Most behavioral-science findings in this hub did not survive scrutiny. Zajonc’s 1968 mere exposure effect did. Across five decades of replication, hundreds of studies, two large meta-analyses, and modern preregistered work, the basic claim that repeated exposure to a stimulus increases liking has held. Here is the honest case for one of cognitive psychology’s most durable findings.
If you have been reading through this hub, you have watched a long parade of canonical behavioral-science results get dismantled. Power posing did not survive Carney’s own recantation. Ego depletion collapsed under Hagger 2016. Money priming evaporated in preregistered replications. The marshmallow test shrank to something much smaller than its original claim once SES was controlled. Stereotype threat, the entire family of social-priming results, the facial-feedback hypothesis --- one after another, the most-cited demonstrations of “this is how the mind works” have either failed to replicate or eroded to a confidence interval that quietly straddles zero.
A rational reader could be forgiven for concluding that all of behavioral and cognitive psychology is suspect. That conclusion would be wrong, and the mere exposure effect is one of the cleanest counter-examples to it.
In the same five-decade window that produced all of those replication failures, the mere exposure effect --- the proposition that simply being exposed to a stimulus, even briefly and even without conscious recognition, increases the subsequent liking of that stimulus --- has held up. It has held up across Zajonc’s original 1968 monograph experiments with Chinese characters and Turkish nonsense words and yearbook faces. It has held up across Bornstein’s 1989 meta-analysis of 208 experiments. It has held up across the perceptual-fluency theoretical refinement of Bornstein and D’Agostino in 1992. It has held up across the 2017 Montoya re-examination of 256 effects across recognition, familiarity, and liking outcomes. And it continues to hold up in modern preregistered replications across paradigms ranging from product packaging to social-network friendship formation.
This is another anti-example article in a hub full of takedowns. It exists for the same three reasons the availability-heuristic and default-effect anti-examples exist. First, calibration --- readers should leave this hub knowing that “behavioral economics and cognitive psychology are mostly broken” is wrong; the more accurate claim is that the fields have produced a small number of robust, mechanism-grounded findings and a much larger number of fragile, contextually thin findings, and the failure mode has been treating those two categories as equivalent. Second, decision-usefulness --- for an executive evaluating which behavioral concepts to actually deploy in advertising, branding, product design, hiring, or market-entry strategy, the mere exposure effect is one of the highest-confidence inputs you have. And third, intellectual honesty --- a hub that catalogs the failures owes readers the parts that worked.
So here is the case for the mere exposure effect, with the legitimate critiques included.
What Zajonc 1968 Actually Demonstrated
The foundational paper is Zajonc, R. B. (1968). “Attitudinal Effects of Mere Exposure.” Journal of Personality and Social Psychology Monograph Supplement, 9(2, Pt. 2), 1—27. DOI: 10.1037/h0025848.
The structure of the paper is unusual in the same way the Tversky-Kahneman 1973 paper is unusual. It is not one experiment. It is a sequence of distinct studies, each using a different stimulus set, each testing a different facet of the proposed exposure-liking relationship, each yielding a result consistent with the same underlying mechanism. The cumulative weight of the studies is much harder to dismiss than any single demonstration. You would need to construct separate artifact explanations for each one, and no critic in the subsequent fifty-eight years has managed to do that across all of them.
The setup is straightforward. Zajonc proposed that the bare fact of repeated exposure to a stimulus, in the absence of any reinforcement or any explicit evaluative content, is sufficient to enhance affective evaluation of that stimulus. The exposure itself --- without learning, without conditioning, without conscious cognitive elaboration --- moves liking. This was a strong claim against the dominant learning-theory framework of the 1960s, which assumed that attitude change required some form of reinforcement or cognitive processing of stimulus-meaning.
The studies tested specific implications of this proposal. In one, undergraduate subjects were shown unfamiliar Turkish-sounding nonsense words --- “iktitaf,” “biwojni,” “afworbu,” and others --- at varying exposure frequencies, from a single presentation to twenty-five presentations. After the exposure phase, subjects were told the words were Turkish adjectives and asked to rate whether each one referred to something good or bad. The result was systematic: subjects rated the more-frequently-exposed words as referring to more-favorable concepts. The exposure had moved evaluative judgment about words whose meaning the subjects did not actually know.
In another study, Zajonc used Chinese characters as the stimulus set. Subjects had no prior knowledge of Chinese and could not read the characters. Characters were presented at varying frequencies during an exposure phase. Subjects were then asked to estimate whether each character represented a good or a bad concept. As with the Turkish-adjective study, frequently-exposed characters were judged to represent more-favorable concepts than rarely-exposed characters. Exposure shifted affect even for completely opaque stimuli.
In a third study, Zajonc used yearbook photographs of unfamiliar faces. Subjects were shown the faces at varying frequencies. Subjects were then asked to rate how much they liked each person depicted. Again, the more frequently a face had been shown, the more the subject reported liking the person. The effect held across stimulus categories --- abstract symbols, novel words, human faces --- and it held with no reinforcement, no meaningful information, and no explicit affective content in the exposure phase itself.
The studies, taken together, established something more durable than any single demonstration could have. They established a regularity of cognition: repeated exposure to a stimulus, by itself, enhances affective response to that stimulus. The regularity held across enough different stimulus types and task structures that it could not be dismissed as a quirk of one paradigm.
The Zajonc paper landed in a strange spot in the 1968 literature. It was published as a monograph supplement rather than as a regular journal article, which limited initial citation. But the demonstrations were clean, the theoretical claim was sharp, and the implications for advertising, branding, political campaigning, and interpersonal attraction were obvious enough that the paper accumulated citations rapidly through the 1970s and 1980s. By the time the first comprehensive meta-analysis appeared in 1989, the mere exposure effect had been replicated dozens of times across dozens of laboratories.
Bornstein 1989 --- The Meta-Analysis That Established Robustness
The question of whether the mere exposure effect would survive systematic meta-analytic scrutiny was settled by Bornstein, R. F. (1989). “Exposure and Affect: Overview and Meta-Analysis of Research, 1968-1987.” Psychological Bulletin, 106(2), 265—289. DOI: 10.1037/0033-2909.106.2.265.
Bornstein collected the published exposure-affect literature from 1968 through 1987 --- twenty years of research --- and performed a quantitative meta-analysis of 208 separate experiments. This is the kind of meta-analysis that would be considered methodologically strong even by modern post-replication-crisis standards: it included a comprehensive literature search across multiple databases, it coded each study for stimulus type and exposure parameters and moderator variables, it computed effect sizes in a common metric, and it explicitly tested for moderating variables rather than just reporting an average effect.
The main result was unambiguous. The overall effect size of mere exposure on affective response was substantial and statistically robust. Across the 208 experiments, the relationship between exposure frequency and stimulus liking was reliably positive, and the magnitude of the effect was large enough to have practical consequence rather than being a statistically-detectable but practically-meaningless perturbation.
But the more interesting contribution of the Bornstein meta-analysis was the moderator-variable analysis, because it sharpened the conditions under which the effect was largest, and it produced several non-obvious empirical findings that have shaped the subsequent literature.
The first moderator finding: the mere exposure effect was larger for brief or subliminal exposures than for prolonged, fully-attentive exposures. This was surprising to the conscious-cognitive-elaboration intuition --- if you assumed liking grew through some form of cognitive engagement with the stimulus, you would predict the opposite. But the finding held across the meta-analytic dataset. Brief, sub-threshold exposures that the subject could not consciously recognize were producing larger affect shifts than long, attentive exposures. This pattern strongly implicated a non-conscious mechanism and ruled out the simple “you think about the thing more, you like it more” explanation.
The second moderator finding: the effect was larger for novel stimuli than for already-familiar stimuli. If a subject already had pre-experimental exposure to the stimulus category --- common English words, for instance, rather than novel Turkish nonsense words --- the marginal effect of additional exposure was smaller. The mechanism appeared to be operating most powerfully in the zero-to-modest-familiarity range, with diminishing returns as familiarity accumulated. This pattern matched a perceptual-processing account: the most pronounced changes in processing fluency happen during the early exposures, when the stimulus is moving from genuinely novel to recognizable.
The third moderator finding: the effect was larger when the exposure stimuli were interspersed with other stimuli rather than presented as a long uninterrupted block of the target stimulus. Twenty exposures to a target stimulus spread across a longer session of mixed-stimulus presentation produced larger affect shifts than twenty consecutive exposures to the target stimulus in isolation. This pattern was consistent with several mechanism accounts, including a habituation-avoidance explanation (long blocks lead to satiation, which dampens or reverses the effect) and a discriminative-attention explanation (interspersed exposures are more informative for perceptual learning).
The fourth moderator finding: the effect was largest in laboratory contexts where the subject had no awareness of the manipulation, and it attenuated as subjects became aware of the exposure manipulation. This is consistent with the broader pattern that mere exposure operates pre-consciously rather than through deliberate evaluation, and it has direct practical implications for the design of exposure-based interventions in advertising and branding contexts.
The Bornstein meta-analysis did something rare in the behavioral-science literature of its era: it took a finding that had been replicated frequently but inconsistently, organized the empirical heterogeneity into mechanism-informative moderators, and produced a synthesis that both confirmed the basic effect and sharpened the theoretical understanding of when and why it was largest. The paper is one of the more sophisticated meta-analyses in the cognitive-affective psychology literature, and the moderator findings have all been replicated in subsequent work.
Montoya 2017 --- The Re-Examination That Updated The Effect Profile
By the mid-2010s, the methodological standards of the field had shifted substantially. The original Zajonc and Bornstein work pre-dated the replication-crisis-era emphasis on publication bias, preregistration, and statistical-power analysis. A natural question was whether the mere exposure effect, when re-examined with modern meta-analytic methods, would retain its robust profile or whether it would shrink the way many other classic findings had shrunk under similar scrutiny.
That re-examination is Montoya, R. M., Horton, R. S., Vevea, J. L., Citkowicz, M., & Lauber, E. A. (2017). “A Re-Examination of the Mere Exposure Effect: The Influence of Repeated Exposure on Recognition, Familiarity, and Liking.” Psychological Bulletin, 143(5), 459—498. DOI: 10.1037/bul0000085.
The Montoya team collected the exposure-affect literature through 2016, more than tripling the time window of the Bornstein meta-analysis. They organized the analysis around three outcome variables, not just liking: recognition (does the subject explicitly recognize the previously-exposed stimulus), familiarity (does the subject report a sense of familiarity with the stimulus), and liking (does the subject report enhanced affective evaluation). This decomposition was theoretically important because it allowed empirical separation of the perceptual-processing path from the affective path, and it allowed testing of whether the exposure-liking relationship operated independently of the exposure-recognition relationship.
The headline result was a confirmation: the mere exposure effect on liking remained robust across the expanded dataset. The relationship between repeated exposure and enhanced affective evaluation was positive and reliably non-zero across 256 separate effect sizes from the literature. This was true after correcting for publication bias using modern selection-model methods, which is a meaningful test because many classic findings shrink substantially or disappear when publication-bias corrections are applied.
But the more interesting Montoya findings concerned the moderator structure. The team confirmed most of the Bornstein moderators --- the effect was larger for novel stimuli, larger for brief presentations, larger when exposures were interspersed --- and added several new moderator findings that further sharpened the empirical profile.
First, the exposure-liking effect was meaningfully attenuated, but not eliminated, when the meta-analysis was restricted to studies with the strongest methodological safeguards (preregistration, large samples, well-controlled stimulus randomization). The robust core of the effect was real, but the effect-size estimate from the early Bornstein-era studies was somewhat inflated relative to the modern-methodology estimate. This is a pattern many robust behavioral findings show under modern re-examination: the basic regularity holds, but the size is smaller than the original-era literature implied. The pattern is consistent with publication-bias-driven over-estimation in the early literature without the underlying phenomenon being illusory.
Second, the exposure-liking effect was largest when the stimuli were affectively neutral or mildly positive at the outset. For stimuli that started with strong negative affective valence, repeated exposure did not reliably increase liking, and in some studies produced reactance --- additional exposure made the negative evaluation worse. This is a moderator that the early Zajonc work had not emphasized but that has obvious practical implications. The mere exposure effect is not a universal mechanism for converting dislike to liking; it operates most reliably on the neutral-to-mildly-positive end of the affective spectrum.
Third, the exposure-liking effect operated independently of explicit recognition. The team’s structural analysis confirmed that the affective shift was not mediated by explicit recall or familiarity ratings --- subjects could fail to recognize a previously-exposed stimulus and still show enhanced liking for it. This is the most striking aspect of the mere exposure literature and the one that most cleanly distinguishes it from a simple “more familiar feels safer” account. The mechanism appears to operate on perceptual-processing fluency at a sub-recognition level, which is the bridge to the next paper.
Fourth, the team’s mediation analysis supported a perceptual-fluency mechanism more strongly than competing accounts. Across the literature, the studies most consistent with the perceptual-fluency mechanism showed the largest and most reliable effects. The studies that had previously been interpreted through alternative mechanisms --- subjective familiarity, classical conditioning, response-bias artifacts --- showed weaker and less reliable patterns once methodological controls were applied. This was a substantial update to the theoretical understanding of why the mere exposure effect operates.
The Montoya re-examination is, by modern standards, the gold-standard meta-analytic treatment of the mere exposure effect. It survives publication-bias correction. It confirms the basic effect at a meaningfully-large effect size. It sharpens the moderator structure. And it provides empirical support for the perceptual-fluency mechanism that had been the leading theoretical account since the early 1990s.
The Perceptual Fluency Mechanism
The theoretical bridge between the exposure manipulation and the affective outcome is the perceptual-fluency hypothesis, articulated in Bornstein, R. F., & D’Agostino, P. R. (1992). “Stimulus Recognition and the Mere Exposure Effect.” Journal of Personality and Social Psychology, 63(4), 545—552. DOI: 10.1037/0022-3514.63.4.545.
Bornstein and D’Agostino proposed that repeated exposure to a stimulus increases the perceptual-processing fluency of that stimulus --- the ease and speed with which the perceptual system can encode and recognize the stimulus on subsequent encounters. Fluent processing feels subjectively good --- it is associated with a positive affective tone, often experienced as a mild sense of ease or familiarity. The subject, lacking explicit awareness of the source of this positive feeling, then mis-attributes the fluency-driven positive affect to the stimulus itself. The stimulus feels more likable because processing it is easier, and the ease-of-processing is interpreted as a property of the stimulus rather than as a property of the perceptual encounter.
This mechanism has several features that make it theoretically attractive and empirically tractable. It is parsimonious --- it does not require positing a special-purpose affect-formation mechanism but instead reduces to a general property of mis-attribution that is well-documented across many cognitive phenomena. It predicts the unconscious-exposure finding --- a subject who has not consciously registered the prior exposure will still process the stimulus more fluently on re-encounter, will still experience the fluency-driven positive affect, and will still mis-attribute it. It predicts the novelty-moderator finding --- the largest fluency gains happen during early exposures, when the perceptual system is moving from genuinely-effortful to easier processing, so the largest affective shifts should also happen at the lower end of the exposure-frequency range. And it predicts the attentive-exposure attenuation --- when subjects are explicitly aware of the source of their fluency, they can discount the affective signal as not being a property of the stimulus, attenuating the effect.
The empirical support for the perceptual-fluency account has accumulated steadily. Studies that manipulate processing fluency through other means --- contrast manipulations that make a stimulus easier to read, font choices that increase legibility, presentation duration adjustments --- produce affective shifts that look like mere exposure effects, supporting the proposition that fluency is the operative variable rather than exposure per se. Neuroscience work using EEG and fMRI has documented that previously-exposed stimuli activate perceptual-processing regions with reduced effort, and that this fluency activation is paralleled by activation in affect-related regions. The mechanism story is not perfectly settled --- there are competing accounts that propose roles for additional mechanisms including subjective familiarity, classical conditioning, and discriminative-attention --- but the perceptual-fluency account is the dominant and best-supported framework as of the most recent meta-analytic treatment.
The mechanism story matters practically for two reasons. First, it predicts the boundary conditions of the effect with more precision than a generic “exposure makes you like things more” framing would. Interventions that disrupt fluency (cluttered presentation, low-contrast formats, very long exposure durations that produce satiation) will attenuate the effect; interventions that enhance fluency (clean presentation, repeated short exposures, contextually-coherent surroundings) will magnify it. Second, it provides a generalizable account that connects mere exposure to other fluency-driven judgment effects in the literature --- the truth-by-repetition effect, the moniker-related fluency effects in stock-price studies, the rhyme-as-reason effect in persuasion --- which means the mere exposure literature is consistent with and reinforces a broader body of cognitive-psychology work on the role of processing ease in judgment.
Real-World Applications
Once you understand the mere exposure effect as a perceptual-fluency-driven enhancement of affective evaluation under repeated exposure, you start to see it operating across a remarkable range of consequential real-world domains.
Advertising frequency. The advertising literature on optimal-frequency-of-exposure --- the long-running debate over how many times a consumer needs to see an ad before purchase intent meaningfully shifts --- is directly informed by mere exposure theory. The classic Krugman three-exposure framing, the more modern frequency-capping rules in programmatic display advertising, the seasonal-recency effects in CPG advertising --- all of these operate within an empirical regime where the mere exposure mechanism is one of the load-bearing inputs. The fluency-driven affect shift is part of why repeated brand impressions over time produce a measurable lift in brand consideration even in the absence of any explicit message comprehension, and why the advertising industry’s intuition that “you need to be seen multiple times” is empirically grounded rather than mythological.
Brand familiarity and category dominance. The dominance of incumbent brands in mature consumer-product categories is partly mediated by the mere exposure effect. Coca-Cola is not just preferred because consumers have learned through extensive product experience that it tastes good --- it is also preferred because its visual identity, sonic logo, and product imagery have been encountered thousands of times across decades, producing a fluency-driven affective baseline that any challenger brand has to overcome. This is part of why new entrants in mature categories typically need to spend disproportionately on awareness-building before they can even compete on product attributes --- the incumbent advantage is partly an exposure-fluency advantage that compounds over time.
Music popularity and radio rotation. The music industry has long understood that radio rotation drives song popularity in a way that is not fully reducible to musical quality --- a song that gets played frequently moves up the charts even when listener surveys show no consistent quality-rating advantage over rotation-matched competitors. This is the mere exposure effect operating in the auditory domain. Modern streaming-platform algorithms have made the relationship even more explicit: a song that gets surfaced to a listener repeatedly will be added to playlists at higher rates than a song with similar audio features that does not get surfaced as often. The fluency-driven affective enhancement makes the song feel more likable on each subsequent play, until the playback frequency saturates and the effect plateaus or reverses through satiation.
Political campaign repetition and name recognition. Political-campaign research has documented that simple name recognition is a meaningful predictor of vote choice in low-information races, and that name recognition is built primarily through repeated exposure --- yard signs, mailers, radio mentions, social-media impressions. The fluency-driven mechanism is part of why a candidate who is more frequently exposed gets read as more likable and more competent by voters who have no other basis for evaluation. The effect attenuates in high-information races where voters have substantive policy positions to evaluate, but in down-ballot races and primaries it can be decisive. The repetition strategy is not pseudo-scientific; it operates on a well-documented mechanism.
Social-network friend formation and proximity-based attraction. The classic Festinger work on friendship formation in a college dormitory --- showing that physical proximity strongly predicts who becomes friends with whom, beyond what would be expected from any deliberate compatibility-screening --- is partly a mere exposure phenomenon. People you encounter frequently feel more likable, even when the encounters themselves carry no relational content. This generalizes to office-seating arrangements, to repeated encounters at recurring meetings, to online-platform encounter frequency. The implication for organizational design is that exposure-frequency among employees is a meaningful input to network formation, with downstream effects on collaboration patterns and informal-network resilience.
Product packaging recognition and grocery-shelf attention. CPG product design has internalized the mere exposure mechanism --- a product whose packaging is encountered repeatedly in a category becomes the easier-to-process visual stimulus on the shelf, and the easier-to-process stimulus is read as more likable and more credible. This is part of the moat that incumbent products enjoy against newer entrants and part of why repackaging redesigns are risky --- they reset the fluency accumulation that the previous design had built. The shelf-attention literature treats this explicitly, and the design conservatism of established CPG brands is partly an empirical response to the fluency-disruption risk.
Algorithmic recommendation and content feeds. Modern recommendation systems, by their nature, produce repeated exposure to certain creators, certain styles, certain product categories. The mere exposure effect operates inside these systems --- a creator who is surfaced repeatedly to a user becomes more likable to that user on each subsequent surfacing, partly through fluency mechanisms, until satiation effects begin to attenuate the gains. This is part of why social-media platforms produce engagement-amplification dynamics that look like preference-formation but are partly fluency-formation, and it is part of why preference-trajectory analytics in recommendation systems need to account for exposure-driven liking shifts rather than treating expressed preferences as exposure-independent.
The general pattern: the mere exposure effect is operating constantly, across consumer markets, political contexts, social-network formation, and algorithmic feed dynamics. It is one of the cognitive mechanisms with the broadest real-world footprint, and the practical interventions that exploit it --- advertising-frequency optimization, awareness-building campaigns, recommendation-system design, brand-identity consistency --- are all evidence-supported strategies that produce real outcome shifts.
What This Means For Strategists
The practical takeaways for someone making real decisions about marketing, branding, product design, advertising spend, or organizational design are:
Mere exposure is exploitable in marketing and brand-building, and the exploitation is well-evidenced. Repeated brand impressions, sustained advertising frequency, packaging-design consistency, and any other input that increases the fluency with which a consumer can encounter and process your brand will move purchase intent and brand-consideration metrics. The marketing-textbook claim that “awareness compounds over time” is, for once, supported by the underlying cognitive science. The implication is that brand-investment decisions that produce repeated exposure across consumer touchpoints --- even when those individual exposures carry no explicit informational content --- are likely to produce returns that the explicit-content-only models would miss.
The optimal exposure structure is not continuous saturation but distributed repetition with novelty preservation. The Bornstein and Montoya moderator findings have a sharp practical implication. Repeated exposure that is interspersed with other stimuli, presented in brief encounters rather than long sessions, and preserves some sense of stimulus novelty produces the largest fluency-driven affective shifts. Continuous high-saturation campaigns produce diminishing returns and can cross into satiation, where additional exposure reverses the effect. The implication for media planning is that frequency-capping rules and creative-rotation strategies are not just nice-to-have hygiene --- they are mechanism-grounded levers for keeping the exposure-liking relationship in its productive range.
The mere exposure effect does not convert dislike to liking. The Montoya finding that the effect operates most reliably on neutral-to-mildly-positive stimuli has direct implications for crisis-recovery and rehabilitation campaigns. A brand that has acquired strong negative valence with a market segment will not be repaired through additional exposure --- additional exposure can deepen the negative association rather than soften it. The effective response in those situations is substantive repositioning (changing the meaning of the brand) followed by exposure-based reinforcement of the new positioning, not pure repetition of the existing brand artifacts.
Mere exposure operates pre-consciously, which has both ethical and tactical implications. Because the effect operates through a fluency mechanism that does not require conscious recognition or deliberate evaluation, consumers cannot easily defend against it through willpower or conscious counter-effort. This means that exposure-based branding strategies will move purchase behavior even with consumers who consciously dislike advertising or who consider themselves immune to brand appeals. It also means that ethical questions about exposure-driven persuasion cannot be resolved by appealing to consumer autonomy in the simple “they can choose to ignore it” framing --- the choosing-to-ignore option is partial at best, because the mechanism operates on processes that are not under conscious control. The mature ethical framing is that exposure-based branding is a legitimate marketing tool when the underlying product proposition is honest and the exposure does not crowd out informational content that consumers need to evaluate the product on its merits.
Be aware of mere exposure operating on your own strategic choices. Decision-makers are not immune to the mechanism. A vendor whose representative shows up at every industry conference for three years will feel more credible and likable to a buyer than a vendor whose product is superior but who has not invested in exposure-frequency. A board candidate whose name has appeared in industry publications will feel more competent than a similarly-qualified candidate who has not been visible. An investment thesis that has appeared repeatedly in your information diet will feel more correct than a competing thesis with similar evidence. The defense, as with availability bias, is structural rather than willpower-based: insert explicit substantive-evaluation steps into the decision process before allowing exposure-driven liking to shape the choice.
Use mere exposure deliberately and ethically. The same mechanism that builds legitimate brand equity through honest sustained presence is what underlies the manipulative variant of repetition-based persuasion --- false claims repeated enough times become subjectively more credible through fluency mechanisms (the related “illusory-truth effect” literature). The line between legitimate brand-building and manipulative repetition is the same line that runs through all marketing ethics: are the claims you are exposing the consumer to substantively accurate, and would the exposure-driven preference persist after the consumer received complete substantive information about the product? The version of mere-exposure marketing that survives ethics review is the version where the underlying proposition is honest and the exposure mechanism is operating on top of an accurate informational base, not as a substitute for it.
The general framing: mere exposure is real, large, and well-evidenced. It is genuinely useful as a marketing and brand-building input and as a diagnostic lens for understanding why certain market structures (incumbent dominance, recommendation-system dynamics, political name-recognition advantages) operate the way they do. It is also one of the cognitive mechanisms you should expect to be operating on your own evaluations, and the defense against it is structural rather than willpower-based.
What This Anti-Example Tells Us About Cognitive Psychology Overall
The replication crisis is real. The catalog of canonical-then-collapsed findings in this hub is long. A reasonable executive could read all of that and conclude that cognitive and behavioral psychology is mostly an academic exercise --- interesting theory, weak evidence, low practical reliability.
That conclusion would be wrong, and the mere exposure effect, alongside the availability heuristic and the default effect, is among the best counter-examples.
What cognitive psychology can produce, when it does the work properly, is a small number of robust, mechanism-grounded, replicable findings that have meaningfully advanced our understanding of how the mind actually works. Mere exposure is one of them. Availability is another. Defaults are a third. The combined set of well-replicated, well-mechanism-specified findings is perhaps a dozen entries across the broader behavioral- and cognitive-economics literature.
The job of a sophisticated reader of this literature is to distinguish the dozen-or-so robust findings from the much larger catalog of fragile findings. The hub you are reading is a guided tour of that distinction. The mere exposure effect is one of the answers you take away. The availability-heuristic and default-effect articles are others. The collapsed-findings catalog (power posing, ego depletion, money priming, the marshmallow test in its inflated form, facial feedback, social priming, and the rest) is the part of the picture that earned the field its post-crisis reputation.
The synthesis: cognitive and behavioral psychology are more useful than the replication-crisis critique alone would suggest, and less useful than the original consulting-deck version implied. The mature, calibrated, post-replication-crisis version of the field is in better shape than either the original version or the pessimistic post-crisis version. The mere exposure effect is one of the findings that demonstrates this is possible.
Sources
- Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology Monograph Supplement, 9(2, Pt. 2), 1—27. DOI: 10.1037/h0025848
- Bornstein, R. F. (1989). Exposure and affect: Overview and meta-analysis of research, 1968-1987. Psychological Bulletin, 106(2), 265—289. DOI: 10.1037/0033-2909.106.2.265
- Bornstein, R. F., & D’Agostino, P. R. (1992). Stimulus recognition and the mere exposure effect. Journal of Personality and Social Psychology, 63(4), 545—552. DOI: 10.1037/0022-3514.63.4.545
- Montoya, R. M., Horton, R. S., Vevea, J. L., Citkowicz, M., & Lauber, E. A. (2017). A re-examination of the mere exposure effect: The influence of repeated exposure on recognition, familiarity, and liking. Psychological Bulletin, 143(5), 459—498. DOI: 10.1037/bul0000085
- Zajonc, R. B. (2001). Mere exposure: A gateway to the subliminal. Current Directions in Psychological Science, 10(6), 224—228. DOI: 10.1111/1467-8721.00154
- Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8(4), 364—382. DOI: 10.1207/s15327957pspr0804_3
Related
Browse the full Replication Crisis Hub for other behavioral-science findings, including:
- The Availability Heuristic --- another anti-example; the Tversky-Kahneman finding that survived five decades
- The Halo Effect --- a related judgment-formation bias with its own evidence profile
- Defaults and Status Quo Bias (Anti-Example) --- what a robust nudge actually looks like in field experiments
- Confirmation Bias --- another well-replicated cognitive bias with broad real-world footprint
- The Framing Effect --- the Tversky-Kahneman demonstration that has held up better than most prospect-theory derivatives
FAQ
Is the mere exposure effect strong enough to drive actual behavior change, not just rated liking?
Yes. The empirical literature includes behavior-outcome studies, not just self-report-rating studies. Repeated exposure to a product brand has been shown to shift product-choice in field-experimental designs. Repeated exposure to a political candidate’s name has been shown to shift vote choice in low-information ballot races. Repeated exposure to a song through radio rotation has been shown to shift purchase behavior in the music-retail era. The behavior-outcome literature is smaller than the laboratory-rating literature, but the direction of the behavioral effects is consistent with the rating effects, and the effect sizes are practically meaningful in commercial contexts.
How does the mere exposure effect interact with the truth-by-repetition effect?
The two effects share the perceptual-fluency mechanism but operate on different judgment dimensions --- mere exposure on affective liking, truth-by-repetition on perceived factual accuracy. A claim that has been repeated multiple times feels more likely to be true, partly because the repeated processing produces fluency, and fluency is read as a credibility cue. This is the mechanism that makes repeated false claims dangerous in political and misinformation contexts. The two effects are theoretically and empirically related, and both are part of the broader fluency-based-judgment literature.
How do I reduce mere exposure bias in my own decision-making?
The defense is structural, not willpower-based. Because the mechanism operates pre-consciously on fluency cues rather than at the deliberative-evaluation stage, “I will try harder to not be influenced by familiarity” does not work. The intervention that works is to force substantive-criteria evaluation before allowing exposure-driven liking to shape choices. For vendor selection, this is structured RFP scoring against pre-specified criteria. For hiring, this is structured interviewing with pre-defined evaluation rubrics. For investment, this is checklists that force evaluation of substantive thesis-strength before relative-familiarity comparisons enter the decision. The structural defense is durable; the awareness-based defense is not.
Is the mere exposure effect universal, or are there cultural moderators?
The basic effect has been replicated across cultures and across very different stimulus types, suggesting the underlying perceptual-fluency mechanism is reasonably universal. There are cultural moderators on the magnitude --- the effect appears somewhat larger in collectivist than individualist cultures for socially-relevant stimuli, and the affective interpretation of fluency may carry different weight in different cultural contexts. But the basic regularity (repeated exposure produces enhanced affective evaluation) is one of the more cross-culturally robust findings in cognitive psychology.
How does the mere exposure effect interact with social-media algorithm design?
Recommendation algorithms produce repeated exposure to certain creators, certain styles, and certain content categories as a structural feature. This means the algorithms are running a continuous mere-exposure intervention on every user, with the result that user-expressed preferences after algorithmic exposure are not a clean measurement of underlying preference --- they are a measurement of underlying preference shifted by the fluency-driven liking enhancement that the exposure pattern has produced. The implication for product analytics is that engagement and preference metrics inside algorithm-driven feeds need to be interpreted with the exposure-shift in mind, rather than treated as exposure-independent indicators of intrinsic preference.
What about advertising satiation --- doesn’t repeated exposure eventually backfire?
Yes, and this is one of the Bornstein moderator findings. The exposure-liking relationship is not monotonically increasing across all frequencies. At very high exposure frequencies, particularly when exposures are massed in time rather than distributed, the effect plateaus and can reverse through satiation. This is part of why advertising-frequency optimization is not just “more is better” --- there is an inverted-U shape in the exposure-liking relationship, with diminishing returns and eventual negative returns at extreme high frequencies. The practical management of this is through frequency-capping rules and creative-rotation strategies that keep the per-creative exposure count in the productive range of the curve.
Does the mere exposure effect explain why famous brands beat generic equivalents in blind taste tests?
The blind-taste-test literature is complicated and the mere exposure effect is one of several mechanisms that has been proposed. In blind comparisons, where the brand identity is removed, famous brands do not consistently outperform generic equivalents on taste --- this is well-documented for cola products, for store-brand vs national-brand grocery items, and for several other product categories. When the brand identity is visible, famous brands outperform substantially, and this is partly attributable to mere exposure operating on brand-imagery rather than on the product itself. The fluency-driven affective enhancement of the brand identity carries over into the consumer’s evaluation of the product, even when the underlying product is similar. This is one of the cleaner real-world demonstrations of mere exposure operating commercially.
How do I distinguish robust cognitive-psychology findings from fragile ones?
The same four-question diagnostic from the default-effect and availability-heuristic articles applies. Is the mechanism over-determined (multiple plausible mechanisms predict the effect, so attacking any single mechanism does not dismantle the prediction)? Are the conditions of application well-specified (so failed replications can be diagnosed rather than dismissed)? Is the effect size large enough to detect without statistical heroics? Has it been measured in high-stakes field experiments with administrative-data outcomes, not just convenience-sample lab work? The mere exposure effect, like availability and defaults, passes all four. Most of the failed findings in this hub fail at least three.
Is the mere exposure effect actionable in B2B and enterprise sales contexts?
Yes. The mechanism does not depend on consumer-product context; it operates on any decision-maker who can be exposed to a brand or vendor repeatedly. The practical implication is that sustained presence at industry events, sustained content publication in trade media, sustained representation in analyst reports, and sustained visibility in customer-relationship contexts will produce a fluency-driven affective baseline that compounds over the multi-year sales cycle typical of enterprise procurement. This is part of why incumbent-vendor advantages in enterprise software are so durable --- the incumbent vendor has accumulated substantial exposure-fluency with the buying committee through years of presence, which an insurgent vendor has to overcome on top of the rational substantive comparison. The exposure-investment strategy in B2B is the analog of the brand-building strategy in B2C, and it operates on the same underlying mechanism.
What is the single best deployment of the mere exposure effect in a startup marketing program?
Sustained brand-identity consistency across all touchpoints, repeated over years. The compound interest on a consistent visual identity, sonic logo, language style, and content cadence accrues through the fluency mechanism in a way that produces affective baseline shifts in the target market that pay back across the lifetime of the company. The most-leveraged version of this is a brand-identity decision made early, executed consistently across every consumer touchpoint (web, product, sales materials, social, advertising), and preserved through the natural temptation to redesign or refresh. This is one of the highest-confidence behavioral-marketing investments available to a startup, and it has the additional advantage of being a marketing asset that compounds over time rather than degrading.
replication-crisis mere-exposure-effect zajonc cognitive-psychology evidence-evaluation
Free Tool
Built for Experimentation Teams
GrowthLayer is the experimentation platform I built for CRO teams --- test management, AI-powered insights, and pattern recognition across your entire program.
Explore GrowthLayer → (opens in new tab)
Share this article
LinkedIn (opens in new tab) X / Twitter (opens in new tab)
Copy link
Go deeper
Methodology The PRISM Method Case Studies $30M+ in Results Work Together Services & Mentoring
Experimentation and growth leader. Builds AI-powered tools, runs conversion programs, and writes about economics, behavioral science, and shipping faster.