least 30 (REF. 84). If groups activation during automobile driving . of individuals are being compared under different treatments or conditions, the sample size will In spatial navigation tasks such as driving, need to be much greater to detect differences between groups and between different treatments. and presumably navigating a building, the • What is the nature of the stimuli? Simple stimuli are the easiest to analyse. Real-world images, as hippo campus has a key role. These early might appear in an advertisement, become difficult to characterize unless one element at a time is virtual reality experiments suggested that varied. For statistical power, a minimum of 10 repetitions within a stimulus category are required, the hippocampus is active when the subject although 20–30 would be more likely to achieve meaningful results. makes navigation decisions but not when 65 • What type of software will be used to analyse the neuroimaging data? Several software packages they are externally cued . Perhaps taking into exist, and although these programmes make neuroimaging seem simple, it takes a minimum of account ‘hippocampal load’ may be a useful 1 year of training to be able to use them and 3 years to become fully competent. tool in architectural design — for example, to • How will motion correction be performed? make buildings easier to navigate. extending this idea by considering the neurobiological • Are conditions balanced in time? If not, how will subjects’ drifting attention be compensated for? changes associated with ageing, it might be • Is this a whole-brain analysis or is a specific part of the brain being examined? These necessitate possible to design buildings and retirement different thresholds of identifying activation. The chance of an activation appearing somewhere in communities that mitigate the memory loss the brain is high due to random noise. associated with Alzheimer’s disease. • Will regions of interest be defined a priori? If so, what is the justification for this? Conclusions based on activation of a single region will have relatively little predictive power over conventional Political candidates. Finally, neuromarketing behavioural methods. might be applied to perhaps the greatest mar- • If multi-voxel pattern analysis (MVPA) methods will be used, will they be completely data-driven keting campaign of all: politics. According (principal component analysis or independent component analyses) or will they be based on classifier training of subject responses (support vector machine, relevance vector regression or to the Federal election commission (see Gaussian process regression)? How will the resulting activity maps be interpreted? Further information), the cost of the 2008 uS Presidential race was approximately $1.6 • How robust are the results? Ask for a ‘bootstrap’ — for example, testing on a ‘fresh’ subsample of data. billion. It was also around that time that neuroimaging made its way into politics, perhaps • What type of scanner will be used? Either 1.5 or 3 Tesla scanners can yield images of acceptable most prominently in the form of a New York quality. Open MRIs do not have the field homogeneity or the gradient technology necessary for 67 fMRI. What quality control checks are performed to make sure the scanner is operating optimally Times op-ed piece . Peer-reviewed studies and consistently from day to day? What steps will be taken to minimize signal artefacts in areas have shown a complex pattern of activation with poor signal? in response to statements about candidates; these patterns have been interpreted as evidence that motivated reasoning involves activation in the ventromedial PFc, the anterior Such stereotypical responses suggest that industry but, as most of this work is unpub- cingulate cortex, the posterior cingulate cor- fMrI could be used during the editing proc- lished, it is difficult to evaluate the quality of tex and the insula68. Subsequent studies have ess. For example, different cuts of a movie the product. However, guidelines for general suggested that activation of the medial PFc could be measured against these cortical quality of scientific work can be formu- might be associated with maintaining a responses, which could then be used to select lated based on two decades of neuroscience subject’s preference for a candidate in the final cut for release. Although it seems research. Thus, without passing judgment on response to advertisements, whereas activity hopelessly complex to interpret such brain whether neuromarketing works, we can at in the lateral PFc might be associated with responses, it may not be necessary if the only least identify the items to look for in a quality changing candidates69. goal is to release the most profitable movie. operation (BOX 4). In marketing terms, the political can- Provided there were a metric of quality (for didates are the products that must be sold example, box office returns or test audience Architecture. A growing number of neuro- to the electorate. Therefore, like other reports), brain activation patterns could scientists and architects have begun to products, candidates and their campaigns be chosen to optimize outcomes without consider the relationships of the brain to the have pre- and post-design phases. Political any knowledge of what the patterns meant. architectural experience64. The neuroscience marketing is aimed at selling an existing Several neuromarketing companies have tar- of architecture could be considered from candidate but, with more foresight, can also geted their efforts towards the entertainment two perspectives: first, the neural activity be used to ‘design’ a better candidate. The

aforementioned neuroimaging studies have hype. However, continuing developments in Dan Ariely is at the Fuqua School of Business, focused on the post-design responses to analytical tools for neuroimaging data — for Center for Cognitive Neuroscience, Department of Economics, and the Department of Psychiatry and 68,69 advertisements for political candidates . example, MvPA — suggest that neuro- Behavioural Sciences, Duke University, Durham, could neuroimaging also be used to imaging will soon be able to reveal hidden North Carolina 2770, USA. design a candidate? Although potential information about consumer preferences. Gregory S. Berns is at the Department of Psychiatry nominees already go through a ‘grooming’ Although this information could boost post- and Behavioural Sciences, Economics Department, process, it is worth examining this prospect. design sales efforts, we think that the real Center for Neuropolicy, Emory University, Atlanta, A candidate’s appearance, trustworthiness pay-off will come during the design proc- Georgia 30322, USA. and message content might determine a ess. using fMrI data during design could Correspondence to G.S.B. voter’s decision. considerable neuroimag- affect a wide range of products, including e-mail: [email protected] ing work has been done on the perception food, entertainment, buildings and political doi:10.1038/nrn2795 of human faces70 and features such as facial candidates. Published online 3 March 2010 symmetry, skin colour and attractiveness. There are two sides to the use of such 1. Beckwith, N. E. & Lehmann, D. R. The importance of halo effects in multi-attribute attitude models. Key brain structures in visual processing information. Product manufacturers could J. Mark. Res. 12, 265–275 (1975). include the fusiform face area for basic use neural information to coerce the public 2. Day, G. S. The threats to marketing research. J. Mark. Res. 12, 462–467 (1975). 71 face processing , the superior temporal into consuming products that they neither 3. Griffin, A. & Hauser, J. R. The voice of the customer. sulcus for gaze direction and intention need nor want. However, we hope that Mark. Sci. 12, 1–27 (1993). 4. 12 Green, P. E. & Srinivasan, V. Conjoint analysis in and the nAc for attractiveness . A recent future uses of neuromarketing will help marketing: new developments with implications for study on the effect of political candidates’ companies to identify new and exciting research and practice. J. Mark. 54, 3–19 (1990). 5. Lindstrom, M. Buyology. Truth and Lies About Why appearance found that insula activation in products that people want and find useful. We Buy (Doubleday, New York, 2008). response to seeing a picture of a candidate One example is a new trend in ‘user design’ 6. Hauser, J. R. & Shugan, S. M. Intensity measures of consumer preference. Oper. Res. 28, 278–320 was associated with a greater likelihood in which companies allow consumers to (1980). of that candidate losing the election72. In participate, through the internet, in the 7. Buchanan, B. & Henderson, P. W. Assessing the bias of preference, detection, and identification measures addition, dorsolateral PFc and anterior design of new products and by doing so of discrimination ability in product design. Mark. Sci. cingulate cortex activation occurred when create products that are more useful for the 11, 64–75 (1992). 8. Kahneman, D., Wakker, P. P. & Sarin, R. Back to subjects viewed images of a candidate companies and for their customers. Perhaps Bentham? Explorations of experienced utility. of a political party different from their a next phase in user design is one that incor- Q. J. Econ. 112, 375–405 (1997). 73 9. Plassmann, H., O’Doherty, J. & Rangel, A. own . The neurobiology of trust has also porates not only what consumers express, Orbitofrontal cortex encodes willingness to pay in become quite popular to study with both but also what they think. everyday economic transactions. J. Neurosci. 27, 9984–9988 (2007). fMrI and, more recently, pharmacologi- Finally, we return to the opening ques- 10. Hare, T. A., Camerer, C. F. & Rangel, A. Self-control cal manipulations74–76. These studies have tion: hope or hype? It is too early to tell but, in decision-making involves modulation of the vmPFC valuation system. Science 324, 646–648 (2009). found that different dimensions of trust, optimists as we are, we think that there is 11. O’Doherty, J. P., Deichmann, R., Critchley, H. D. & such as reputation, fairness and uncertainty, much that neuromarketing can contrib- Dolan, R. J. Neural responses during anticipation of a primary taste reward. Neuron 33, 815–826 (2002). correlate with activity in different brain ute to the interface between people and 12. Aharon, I. et al. Beautiful faces have variable reward regions. Moreover, the hormone oxytocin businesses and in doing so foster a more value: fMRI and behavioral evidence. Neuron 32, 537–551 (2001). affects human behaviour in various eco- human-compatible design of the prod- 13. Zatorre, R. J., Chen, J. L. & Penhume, V. B. When nomic exchanges that depend on social ucts around us. At the same time, neuro- the brain plays music: auditory-motor interactions 77 in music perception and production. Nature Rev. interactions . Finally, a candidate’s message marketing as an enterprise runs the risk of Neurosci. 8, 547–558 (2007). content could be viewed as an experiential quickly becoming yesterday’s fad. Seasoned 14. Knutson, B., Adams, C. M., Fong, G. W. & Hommer, D. Anticipation of increasing monetary reward selectively product. One could theoretically attempt marketers still remember the hype around recruits nucleus accumbens. J. Neurosci. 21, RC159 to maximize striatal and OFc responses subliminal advertising, which quickly faded (2001). 15. O’Doherty, J., Kringelbach, M. L., Rolls, E. T., Hornak, J. to platform statements although, for the and died despite the research interest that & Andrews, C. Abstract reward and punishment reasons stated above, this is not necessarily surrounded it (and research on subliminal representations in the human orbitofrontal cortex. Nature Neurosci. 4, 95–102 (2001). predictive of success. priming remains a large part of academic 16. Izuma, K., Saito, D. N. & Sadato, N. Processing of research in social psychology). How can social and monetary rewards in the human striatum. Neuron 58, 284–294 (2008). conclusions and future directions we make sure that neuromarketing will not 17. Rilling, J. K. et al. A neural basis for social neuromarketing has received considerable suffer a similar fate? For one, the academic cooperation. Neuron 35, 1–20 (2002). 18. Ploghaus, A., Becerra, L., Borras, C. & Borsook, D. attention in both the scientific community community should take this topic seriously Neural circuitry underlying pain modulation: and the media. Although few scientific and not leave it to the neuromarketers and expectation, hypnosis, placebo. Trends Cogn. Sci. 7, 197–200 (2003). neuro marketing studies have been con- the op-ed page of the New York Times. 19. Ploghaus, A. et al. Dissociating pain from its ducted, the existing evidence suggests that we should also ask deeper questions on anticipation in the human brain. Science 284, 1979–1981 (1999). neuroimaging could be used advantageously how marketing works — and not simply 20. Koyama, T., McHaffie, J. G., Laurienti, P. J. & in several domains of marketing. For a examine whether type X of advertising Coghill, R. C. The subjective experience of pain: where expectations become reality. Proc. Natl Acad. Sci. USA marketer, neuroimaging could be attractive works better or worse than type Y. If we 102, 12950–12955 (2005). because it might be cheaper and faster than take neuromarketing as the examination of 21. Craig, A. D. How do you feel? Interoception: the sense of the physiological condition of the body. Nature Rev. current marketing tools, and because it could the neural activities that underlie the daily Neurosci. 3, 655–666 (2002). provide hidden information about products activities related to people, products and 22. Hare, T. A., O’Doherty, J., Camerer, C. F., Schultz, W. & Rangel, A. Dissociating the role of the orbitofrontal that would otherwise be unobtainable. we marketing, this could become a useful and cortex and the striatum in the computation of goal think it unlikely that neuroimaging will be interesting path for academic research values and prediction errors. J. Neurosci. 28, 5623–5630 (2008). more cost-effective than traditional market- and at the same time provide useful inputs 23. Plassmann, H., O’Doherty, J., Shiv, B. & Rangel, A. ing tools, and so the first point is mostly to marketers. Marketing actions can modulate neural

representations of experienced pleasantness. 46. Ambler, T., Ioannides, A. & Rose, S. Brands on the 69. Kato, J. et al. Neural correlates of attitude change Proc. Natl Acad. Sci. USA 105, 1050–1054 (2008). brain: neuro-images of advertising. Bus. Strategy Rev. following positive and negative advertisements. 24. Poldrack, R. A. Can. cognitive processes be inferred 11, 17–30 (2000). Front. Behav. Neurosci. 3, 6 (2009). from neuroimaging data? Trends Cogn. Sci. 10, 47. Rossiter, J. R., Silberstein, R. B., Harris, P. G. & Nield, G. 70. Tsao, D. Y. & Livingstone, M. S. Mechanisms of face 59–63 (2006). Brain-imaging detection of visual scene encoding in perception. Ann. Rev. Neurosci. 31, 411–437 (2008). 25. Poldrack, R. A. The role of fMRI in cognitive long-term memory for TV commercials. J. Advert. Res. 71. Kanwisher, N. & Yovel, G. The fusiform face area: neuroscience: where do we stand? Curr. Opin. 41, 13–22 (2001). a cortical region specialized for the perception of Neurobiol. 18, 223–227 (2008). 48. Astolfi, L. et al. Neural basis for brain responses to TV faces. Philos. Transact. Roy. Soc. B Biol. Sci. 361, 26. Delgado, M. R. Reward-related responses in the commercials: a high-resolution EEG study. IEEE Trans. 2109–2128 (2006). human striatum. Ann. NY Acad. Sci. 1104, 70–88 Neural Syst. Rehabil. Eng. 16, 522–531 (2008). 72. Spezio, M. L. et al. A neural basis for the effect of (2007). 49. Tverksy, A. & Kahneman, D. The framing of decisions candidate appearance on election outcomes. Soc. 27. Fox, P. T. & Lancaster, J. L. Mapping context and and the psychology of choice. Science 211, 453–458 Cogn. Affect. Neurosci. 3, 344–352 (2008). content: the BrainMap model. Nature Rev. Neurosci. (1981). 73. Kaplan, J. T., Freedman, J. & Iacoboni, M. Us versus 3, 319–321 (2002). 50. De Martino, B., Kumaran, D., Seymour, B. & them: political attitudes and party affiliation influence 28. Knutson, B., Rick, S., Wimmer, G. E., Prelec, D. & Dolan, R. J. Frames, biases, and rational decision- neural response to faces of presidential candidates. Loewenstein, G. Neural predictors of purchases. making in the human brain. Science 313, 684–687 Neuropsychologia 45, 55–64 (2007). Neuron 53, 147–156 (2007). (2006). 74. Fehr, E. & Camerer, C. Social neuroeconomics: 29. Grosenick, L., Greer, S. & Knutson, B. Interpretable 51. Waber, R. L., Shiv, B., Carmon, Z. & Ariely, D. the neural circuitry of social preferences. Trends classifiers for FMRI improve prediction of purchases. Commercial features of placebo and therapeutic Cogn. Sci. 11, 419–427 (2007). IEEE Trans. Neural Syst. Rehabil. Eng. 16, 539–548 efficacy. JAMA 299, 1016–1017 (2008). 75. Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher, U. (2008). 52. Montague, P. R. & Berns, G. S. Neural economics & Fehr, E. Oxytocin increases trust in humans. 30. Rangel, A., Camerer, C. & Montague, P. R. and the biological substrates of valuation. Neuron Nature 435, 673–676 (2005). A framework for studying the neurobiology of 36, 265–284 (2002). 76. King-Casas, B. et al. Getting to know you: reputation value-based decision making. Nature Rev. Neurosci. 53. Schultz, W., Dayan, P. & Montague, P. R. A neural and trust in a two-person economic exchange. Science 9, 545–556 (2008). substrate of prediction and reward. Science 275, 308, 78–83 (2005). 31. Camerer, C., Loewenstein, G. & Prelec, D. 1593–1599 (1997). 77. Baumgartner, T., Heinrichs, M., Vonlanthen, A., Neuroeconomics: how neuroscience can inform 54. Colloca, L. & Benedetti, F. Placebos and painkillers: Fischbacher, U. & Fehr, E. Oxytocin shapes the economics. J. Econ. Lit. 43, 9–64 (2005). is mind as real as matter? Nature Rev. Neurosci. 6, neural circuitry of trust and trust adaptation in 32. Glimcher, P. W. Decisions, decisions, decisions: 545–552 (2005). humans. Neuron 58, 639–650 (2008). choosing a biological science of choice. Neuron 36, 55. Wager, T. D. et al. Placebo-induced changes in fMRI in 78. Davidson, R. J., Ekman, P., Saron, C. D., Senulis, J. A. & 323–332 (2002). the anticipation and experience of pain. Science 303, Friesen, W. V. Approach-withdrawal and cerebral 33. Yin, H. H., Ostlund, S. B. & Balleine, B. W. Reward- 1162–1167 (2004). asymmetry: emotional expression and brain physiology I. guided learning beyond dopamine in the nucleus 56. Benedetti, F., Mayberg, H. S., Wager, T. D., Stohler, J. Pers. Soc. Psychol. 58, 330–341 (1990). accumbens: the integrative functions of cortico-basal C. S. & Zubieta, J. K. Neurobiological mechanisms of 79. Harmon-Jones, E. Clarifying the emotive ganglia networks. Eur. J. Neurosci. 28, 1437–1448 the placebo effect. J. Neurosci. 25, 10390–10402 functions of asymmetrical frontal cortical activity. (2008). (2005). Psychophysiology 40, 838–848 (2003). 34. Padoa-Schioppa, C. & Assad, J. A. Neurons in the 57. Kenning, P. H. & Plassmann, H. How neuroscience can 80. Huster, R. J., Stevens, S., Gerlach, A. L. & Rist, F. orbitofrontal cortex encode economic value. Nature inform consumer research. IEEE Trans. Neural Syst. A spectroanalytic approach to emotional 441, 223–226 (2006). Rehabil. Eng. 16, 532–538 (2008). responses evoked through picture presentation. 35. Schoenbaum, G. & Roesch, M. Orbitofrontal cortex, 58. Lee, N., Broderick, A. J. & Chamberlain, L. What is Int. J. Psychophysiol. 72, 212–216 (2008). associative learning, and expectancies. Neuron 47, neuromarketing? A discussion and agenda for future 81. Ohme, R., Reykowska, D., Wiener, D. & Choromanska, A. 633–636 (2005). research. Int. J. Psychophysiol. 63, 199–204 Analysis of neurophysiological reactions to advertising 36. Norman, K. A., Polyn, S. M., Detre, G. J. & (2007). stimuli by means of EEG and galvanic skin response Haxby, J. V. Beyond mind-reading: multi-voxel 59. Yoon, C., Gutchess, A. H., Feinberg, F. & Polk, T. A. measures. J. Neurosci. Psychol. Econ. 2, 21–31 (2009). pattern analysis of fMRI data. Trends Cogn. Sci. 10, A functional magnetic resonance imaging study of 82. Kobayashi, M. & Pascual-Leone, A. Transcranial 424–430 (2006). neural dissociations between brand and person magnetic stimulation in neurology. Lancet Neurol. 37. Haynes, J.D. & Rees, G. Decoding mental states from judgments. J. Consum. Res. 33, 31–40 (2006). 2, 145–156 (2003). activity in humans. Nature Rev. Neurosci. 7, 523–534 60. Small, D. M. & Prescott, J. Odor/taste integration 83. Illes, J. et al. Incidental findings in brain imaging (2006). and the perception of flavor. Exp. Brain Res. 166, research. Science 311, 783–784 (2006). 38. Kamitani, Y. & Tong, F. Decoding the visual and 345–357 (2005). 84. Thirion, B. et al. Analysis of a large fMRI cohort: subjective contents of the human brain. Nature 61. Rolls, E. T. Brain mechanisms underlying flavour and statistical and methodological issues for group Neurosci. 8, 679–685 (2005). appetite. Philos. Trans. R. Soc. Lond. B Biol. Sci. 361, analyses. NeuroImage 35, 105–120 (2007). 39. Kay, K. N., Naselaris, T., Prenger, R. J. & Gallant, J. L. 1123–1136 (2006). Identifying natural images from human brain activity. 62. De Araujo, I. E. T. & Rolls, E. T. The representation Acknowledgements Nature 452, 352–356 (2008). in the human brain of food texture and oral fat. This work was supported by grants to G.S.B. from the 40. Hasson, U., Nir, Y., Levy, I., Fuhrmann, G. & J. Neurosci. 24, 3086–3093 (2004). National Institute on Drug Abuse (R01DA016434 and Malach, R. Intersubject synchronization of 63. Hasson, U., Furman, O., Clark, D., Dudai, Y. & R01DA025045), the Office of Naval Research and Air Force cortical activity during natural vision. Science Davachi, L. Enhanced intersubject correlations Office of Scientific Research, and the National Science 303, 1634–1640 (2004). during movie viewing correlate with successful Foundation (BCS0827313). 41. Mitchell, T. M. et al. Predicting human brain activity episodic encoding. Neuron 57, 452–462 (2008). associated with the meanings of nouns. Science 64. Eberhard, J. P. Applying neuroscience to architecture. Competing interests statement 320, 1191–1195 (2008). Neuron 62, 753–756 (2009). The authors declare competing financial interests: see web 42. Speer, N. K., Zacks, J. M. & Reynolds, J. R. 65. Spiers, H. J. & Maguire, E. A. Neural substrates version for details. Human brain activity time-locked to narrative even of driving behaviour. NeuroImage 36, 245–255 boundaries. Psychol. Sci. 18, 449–455 (2007). (2007). FURtHeR inFoRMation 43. Spiers, H. J. & Maguire, E. A. Spontaneous 66. Calhoun, V. D. et al. Different activation dynamics in Dan Arieley’s homepage: mentalizing during an interactive real world task: multiple neural systems during simulated driving. http://www.predictablyirrational.com/ an fMRI study. Neuropsychologia 44, 1674–1682 Hum. Brain Map. 16, 158–167 (2002). Gregory s. Berns’s homepage: (2006). 67. Freedman, J. This is your brain on politics. New York http://www.neuropolicy.emory.edu/overview.html 44. Spiers, H. J. & Maguire, E. A. Decoding human brain Times (18 Jan 2005). Federal election commission: http://www.fec.gov/ activity during real-world experiences. Trends Cogn. 68. Westen, D., Blagov, P. S., Harenski, K., Kilts, C. & Sci. 11, 356–365 (2007). Hamann, S. Neural bases of motivated reasoning: sUPPLeMentaRy inFoRMation 45. McClure, S. M. et al. Neural correlates of behavioral an FMRI study of emotional constraints on partisan See online article: s1 (box) preference for culturally familiar drinks. Neuron 44, political judgment in the 2004 US Presidential All liNks Are AcTive iN The oNliNe PdF 379–387 (2004). election. J. Cogn. Neurosci. 18, 1947–1958 (2006).