ArticlePDF Available

Placebo Effects of Marketing Actions: Consumers May Get What They Pay For

Authors:

Abstract and Figures

The authors thank Himanshu Mishra, Chris Tay-lor (parts of this research formed the basis of his honors thesis), and Mon-ica Wadhwa for their help in administering the experiments and coding participants' responses. They also thank Eric Johnson, participants of the Distinguished Scholars Retreat at the University of Alberta, and the three anonymous JMR reviewers for their comments and suggestions. We dedi-cate this article to the late Dick Wittink for his exceptional guidance and encouragement. May he rest in peace. Editor's note: In contrast to other lead articles that were invited by for-mer editor Dick Wittink, this article is based on a regular submission. The authors demonstrate that marketing actions, such as pricing, can alter the actual efficacy of products to which they are applied. These placebo effects stem from activation of expectancies about the efficacy of the product, a process that appears not to be conscious. In three experi-ments, the authors show that consumers who pay a discounted price for a product (e.g., an energy drink thought to increase mental acuity) may derive less actual benefit from consuming this product (e.g., they are able to solve fewer puzzles) than consumers who purchase and con-sume the exact same product but pay its regular price. The studies con-sistently support the role of expectancies in mediating this placebo effect. The authors conclude with a discussion of theoretical, managerial, and public policy implications of the findings.
Content may be subject to copyright.
Journal of Marketing Research
Vol. XLII (November 2005), 383–393
383
©2005, American Marketing Association
ISSN: 0022-2437 (print), 1547-7193 (electronic)
*Baba Shiv is Associate Professor of Marketing, Graduate School of
Business, Stanford University (e-mail: shiv_baba@gsb.stanford.edu). Ziv
Carmon is Associate Professor of Marketing, INSEAD, Asia Campus
(e-mail: ziv.carmon@insead.edu). Dan Ariely is Luis Alvarez Renta Pro-
fessor of Management Science, Massachusetts Institute of Technology
(e-mail: ariely@mit.edu). The authors thank Himanshu Mishra, Chris Tay-
lor (parts of this research formed the basis of his honors thesis), and Mon-
ica Wadhwa for their help in administering the experiments and coding
participants’ responses. They also thank Eric Johnson, participants of the
Distinguished Scholars Retreat at the University of Alberta, and the three
anonymous JMR reviewers for their comments and suggestions. We dedi-
cate this article to the late Dick Wittink for his exceptional guidance and
encouragement. May he rest in peace.
Editor’s note: In contrast to other lead articles that were invited by for-
mer editor Dick Wittink, this article is based on a regular submission.
BABA SHIV, ZIV CARMON, and DAN ARIELY*
The authors demonstrate that marketing actions, such as pricing, can
alter the actual efficacy of products to which they are applied. These
placebo effects stem from activation of expectancies about the efficacy of
the product, a process that appears not to be conscious. In three experi-
ments, the authors show that consumers who pay a discounted price for
a product (e.g., an energy drink thought to increase mental acuity) may
derive less actual benefit from consuming this product (e.g., they are
able to solve fewer puzzles) than consumers who purchase and con-
sume the exact same product but pay its regular price. The studies con-
sistently support the role of expectancies in mediating this placebo effect.
The authors conclude with a discussion of theoretical, managerial, and
public policy implications of the findings.
Placebo Effects of Marketing Actions:
Consumers May Get What They Pay For
Pro tali numismate tales merces. [One gets what one
pays for.]
—Gabriel Biel (Dictionary of Clichés, Rogers 1985)
Consumers’ beliefs and expectations, shaped by experi-
ences in their daily lives, often influence their judgments of
products and services. For example, consumers often
believe and, therefore, judge lower-priced items to be of
lower quality (see, e.g., Gerstner 1985; Huber and McCann
1982; Rao and Monroe 1989). Consumers’ beliefs and
expectations can also affect their subjective experiences.
For example, a drink may taste better if it has a favorite
brand’s label than if it is unlabeled (Allison and Uhl 1964;
McClure et al. 2004). Similarly, meat that is labeled 75% fat
free tastes better than the same meat that is labeled as con-
taining 25% fat (Levin and Gaeth 1988). The question that
we address in this research is whether beliefs and expecta-
tions that marketing actions evoke can affect more than
judgments and subjective consumption experiences. Specif-
ically, can they also influence the actual efficacy of the mar-
keted product? For example, can consuming an energy
drink that is purchased at a discount lead not only to judg-
ments of lower quality or to a less favorable consumption
experience but also to diminished performance in, for
example, a cardiovascular workout or a puzzle-solving
task?
We began to explore these questions in a preliminary
study in which 38 members of a fitness center who exer-
cised regularly (at least three times a week) consumed
Twinlab Ultra Fuel before and during a workout session.
Before consuming the energy drink, participants were
shown the list of its ingredients and were told that the drink
was from the most recently manufactured batch. One group
of participants was told that we purchased the drink at the
regular price of $2.89; another group was told that the regu-
lar price of the drink was $2.89 but that we had purchased it
at a discounted price of $.89 because we bought it in bulk as
an institutional purchase. After exercising, participants rated
the intensity of their workout on a scale that ranged from –3
(“not at all intense”) to +3 (“very intense”) and how
fatigued they felt on a scale that ranged from 1 (“not at all”)
to 7 (“very”). The results show that participants in the
reduced-price condition rated their workout intensity as
lower (M = –.4) than did those in the regular-price condi-
tion (M = .6; F(1, 36) = 7.5, p< .01), and participants in the
reduced-price condition indicated that they were more
fatigued (M = 4.5) than did those in the regular-price condi-
tion (M = 3.7; F(1, 36) = 3.5, p< .10). Finally, when asked
during debriefing if the price of the drink affected the work-
out, not a single participant answered affirmatively.
The findings of our preliminary study share a kinship
with the well-known placebo phenomenon in the medical
384 JOURNAL OF MARKETING RESEARCH, NOVEMBER 2005
1Although some researchers (e.g., Hahn 1997) distinguish between
placebo and nocebo effects, we use the term placebo for both, in line with
the common view that the desirability of the effect should not influence its
labeling (Stewart-Williams and Podd 2004).
domain (see Stewart-Williams and Podd 2004). Specifi-
cally, patients’ beliefs and expectations about the treatment
they are receiving (e.g., an antidepression medication) can
yield real changes to their health, even if the treatment is
actually inert and has no inherent power to produce health
effects (e.g., an inert sugar pill that looks like the anti-
depression medication). A large body of research has exam-
ined effects of aspects that are inherent to a placebo, such as
beliefs about the efficacy of a drug for which the placebo is
a substitute (e.g., Kirsch 1999) or about the form in which
the placebo is received (e.g., Kaptchuk et al. 2000). The
results of our preliminary study suggest that features that
are not inherent to a product, such as its price, can also trig-
ger a placebo effect.
The preliminary study addresses the questions we raised
in the first paragraph, and its results suggest that price dis-
counts can lead to a behavioral effect—we refer to this as a
placebo effect of marketing action in this research—and
that this effect may occur beyond awareness. However, sev-
eral criticisms can be leveled against this study. First, our
dependent measure in this study is our participants’ percep-
tions of their behavior (e.g., perceived workout intensity)
rather than the behavior itself. Second, our study does not
include a no-treatment control group. Therefore, we cannot
distinguish between a desirable placebo effect (i.e., the
regular-price fitness drink boosted the efficacy of the drink)
and an undesirable placebo effect (i.e., the sale price
detracted from the efficacy of the energy drink).1Third,
participants did not actually pay for the drink but were
merely informed of the prices. Finally, and most important,
this study gives no indication of what underlies the effect.
We address all these criticisms in our subsequent studies
and also explore whether the effect is nonconscious, as the
preliminary study suggests.
In the next section, we draw on research on the placebo
effect and on the price–quality association to predict how
beliefs and expectations that arise from marketing actions,
such as price promotions, may produce effects on behavior.
Following this, we present three experiments that document
undesirable placebo effects that result from price discounts.
In the third experiment, we also document a desirable
placebo effect that is ignited by advertising claims. In all
three experiments, we find support for the role of expectan-
cies as underlying this effect, and we rule out alternative
accounts.
THEORETICAL BACKGROUND
Voluminous research on placebo effects has shown that
successfully conveying the false belief that patients received
a particular treatment can bestow some of the benefits of the
genuine treatment (for a review, see Stewart-Williams and
Podd 2004). Credible placebos can help relieve and some-
times even cure physical and mental illness, such as pain
(e.g., Montgomery and Kirsch 1996), cardiovascular dis-
ease (e.g., Bienenfeld, Frishman, and Glasser 1996), and
depression (Kirsch and Sapirstein 1999). Placebo effects
have also been detected with functional magnetic resonance
imaging (fMRI; Wager et al. 2004).
Two notions are believed to account for placebo effects:
expectancy theory and classical conditioning. According to
expectancy theory, placebo effects arise because beliefs
about a substance/procedure serving as a placebo activate
expectations that a particular effect will occur, which then
affect the subsequent effectiveness of the substance/
procedure. The classical conditioning view considers con-
suming substances with known therapeutic effects to be
conditioning trials. The active substances giving rise to
these effects serve as unconditioned stimuli, and the vehi-
cles by which they are delivered (e.g., pills, capsules,
drinks) serve as the conditioned stimuli. Pairing the uncon-
ditioned stimuli and the conditioned stimuli over time
endows the vehicles with a capacity to evoke therapeutic
effects in the form of conditioned responses. These two
views have been contrasted and debated, but an emerging
view is that expectancies mediate all placebo effects, and
conditioning is one means by which people initially form
and then activate expectancies (Kirsch 2004; Rescorla
1988). The growing acceptance of expectations as the basic
mechanism for placebo effects has led to an increased inter-
est in how beliefs lead to placebo effects and the role of
expectancies in mediating this effect.
The Mediating Role of Expectancies
Figure 1 highlights the process associated with placebo
effects, a framework that we derive from work in the med-
ical domain (Kirsch 1999; Kirsch and Lynn 1999; Kirsch
Response
Expectancies
Subjective/Behavioral
Outcomes
Salient Beliefs
•Intrinsic aspects
•Extrinsic aspects
Other Factors
(e.g., self-efficacy beliefs and
related expectancies, as
reflected in no-treatment
control conditions)
Figure 1
FRAMEWORK FOR PLACEBO EFFECTS
Placebo Effects of Marketing Actions 385
and Sapirstein 1999; Stewart-Williams and Podd 2004). The
figure presents factors that are expected to influence
placebo effects of marketing actions. Briefly, when a person
receives what is purportedly an active substance or treat-
ment, his or her salient beliefs about the substance or treat-
ment activate response expectancies, or anticipations of
subjective and/or behavioral consequences of using the sub-
stance or being treated. These response expectancies along
with contextual factors that are unrelated to the substance or
treatment then lead to the subjective and behavioral out-
comes, or placebo effects.
Several aspects of this process warrant elucidation. First,
critical to the placebo effect are specific beliefs that are
salient when a person receives the purportedly active sub-
stance or treatment. These beliefs, for example, could be
related to intrinsic aspects of the active substance or treat-
ment, such as its potential therapeutic effects or its deleteri-
ous side effects, yielding a desirable placebo effect in the
former case and an undesirable placebo effect in the latter
case (see Hahn 1997). Similarly, extrinsic aspects can shape
salient beliefs about the substance or treatment and lead to
stronger (weaker) placebo effects. An example of this is
whether a person receives a medication through injections
or capsules (Kaptchuk et al. 2000).
Second, the magnitude of the subjective and/or behav-
ioral consequences depends on the strength of the activated
response expectancies, which in turn can be influenced by a
variety of factors. For example, encouraging people to elab-
orate on their expectations may enhance the magnitude of
the placebo effect (Fillmore and Vogel-Sprott 1992). Simi-
larly, the magnitude of the placebo effect may be influenced
by the strength of a person’s salient beliefs about the active
substance or treatment. In turn, for example, this can be
enhanced by greater familiarity through prior usage (Kirsch
1985). The strength of these beliefs and, therefore, the mag-
nitude of the placebo effect may also be diminished by
instructions that cast doubts about these beliefs. For exam-
ple, alerting people that they are participating in a double-
blind study and that the substance they are receiving may be
inert gives rise to diminished placebo effects (Kirsch and
Weixel 1988).
Third, the subjective and behavioral outcomes can be
shaped by self-efficacy beliefs (Kirsch 1985), which,
together with other extraneous factors, are reflected in out-
comes of no-treatment control conditions that are some-
times included in placebo studies. Finally, the process by
which expectancies are elicited to lead to the placebo effect
can be either conscious or nonconscious. Consistent with
one aim of our research, Stewart-Williams and Podd (2004)
call for research on placebo effects that will identify situa-
tions in which the mediating role of expectancies occurs
nonconsciously. Next, we examine implications of the
framework for placebo effects that may arise from market-
ing actions, such as price discounts.
Marketing Actions and the Placebo Effect
If marketing actions, such as price discounts, lead to a
placebo effect, as our preliminary study suggests, what
might be the nature of beliefs that trigger response
expectancies that, in turn, lead to the placebo effect? Fur-
thermore, how will contextual factors influence the strength
of the expectancies and, thus, the magnitude of the effect?
To answer these questions, consider the context of the pre-
liminary study. Recall that participants received an energy
drink (Twinlab Ultra Fuel), were shown a list of its ingredi-
ents, and were informed that we purchased the drink at
either its regular price or a discounted price. The stimulus
materials could have made several beliefs salient. For exam-
ple, intrinsic aspects related to the ingredients could have
activated beliefs about their effects. In addition, the brand
name (an extrinsic cue) could have activated beliefs about
the product’s superior quality (Rao and Monroe 1989). Fur-
thermore, given that consumers often believe that price lev-
els tend to reflect quality (e.g., Huber and McCann 1982;
Rao and Monroe 1988, 1989), the price discount (another
extrinsic cue) may have triggered beliefs that the product’s
quality was inferior. According to the framework we pre-
sented previously, all these beliefs could have been salient,
triggering various types of response expectancies. These
response expectancies along with other factors, such as
non-product-related beliefs (e.g., self-efficacy beliefs, such
as how good a person is at fitness workouts) and partici-
pants’ abilities, could have affected respondents’ perform-
ance in their fitness workout. However, because price was
the only manipulated factor in the preliminary experiment
and because participants were randomly assigned to the two
levels of this factor (thereby controlling for other factors,
such as beliefs about the ingredients, the brand name, or
self-efficacy), the difference we observed in participants’
performance is likely to have been due to the salient beliefs
about price, that is, a placebo effect of price discounts. Con-
versely, participants’ performance in a no-treatment control
condition would have reflected effects of the other factors,
such as non-product-related beliefs, participants’ abilities,
and so forth.
Next, we consider the implications of other aspects of the
framework in relation to our preliminary study, and we
draw predictions for similar studies we present subse-
quently. The magnitude of the placebo effect could be
affected by a host of factors. First, encouraging participants
to elaborate on their expectations could increase the
strength of those expectations and, thus, the magnitude of
the observed placebo effect. We test this prediction in our
first study. Second, beliefs about the brand name and/or the
ingredients should be stronger with greater (rather than
lower) frequency of prior usage. Thus, in addition to the
observed price effect, we expect that frequency of prior
usage influences the behavioral outcome. We provide evi-
dence relating to prior usage in our first two studies. Third,
if we draw participants’ attention to their price–quality
beliefs, many may realize that the price–quality relationship
is not applicable, which should weaken the impact of such
beliefs and, thus, the magnitude of the placebo effect. We
test this prediction in our second study. Finally, consider
what might happen if we presented advertising claims that
either strengthened or weakened participants’ beliefs in the
efficacy of the ingredients (an intrinsic cue). This manipula-
tion should have independent effects on participants’ per-
formance, in addition to the observed placebo effect of
price discounts. We examine this in the third study.
The discussion in the previous section also suggests that
the process by which expectancies are elicited to lead to the
placebo effect can be either conscious or nonconscious (see
Kirsch 2004; Stewart-Williams and Podd 2004). Thus, the
question is whether the placebo effects of price discounts
are conscious. Rao and Monroe (1988) argue that the rela-
386 JOURNAL OF MARKETING RESEARCH, NOVEMBER 2005
tionship between price and perceived quality is a belief that
is activated and used when people make rapid judgments
about a product’s quality. Adaval and Monroe (2002) sug-
gest that price–quality beliefs are activated and affect judg-
ments at a nonconscious level.
Building on these ideas, we conducted three experiments.
In these experiments, we investigated whether price dis-
counts lead to placebo effects by activating response expec-
tations and whether the process by which these activated
expectations lead to the placebo effect is in line with our
previous predictions.
EXPERIMENT 1
The purpose of Experiment 1 was to (1) document evi-
dence of a placebo effect caused by price discounts; (2)
explore what underlies the placebo effect by examining
whether the observed effects are mediated by expectancies,
in line with findings in the medical domain; and (3) deter-
mine whether the impact of expectancies on the observed
placebo effect occurs nonconsciously.
In this experiment, participants first consumed SoBe
Adrenaline Rush (a drink that claims to help increase men-
tal acuity on its package) and then solved a series of puz-
zles. Note that such drinks are familiar to the student popu-
lation from which we derived our sample. Indeed, in
response to measures we collected at the end of this experi-
ment, 92% of the participants stated that they had heard of
SoBe before, and 48% stated that they had consumed this
drink before. To accomplish the first goal of this experi-
ment, we collected a measure of performance, namely, the
number of puzzles solved correctly. To accomplish the sec-
ond goal, we adapted a procedure that Fillmore and Vogel-
Sprott (1992) used to vary the strength of activated response
expectancies. Specifically, one group of participants elabo-
rated on their expectancies by rating the expected efficacy
of the drink before solving the puzzles, and a second group
did not perform this rating task. If expectancies indeed
mediate an observed placebo effect, strengthening response
expectancies should amplify the basic effects. To accom-
plish the third objective, we used a straightforward depend-
ent variable approach that we adapted from previous work
on nonconscious processes (see, e.g., Fitzsimons, Char-
trand, and Fitzsimons 2004). After participants completed
the puzzle task, we asked them to rate how effective SoBe
was at improving their puzzle-solving performance on a
scale that ranged from 1 (“not at all effective”) to 7 (“very
effective”). If participants were (non)conscious of the
impact of expectancies related to the efficacy of SoBe on
their subsequent performance, this measure should (not)
mediate the effects of the independent variables on the
number of puzzles solved. Note that the underlying process
has two components: (1) activation of expectancies and (2)
subsequent impact of those expectancies on participants’
performance in the puzzle task. Even if the first component
occurs at a nonconscious level, the procedure we used to
strengthen expectancies would make expectancies con-
scious (in conditions in which participants were asked to
rate the efficacy of the drink). Thus, lack of mediation
would suggest only that the second component occurs non-
consciously. We explore the nature of the first component in
Experiment 2 because our approach in Experiment 1 does
not enable us to learn about it.
Design and Procedure
We used a 2 (price: regular versus discounted) ×2
(expectancy strength: high versus low) between-subjects
design. We randomly assigned 125 participants to the four
conditions. At the beginning of the session, participants
were told that as part of the study, they would consume
SoBe. As in the preliminary study, they were shown the
packaging and the ingredients it contained and were told
that the drink was from the most recently manufactured
batch. To reinforce the sense that the energy drink would
influence their performance, participants were then told that
they would watch a video for about ten minutes, purport-
edly to allow the ingredients to have their effects. They
were also told that after watching the video, they would
solve a series of word-jumble puzzles (e.g., TUPPIL, the
solution for which is PULPIT); their goal was to solve as
many puzzles as possible in the allotted 30 minutes. Before
the drink was distributed, participants were given a form
that authorized us to charge their university billing account
for the drink that they would consume. For some partici-
pants (regular-price conditions), the form stated that they
would be charged $1.89 and that this was the regular price
of the drink in retail outlets. For other participants (reduced-
price conditions), the form stated that the regular price of
the drink at retail outlets was $1.89 but that they would be
charged $.89 because we purchased the drink at a discount
as part of an institutional purchase.
Participants consumed the drink and then watched a
video for about ten minutes. They then received a booklet
that contained instructions on the cover page, followed by
the puzzles. The instructions on the cover sheet stated that
participants would have 30 minutes to solve 15 puzzles.
After the cover page, some respondents (high-expectancy-
strength conditions) were shown a page with the following
statements: “I feel that SoBe is ‘very bad’ (1)/‘very good’
(7) at improving concentration,” and “I feel that SoBe is
‘very bad’ (1)/‘very good’ (7) at improving mental perform-
ance.” Respondents in the low-expectancy-strength condi-
tions were not shown these statements. Subsequently, par-
ticipants engaged in the puzzle task and then responded to a
series of measures; finally, they were debriefed.
Other Measures
After solving the puzzles, participants indicated their
gender; whether they were familiar with SoBe; whether
they had consumed this drink before; and, in general, how
good and how experienced they were at solving puzzles
such as word jumbles. These measures served as covariates
in the various analyses. At the end of the instrument, we
asked participants to recall the price they paid for the drink
(in this and in our other experiments, all participants in the
treatment conditions recalled the price they had paid within
a range of +/–11% accuracy, and there were no differences
in recall across the various treatment conditions). After the
experiment, an independent coder determined the number
of puzzles that each respondent solved correctly.
Results
Pilot study. We first conducted a pilot study to assess par-
ticipants’ performance in a no-treatment (control) condition.
For the pilot study, we used 31 participants from the same
population as that of the main experiment. The procedure
Placebo Effects of Marketing Actions 387
closely followed the one we used in the main experiment,
except that the participants were not told about the SoBe
drink and did not consume it; they merely solved the puz-
zles and responded to a relevant subset of the measures. The
average number of puzzles solved by the participants of the
pilot study was 9.1.
Main experiment. The average number of puzzles solved
across the various conditions appears in Figure 2. An analy-
sis of covariance (ANCOVA) on the number of puzzles
solved revealed an interaction between price and
expectancy strength (F(1, 120) = 5.6, p< .05) and a main
effect of price (F(1, 120) = 34.7, p< .0001). The pattern of
results in the low-expectancy-strength conditions was con-
sistent with a placebo effect; the number of puzzles solved
was lower in the reduced-price condition (M = 7.7) than in
the regular-price condition (M = 9.5; F(1, 120) = 5.9, p<
.05).
The results in the high-expectancy-strength conditions
suggest that the observed placebo effect was indeed medi-
ated by expectancies about the efficacy of the drink. As
Figure 2 shows, when expectancy strength was high, the
magnitude of the undesirable placebo effect in the low-
expectancy-strength conditions increased. Specifically,
when the price was discounted, the number of puzzles
solved decreased further at higher (M = 5.8) rather than
lower (M = 7.7) levels of expectancy strength (F(1, 120) =
7.7, p< .01).
A Sobel test (e.g., MacKinnon et al. 2002) within the
high-expectancy-strength conditions further supports medi-
ation by expectancies. Recall that we operationalized
expectancy strength by asking respondents in the high-
expectancy-strength conditions to rate how good SoBe was
at improving concentration and mental performance before
they solved the puzzles (we did not ask respondents in the
low-expectancy-strength conditions this question). In line
with our conceptualization, these ratings were higher in the
regular-price condition (M = 4.3) than in the reduced-price
condition (M = 3.5; F(1, 64) = 11.1, p< .01). The Sobel test
revealed that these ratings also mediated the effects of the
independent variable, price, on the number of puzzles
solved within the high-expectancy-strength conditions (z =
3.0, p< .01).
The only covariate that was relevant in the ANCOVA was
prior consumption of SoBe, a variable that did not interact
with any of the other independent factors. To explore the
effects of prior usage further, we performed an analysis of
variance with prior usage, a categorical variable, as a third
independent factor. Again, the only effect we observed for
prior usage was a main effect; on average, participants who
had consumed SoBe before solved more puzzles (M = 8.7)
than did those who had not (M = 7.7; F(1, 117) = 4.2, p<
.05), as we predicted in our framework.
Was the underlying process conscious? To answer this
question, we examined participants’ responses to a measure
that we collected after they solved the puzzles. Recall that
after participants solved the puzzles, they were asked how
effective they believed SoBe was at improving their puzzle-
solving performance (1 = “not at all effective,” 7 = “very
effective”). Had participants been conscious that their per-
formance was affected by consumption of the drink, this
measure would have mediated the effects of the independ-
ent variables on the number of puzzles solved. A Sobel test
did not support this possibility (p= .64), suggesting that
expectancies were not conscious when participants solved
the puzzles. However, note that this null result is difficult to
interpret because it could have occurred for reasons other
than the process being nonconscious. We test the potential
nonconscious nature of the underlying process more explic-
itly in the next experiment.
Discussion
The results of Experiment 1 support our basic prediction
that price discounts can lead to an undesirable placebo
effect. Offering a price discount on a product that claims to
be beneficial for mental acuity negatively affected perform-
ance on a subsequent task, namely, the number of puzzles
solved correctly. The performance of participants who con-
sumed the discounted drink was worse than that of partici-
pants who consumed the regular-priced drink and of those
who were in the no-treatment pilot study (and did not con-
sume the drink). This detrimental effect was accentuated
when we reinforced expectations about the efficacy of the
product. Furthermore, on average, participants who had
consumed the energy drink before solved more puzzles than
did those who had not. However, this variable did not inter-
act with any of the other independent variables. Finally, the
results suggest that the underlying process leading to our
observed placebo effect occurred nonconsciously.
A noteworthy finding in Experiment 1 was that we
observed only an undesirable placebo effect in the reduced-
price conditions. The results in the regular-price conditions
were no different than the results in the no-treatment control
condition (which we administered as a pretest). In the next
experiment, we examine whether the findings of Experi-
ment 1 replicate and shed more light on the cause of the
observed placebo effect.
EXPERIMENT 2
One goal of Experiment 2 was to examine why we did
not observe a desirable placebo effect of the regular-priced
drink in Experiment 1. A second goal was to rule out two
alternative accounts: (1) that participants paying the regular
7.7
9.5
5.8
9.9
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
Low Expectancy Strength High Expectancy Strength
Discounted price Full price
Notes: The number of puzzles solved in the control condition = 9.1.
Figure 2
NUMBER OF PUZZLES SOLVED: EXPERIMENT 1
388 JOURNAL OF MARKETING RESEARCH, NOVEMBER 2005
2This assumption was supported in a separate pretest. We used 33 par-
ticipants from the same population as that of the main studies and asked
them to perform a task similar to that in Experiment 1. Participants were
presented with the energy drink, SoBe, and informed that its regular price
was $1.89, but they would buy it from us at a discounted price of $.89. Fol-
lowing a filler task, participants rated the perceived efficacy of the drink
(as in the high-expectancy-strength conditions of Experiment 1). However,
for one group of participants, the price–efficacy link was made salient
before they rated the drink (they were told, “Given the price of SoBe,
please rate…”). Consistent with our assumption, the mean ratings were
lower when the price–efficacy link was made salient (M = 3.8) than when
it was not (M = 4.9; F(1, 31) = 7.6, p< .01).
price might have worked harder on the puzzle task to reduce
the greater dissonance they might have experienced as a
result of the regular price they paid and (2) that compared
with participants in the regular-price condition, those in the
reduced-price conditions concentrated less on the puzzle
task as a result of entertaining distracting thoughts about
getting the drink at a lower price. A third goal of Experi-
ment 2 was to examine whether drawing attention to price–
quality beliefs would affect the observed placebo effect.
This enables us to test several predictions. First, drawing
participants’ attention to price–efficacy beliefs is likely to
help some of them realize that these beliefs may not be
applicable to all contexts. In turn, this should weaken their
response expectations and, thus, the magnitude of the
placebo effect (see Kirsch and Weixel 1988).2Second, the
procedure enables us to shed more light on whether the
underlying process is nonconscious. Research has consis-
tently shown that if the activation of information in memory
occurs nonconsciously, drawing attention to the priming
source (in our case, the relationship between price and
expected efficacy) reduces subsequent effects of this infor-
mation (e.g., Strack et al. 1993). This attenuating effect is
likely to occur when drawing attention to the priming
source casts doubts on the relevance of the priming source.
Conversely, if the activation of the information occurs con-
sciously, drawing attention to the priming source enhances
subsequent effects of the information. Third, if drawing par-
ticipants’ attention to the price–efficacy beliefs reduced the
magnitude of the observed placebo effect, it would reduce
the viability of the alternative cognitive dissonance account.
According to the dissonance explanation, drawing attention
to the price of the drink should increase dissonance in the
regular-price condition, thus enhancing rather than attenuat-
ing the magnitude of the basic effect.
To accomplish the third goal of examining whether draw-
ing attention to price–quality beliefs affects the placebo
effect, we modified the procedure that we used in Experi-
ment 1 to strengthen expectancies before the puzzle task.
Recall that in Experiment 1, we manipulated the strength of
expectancies by having one group of participants respond to
the following statements before the puzzle task: “I feel that
SoBe is ‘very bad’ (1)/‘very good’ (7) at improving concen-
tration,” and “I feel that SoBe is ‘very bad’ (1)/‘very good’
(7) at improving mental performance.” In Experiment 2,
one group of participants rated similar statements, except
that their attention was also drawn to the price–efficacy link
by the following words that we added to the beginning of
the statements: “Given the price I was charged for SoBe,...
To summarize, Experiment 2 used a 2 (price: regular ver-
sus discounted) ×2 (price–efficacy salience: low versus
high) between-subjects design and a control condition.
Apart from modifying the task to incorporate the price–
efficacy salience factor, adding measures to serve as covari-
ates, and conducting Experiment 2 using computers, the
procedure paralleled the one we used in Experiment 1 (our
using a computerized rather than a paper-and-pencil task as
in Experiment 1 may account for some differences in the
performance levels across Experiments 1 and 2). A total of
193 undergraduate students participated in the study.
Results
The average number of puzzles solved across the various
conditions appears in Figure 3. An ANCOVA on the number
of puzzles solved revealed an interaction between price and
price–efficacy salience (F(1, 189) = 5.1, p< .05) and a main
effect of price (F(1, 189) = 3.1, p< .10). As in Experiment
1, prior consumption of SoBe was a relevant covariate that
did not interact with any other factor, and on average, par-
ticipants who had consumed SoBe before solved more puz-
zles than did those who had not (see Kirsch 1985).
The results in the low price–efficacy salience conditions
paralleled those in the low-expectancy-strength conditions
of Experiment 1. As in Experiment 1 and consistent with a
placebo effect of price discounts, the number of puzzles
solved was lower in the reduced-price condition (M = 6.8)
than in the regular-price condition (M = 8.3; F(1, 189) =
7.8, p< .01) and the control condition (M = 8.0; F(1, 189) =
5.0, p< .05). In addition, the number of puzzles solved was
not different between the regular-price and the control con-
ditions (F < 1). As in Experiment 1, we obtained only an
undesirable placebo effect of discounted prices and no
desirable placebo effect of the regular-priced drink. We dis-
cuss this finding subsequently.
The results in the high price–efficacy salience conditions
suggest that drawing attention to price–efficacy beliefs
weakens these beliefs, thus eliminating the placebo effect
(Ms = 8.2 and 8.0, respectively, in the reduced- and regular-
price conditions). In addition, that drawing attention to the
6.8
8.3 8.2 8.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Low Price–Efficacy
Salience
High Price–Efficacy
Salience
Discounted price Full price
Notes: The number of puzzles solved in the control condition = 8.0.
Figure 3
NUMBER OF PUZZLES SOLVED: EXPERIMENT 2
Placebo Effects of Marketing Actions 389
price–efficacy link eliminated the placebo effect is consis-
tent with the initial activation of expectancies being non-
conscious (see Strack et al. 1993). In further reducing the
viability of this alternative account, drawing attention to the
price of the drink not only eliminated the basic effect but
also resulted in means that were in the opposite direction of
what the cognitive dissonance and the distraction explana-
tions predict. Finally, note that both of these alternative
accounts rely on participants paying for their drink in our
experiments. However, although participants in the fitness
study we presented in the introduction did not pay for the
energy drink, we observed a placebo effect in that study.
We found further support for our conceptualization by
examining the expectancy measures that we collected in the
high price–efficacy salience conditions before the puzzle-
solving task (to operationalize the price–efficacy salience
factor). According to our conceptualization, drawing atten-
tion to the price–efficacy link should make the expectancies
in the reduced-price condition parallel to those in the
regular-price condition. Indeed, the expectancy ratings were
not different across the two price levels (Ms = 4.3; F < 1).
Why did we not observe a desirable placebo effect? A
question that arises from Experiments 1 and 2 is why there
was no desirable placebo effect of the regular-priced drink
and an undesirable placebo effect of the reduced-price
drink. Part of the answer to this question comes from exam-
ining participants’ expectancies before the puzzle-solving
task in the no-treatment control condition. Note that unlike
the treatment participants, the control participants were
given neither SoBe nor the accompanying instructions that
emphasized the price of the drink before they completed the
puzzle-solving task. Thus, beliefs about the price of SoBe
could not be activated in these participants. However, as we
show in Figure 1, other beliefs (e.g., about participants’ nat-
ural [unaided] ability in tasks such as solving word puzzles)
may have been more salient to control participants. Among
treatment participants, these self-efficacy beliefs may have
been less salient as a result of the external cues that were
presented (e.g., the drink’s price having been presented
more than once at the beginning of the experiment). In sum-
mary, a combination of two reasons may account for our
results: (1) Price-related (self-efficacy) beliefs may have
been the most salient to participants in the treatment (con-
trol) conditions as a result of the procedure we used, and
therefore (2) the resulting expectancies related to self-
efficacy beliefs in the control conditions may have been as
high as those related to price in the regular-price conditions.
To test these accounts, we first examined a measure related
to self-efficacy beliefs that was collected at the end of the
experiment; participants were asked to rate how good they
were in general at solving word puzzles on a scale that
ranged from 1 (“not at all good”) to 7 (“very good”). As we
expected, the mean rating on this measure was higher in the
control condition (M = 3.9) than in the regular-price condi-
tion (M = 3.2; F(1, 189) = 4.5, p< .05) and the reduced-
price condition (M = 3.3; F(1, 189) = 3.8, p< .05).
To explore this finding further, we conducted a separate
study to assess expectancies before the puzzle-solving task
(those expectancies were not administered in the control
conditions of the main experiments). We had 61 participants
from the same population as that in the main studies per-
form a task that was similar to those of Experiments 1 and
2. Participants were randomly assigned to one of three con-
ditions: two treatment conditions (regular price and reduced
price) and one no-treatment control condition. After partici-
pants received the initial instructions and paid for their
drinks (in the treatment conditions), they performed a filler
task and then rated their expected concentration and mental
performance during the subsequent puzzle-solving task on
two scales that we adapted from those administered in the
high-expectancy-strength conditions of Experiment 1. As
we expected, the mean expectations were not different in
the regular-price condition (M = 4.6) and control condition
(M = 4.4; F < 1) but were lower in the reduced-price condi-
tion (M = 3.3) than in the control condition (F(1, 58) = 22.4,
p< .0001). These findings suggest that a possible reason we
did not observe a desirable placebo effect of the drink in the
regular-price conditions of Experiments 1 and 2 was that
the expectations before the puzzle-solving task were not
different in these conditions from those in the no-treatment
control conditions.
Discussion
The results of Experiment 2 provide further support for a
placebo effect due to price discounts. We replicated the
findings of Experiment 1 in the low price–efficacy salience
conditions of Experiment 2; that is, the number of puzzles
participants solved was lower when they purchased the
product at a discounted price than when they purchased it at
the regular price. Furthermore, as in Experiment 1, we
observed an undesirable placebo effect in the reduced-price
condition but not a desirable placebo effect in the regular-
price condition, compared with the no-treatment control.
Experiment 2 sheds additional light on the underlying
process. Specifically, drawing attention to the price–
efficacy beliefs before participants solved puzzles reduced
the strength of response expectancies, eliminating the unde-
sirable placebo effect we observed in the reduced-price con-
dition. Furthermore, in line with prior work on noncon-
scious effects of biasing information, when participants’
attention was drawn to the relationship between price and
the product’s efficacy, the placebo effect did not occur. This
result supports the notion that the process leading to the
placebo phenomenon we observed occurs nonconsciously.
Finally, that the basic effect was attenuated rather than
enhanced reduces the viability of an alternative account
related to cognitive dissonance.
Experiment 2 also sheds light on a potential reason for
our observing only an undesirable placebo effect in the first
two experiments. First, self-efficacy ratings were higher in
the control condition than in the treatment conditions. We
discuss this finding further in the “General Discussion” sec-
tion. Second, a separate study revealed that expectancies in
the regular-price condition were not different from those in
the no-treatment control, which can account for the finding
that the number of puzzles solved was not different across
these conditions.
EXPERIMENT 3
Experiment 3 had several goals. One goal was to obtain
further support for the role of response expectancies in the
placebo effect of price discounts that we observed. Enhanc-
ing the antecedent intrinsic beliefs about the active sub-
stance or treatment through advertising claims, for example,
should strengthen expectancies and, therefore, the subse-
390 JOURNAL OF MARKETING RESEARCH, NOVEMBER 2005
4.2
5.8
7.4
10.1
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
Low Expectancy Strength High Expectancy Strength
Discounted price Full price
Notes: The number of puzzles solved in the control condition = 6.8.
Before solving the puzzles, participants in all treatment conditions rated
their drink-related expectancies as did those in the high-expectancy-
strength conditions of Experiment 1.
Figure 4
NUMBER OF PUZZLES SOLVED: EXPERIMENT 3
quent placebo effect. Accordingly, we manipulated not only
the price factor as in the previous two experiments but also
another marketing variable related to intrinsic beliefs about
the effectiveness of SoBe. We presented the following
instructions on the cover page: “Drinks such as SoBe have
been shown to improve mental functioning, resulting in
improved performance on tasks such as solving puzzles. In
fact, the Web site of SoBe includes references to over 50
scientific studies suggesting that consuming drinks like
SoBe can significantly improve mental functioning (in the
high-expectancy-strength conditions)/slightly improve men-
tal functioning (in the low-expectancy-strength condi-
tions).” We expected that adding this second factor related
to expectancy strength would enable us to observe a desir-
able placebo effect (in the regular-price, high-expectancy-
strength condition), which we had not observed in the pre-
vious two experiments.
Asecond goal of Experiment 3 was to examine the role
of other mediators of the observed effects. Note that in
Experiments 1 and 2, we found that response expectancies
that were measured before the puzzle-solving task medi-
ated the observed effects. However, we did not examine
how respondents felt during the puzzle-solving task.
Specifically, did respondents in the regular-price condi-
tions feel more motivated and alert during the task than
respondents in the reduced-price conditions? To accom-
plish the second goal, we administered measures after the
puzzle-solving task. We asked participants to rate how alert
and how motivated they felt during the puzzle-solving task
using seven-point scale items, with the last item reverse
scaled.
A third goal of Experiment 3 was to rule out an alterna-
tive account related to mood states. It is possible that partic-
ipants in the reduced-price conditions of Experiments 1 and
2 were in a more positive mood state (because they had
received the drink at a discount) than those in the regular-
price conditions. Research on mood state effects suggests
that positive mood states can impair cognitive capacity and
evoke less careful and substantive process styles than less
positive mood states (for a review, see Forgas 1995). This
might account for why participants in the reduced-price
conditions solved fewer puzzles. We tested the validity of
this account by collecting measures of mood states (in line
with the work of Watson, Clark, and Tellegen [1988])
before participants completed the puzzle-solving task.
To summarize, Experiment 3 used a 2 (price: regular ver-
sus discounted) ×2 (expectancy strength: low versus high)
between-subjects design and a control condition. Apart
from modifying the task to incorporate the expectancy-
strength factor, including expectancy measures before the
puzzle-solving task in all conditions (as in the high-
expectancy-strength conditions of Experiment 1) and
changing some of the puzzles for logistical reasons, the pro-
cedure paralleled those we used in Experiments 1 and 2.
Our changes to some of the puzzles may account for some
differences in the basic results across Experiments 1, 2, and
3. A total of 204 undergraduate students participated in the
study.
Results
The average number of puzzles solved in the various con-
ditions appears in Figure 4. An ANCOVA on the number of
puzzles solved revealed an interaction between price and
price–efficacy salience (F(1, 196) = 3.8, p< .05) and main
effects of price (F(1, 196) = 54.6, p< .0001) and
expectancy strength (F(1, 196) = 132.8, p< .0001). Note
that unlike Experiments 1 and 2, prior consumption of SoBe
was not a relevant covariate, probably because virtually all
participants indicated that they had consumed this drink
before.
The results in the low-expectancy-strength conditions
paralleled those in the corresponding conditions of Experi-
ment 1 and the low price–efficacy salience conditions of
Experiment 2. As in Experiments 1 and 2 and consistent
with a placebo effect, the number of puzzles solved was
lower in the reduced-price condition (M = 4.2) than in the
regular-price condition (M = 5.8; F(1, 196) = 15.1, p<
.0001) and the control condition (M = 6.8; F(1, 196) = 33.1,
p< .0001). In addition, the number of puzzles solved in the
regular price, low-expectancy-strength condition was lower
(M = 5.8) than in the control condition (M = 6.8;
F(1, 196) = 3.6, p< .10), indicating that presenting weak
claims about the efficacy of SoBe resulted in an undesirable
placebo effect even when the drink was sold at its regular
price.
The results in the high-expectancy-strength conditions
were also consistent with our conceptualization. Specifi-
cally, presenting strong claims about the efficacy of SoBe
to strengthen response expectancies increased the number
of puzzles solved in those conditions compared with the
low-expectancy-strength conditions. The number of puz-
zles solved in the reduced-price condition was again lower
(M = 7.4) than that in the regular-price condition (M =
10.1; F(1, 196) = 42.8, p<.0001), but the number was
higher in the regular-price condition than in the control
condition (F(1, 196) = 73.1, p<.0001), a desirable placebo
effect that we had not observed in our previous
experiments.
Placebo Effects of Marketing Actions 391
We found further support for our conceptualization by
examining the expectancy measures that we collected
across all conditions before the puzzle-solving task. In line
with our conceptualization, when the expectancy strength
was low, these ratings were lower in the reduced-price con-
dition (M = 3.2) than in the regular-price condition (M =
3.6; F(1, 196) = 3.4, p< .10) and the control condition (M =
4.2; F(1, 196) = 15.1, p< .0001). When the expectancy
strength was high, these ratings were again lower in the
reduced-price condition (M = 4.7) than in the regular-price
condition (M = 5.9; F(1, 196) = 25.4, p< .0001). Further-
more, a Sobel test revealed that these ratings mediated the
effects of the independent variables on the number of puz-
zles solved (z = 2.3, p< .05).
Alertness and motivation during the task. Recall that
after the puzzle-solving task, we asked participants how
alert and motivated they felt during the task. Separate Sobel
tests revealed that only the alertness measure mediated the
effects of the independent variables on the number of puz-
zles solved (z = 2.0, p< .05; in addition, the pattern of
results on this measure mirrored that of the number of puz-
zles solved). This finding is reasonable given that a major
claim of this drink is that it boosts alertness. The other
measure (which, incidentally, the drink does not claim to
influence) was not relevant as a mediator (p> .60).
Did mood mediate the observed effects? An alternative
account for our findings is that participants who bought the
drink at a discount were in a more positive mood state than
those who purchased it at the regular price. However, the
mood explanation predicts only a main effect of price and
not an interaction with a second factor, such as expectancy
strength, that we find in all three experiments. To reduce the
viability of the mood account further, we collected meas-
ures of mood states before the puzzle-solving task. Separate
ANCOVAs with the positive and negative mood measures
revealed no relevant effects.
Discussion
Experiment 3 documented not only an undesirable
placebo effect as in our previous experiments but also a
desirable placebo effect. Specifically, participants who pur-
chased the drink at its regular price and were presented with
strong advertising claims about the drink solved more puz-
zles than did participants in the control condition. Results of
this experiment also suggest that participants felt more alert
in the regular-price condition than in the reduced-price con-
ditions and that this mediated the placebo effect, consistent
with a claimed effect of the drink (i.e., helping people feel
more alert). Finally, Experiment 3 reduces the viability of
an alternative account related to mood states. First, as in the
previous two experiments, it is difficult to account for the
interactive pattern of results that we obtained in this experi-
ment with the mood explanation. Furthermore, the meas-
ures of mood states we obtained before the puzzle solving
task revealed that participants’ mood was not different
across the various conditions.
This study suggests the possibility that placebo effects of
marketing actions could create ethical dilemmas. For exam-
ple, a marketer may falsely claim that a product offers a
particular benefit, or similarly, a marketer could repackage
a product and significantly increase its price, suggesting
(explicitly or implicitly) that the higher cost of the cosmeti-
cally different product is justified by its greater efficacy.
Because of placebo effects of marketing actions, con-
sumers’ misplaced beliefs in such seemingly baseless
claims may paradoxically make those claims partly true;
indeed, given that unlike false statements, puffery is often
considered acceptable, even modest placebo effects may
make false claims legitimate.
GENERAL DISCUSSION
Marketing actions can have powerful perceptual effects
(e.g., Allison and Uhl 1964; McClure et al. 2004). For
example, cola can taste different when a person knows it is
“the real thing” (i.e., a Coke) from the same product when it
is mislabeled as a generic brand. More generally, it is
widely known that marketers can significantly influence
variables such as (perceptions of) consumption experiences
and purchase behavior. This article extends the scope of
effects that marketing actions are known to be capable of
evoking, showing that such actions can also influence the
actual efficacy of a marketed product. For example, we
found that participants who consumed an energy drink pur-
ported to improve mental acuity that they purchased at a
discounted price subsequently performed worse on a
puzzle-solving task than did equivalent participants who
purchased the same drink at its regular price.
We provide evidence that the effect of the documented
marketing actions is mediated by expectations. For exam-
ple, in Experiments 1, 2 and 3, we show that varying the
strength of response expectancies affects the magnitude of
the effect. Specifically, in Experiment 1, we show that per-
formance (after a participant consumed an energy drink he
or she purchased at a discount) was worse when the expec-
tations related to the drink’s efficacy were strengthened than
when they were not. In Experiment 2, drawing participants’
attention to their beliefs about the price–efficacy link weak-
ened their beliefs (some respondents presumably realized
that the beliefs may not be applicable to that situation) and,
thus, the magnitude of the effect. Experiment 3 revealed
that strengthening response expectancies by presenting
strong advertising claims enhanced the magnitude of the
basic effect. Across the three experiments, we also rule out
several alternative accounts, such as those relating to cogni-
tive dissonance, distraction, and mood states. Our findings
also suggest that the process by which expectations lead to
the observed effects occurs nonconsciously.
The effect we document is akin to placebo effects in
medicine. We contribute to the placebo literature by extend-
ing the types of features that are known to evoke such
effects from those that are inherent to the placebo (e.g.,
information about the placebo, the substance/treatment it
replaces, how a placebo is administered) to price, a feature
that is not inherent to the placebo. Given placebo
researchers’ interest in when placebo effects occur noncon-
sciously, it is also significant that we found that the process
by which expectations lead to the placebo effects occurs
nonconsciously.
Our findings extend what is known about the association
between price and quality in a significant way, showing that
price affects not only perceived quality but also actual qual-
ity (i.e., the actual efficacy of the product). Indeed, the
effect we found and its dependence on expectancies shed
light on a puzzling disparity between two conclusions of the
392 JOURNAL OF MARKETING RESEARCH, NOVEMBER 2005
large body of research on the relationship between price and
quality (see Bettman, John, and Scott 1986; Gerstner 1985;
John, Scott, and Bettman 1986). On the one hand, there is
vast empirical evidence that consumers often perceive
lower-priced products and services to be of lower quality,
especially if they have no simple alternative way to assess
quality (Rao and Monroe 1989). On the other hand, in gen-
eral, investigations of the relationship between price and
objective indications of quality, such as Consumer Reports
ratings, arrive at a different conclusion. For example, Riesz
(1979) examines the correspondence between prices and
indications of objective quality from Consumer Reports for
679 brands in 40 packaged food product classes over a 15-
year period. He concludes that the correlation was near
zero, and in cases such as frozen foods, it was even nega-
tive. More generally, such investigations conclude that the
empirical relationship between those two variables is weak
at best (Gerstner 1985). Why would consumers perceive the
relationship between price and quality to be significant
when, in general, it is not so? An explanation that is implied
by our research for this discrepancy may be a self-fulfilling
nature of consumer expectations. Such expectations may
lead lower-priced products to perform worse, regardless of
whether their objective indications of quality (research of
the type that Consumer Reports examines) are actually
worse. In other words, the well-known cliché that “one gets
what one pays for,” may have more merit than has been
believed. Exploring this is a worthwhile direction for fur-
ther research.
A related direction for further research is to delve deeper
into why we observed only undesirable placebo effects
related to discounted prices in Experiments 1 and 2. Our
findings in Experiment 2 suggest that one reason that could
account for our results is that people normally focus on self-
efficacy beliefs in tasks such as solving puzzles but partly
shift their focus away from their own abilities toward
beliefs about external stimuli when they are presented with
a performance enhancer such as SoBe. An interesting
research question is, Can being offered a performance
enhancer lower expectations about or the salience of a per-
son’s abilities and, thus, potentially lead to diminished per-
formance (particularly if beliefs about the efficacy of the
product are not as strong as the self-efficacy beliefs)?
We believe that another promising direction for further
research is to demonstrate additional situations in which
marketing actions have placebo effects. For example, repli-
cating our results with price promotions on medical prod-
ucts is another worthwhile direction for further research,
and there would be considerable implications for the mar-
keting of such products and for public policy. As a first step
in this direction, we conducted a small preliminary study.
We asked undergraduate marketing students to maintain
diaries of when they caught a cold over the course of a
semester and used an over-the-counter medication (i.e., a
prescription was not necessary) to treat the symptoms. At
the end of the semester, 29 students who had fallen ill dur-
ing the semester and who had bought national over-the-
counter brands indicated how effective the medication they
had bought was in treating their symptoms on a scale that
ranged from 1 (“not at all”) to 7 (“very”). In a separate
question, we asked them to indicate whether they had
bought the medication at its regular price or at a discounted
price. Consistent with a placebo effect of price discounts,
the 16 students who had bought their medication at a dis-
counted price rated the effectiveness of the medication to be
lower (M = 3.6) than the 13 students who had bought their
medication at the regular price (M = 5.5; F(1, 27) = 18.8,
p< .01).
More generally, it seems reasonable to speculate that
marketing decisions ranging from product features, such as
color and texture, to marketing-mix decisions, such as
advertising messages and distribution channels, may influ-
ence the physical effectiveness of the products to which
they are applied. If this is the case, the implications could
be immense. As an admittedly speculative possibility, if two
consumers purchase the same car but one does so at a sub-
stantial discount, the two consumers may drive differently.
A possible result could be that the consumer who purchases
the car at a discount is more accident-prone. Alternatively,
if two consumers purchase the same car but only one is
exposed to advertising messages that stress the safety bene-
fits of the car, the possible result could be that the consumer
exposed to the advertising drives differently than the other
consumer. Further research should carefully examine such
possibilities.
Another worthwhile research direction is to identify
additional moderators of the effect. Beyond the obvious
theoretical importance, this would also be significant from
apractical viewpoint because it might help reduce or even
eliminate undesirable placebo effects (as we did in Experi-
ment 2) in such cases as selling subsidized medications to
consumers. For example, will a delay between consump-
tion of a health-related product and subsequent engage-
ment in a task (see Nowlis, Mandel, and McCabe 2004)
diminish the magnitude of the placebo effect we docu-
ment? An investigation of such questions will help identify
boundary conditions, shed more light on the underlying
process, and explore the scope of placebo effects on mar-
keting actions.
REFERENCES
Adaval, Rashmi and Kent B. Monroe (2002), “Automatic Con-
struction and Use of Contextual Information for Product and
Price Evaluations,Journal of Consumer Research, 28 (March),
572–88.
Allison, Ralph I. and Kenneth P. Uhl (1964), “Influence of Beer
Brand Identification on Taste Perception,” Journal of Marketing
Research, 1 (August), 36–39.
Bettman, James R., Deborah R. John, and Carol A. Scott (1986),
“Covariation Assessment by Consumers,” Journal of Consumer
Research, 13 (December), 316–26.
Bienenfeld, Laura A., William H. Frishman, and Stephen P.
Glasser (1996), “The Placebo Effect in Cardiovascular Dis-
ease,” American Heart Journal, 132 (December), 1207–1221.
Fillmore, Mark and Muriel Vogel-Sprott (1992), “Expected Effect
of Caffeine on Motor Performance Predicts the Type of Response
to Placebo,” Psychopharmacology,106 (February), 209–214.
Fitzsimons, Gavan J., Tanya L. Chartrand, and Gráinne M. Fitz-
simons (2004), “Behavior Responses to Subliminal Brand
Exposure,” working paper, Fuqua School of Business, Duke
University.
Forgas, Joseph P. (1995), “Mood and Judgment: The Affect Infusion
Model (AIM),” Psychological Bulletin,117 (January), 39–66.
Gerstner, Eitan (1985), “Do Higher Prices Signal Higher Qual-
ity?” Journal of Marketing Research, 22 (May), 209–215.
Placebo Effects of Marketing Actions 393
Hahn, Robert A. (1997), “The Nocebo Phenomenon: Scope and
Foundations,” in Placebo Effect: An Interdisciplinary Explo-
ration, Anne Harrington, ed. Cambridge, MA: Harvard Univer-
sity Press, 56–76.
Huber, Joel and John McCann (1982), “The Impact of Inferential
Beliefs on Product Evaluations,Journal of Marketing
Research, 19 (August), 324–33.
John, Deborah R., Carol A. Scott, and James R. Bettman (1986),
“Sampling Data for Covariation Assessment: The Effect of Prior
Beliefs on Search Patterns,Journal of Consumer Research, 13
(June), 38–47.
Kaptchuk, Ted J., Peter Goldman, David A. Stone, and William B.
Stason (2000), “Do Medical Devices Have Enhanced Placebo
Effects, Journal of Clinical Epidemiology, 53 (August),
786–92.
Kirsch, Irving (1985), “Response Expectancy as a Determinant of
Experience and Behavior,American Psychologist, 40 (Novem-
ber), 1189–1202.
——— (1999), How Expectancies Shape Experience. Washing-
ton, DC: American Psychological Association.
——— (2004), “Conditioning, Expectancy, and the Placebo
Effect: Comment on Stewart-Williams and Podd,” Psychologi-
cal Bulletin, 130 (March), 341–44.
——— and Steven Jay Lynn (1999), “Automaticity in Clinical
Psychology,American Psychologist, 54 (July), 504–515.
——— and Guy Sapirstein (1999), “Listening to Prozac But Hear-
ing Placebo: A Meta-Analysis of Antidepressant Medications,”
in How Expectancies Shape Experience, Irving Kirsch, ed.
Washington, DC: American Psychological Association,
303–320.
——— and Lynne J. Weixel (1988), “Double-Blind Versus Decep-
tive Administration of a Placebo,” Behavioral Neuroscience,
102 (April), 319–23.
Levin, Irwin P. and Gary J. Gaeth (1988), “How Consumers Are
Affected by the Framing of Attribute Information Before and
After Consuming the Product,” Journal of Consumer Research,
15 (December), 374–78.
MacKinnon, David P., Chondra M. Lockwood, Jeanne M. Hoff-
man, Stephen G. West, and Virgil Sheets (2002), “A Compari-
son of Methods to Test Mediation and Other Intervening Vari-
able Effects,Psychological Methods, 7 (March), 83–104.
McClure, Samuel M., Jiam Li, Damon Tomlin, Kim S. Cypert,
Latané M. Montague, and P. Read Montague (2004), “Neural
Correlates of Behavioral Preference for Culturally Familiar
Drinks,” Neuron, 44 (October), 379–87.
Montgomery, Guy and Irving Kirsch (1996), “Mechanisms of
Placebo Pain Reduction: An Empirical Investigation,” Psycho-
logical Science, 7 (May), 174–76.
Nowlis, Stephen M., Naomi Mandel, and Deborah B. McCabe
(2004), “The Effect of a Delay Between Choice and Consump-
tion on Consumption Enjoyment,” Journal of Consumer
Research, 31 (December), 502–510.
Rao, Akshay R. and Kent B. Monroe (1988), “The Moderating
Effect of Prior Knowledge on Cue Utilization in Product Evalu-
ations,” Journal of Consumer Research, 15 (September),
253–64.
——— and ——— (1989), “The Effect of Price, Brand Name,
and Store Name on Buyers’ Perceptions of Product Quality: An
Integrative Review,Journal of Marketing Research, 26
(August), 351–57.
Rescorla, Robert A. (1988), “Pavlovian Conditioning: It’s Not What
You Think It Is,” American Psychologist,43(March), 151–60.
Riesz, Peter C. (1979), “Price-Quality Correlations for Packaged
Food Products,Journal of Consumer Affairs, 13 (Winter),
236–47.
Rogers, James T. (1985), Dictionary of Clichés. New York: Facts
on File Publications.
Stewart-Williams Steve and John Podd (2004), “The Placebo
Effect: Dissolving the Expectancy Versus Conditioning
Debate,” Psychological Bulletin, 130 (March), 324–40.
Strack, Fritz, Norbert Schwarz, Herbert Bless, Almut Kübler, and
Michaela Wänke (1993), “Awareness of the Influence as a
Determinant of Assimilation Versus Contrast,” European Jour-
nal of Social Psychology, 23 (January–February), 53–62.
Wager Tor D., James K. Rilling, Edward E. Smith, Alex Sokolik,
Kenneth L. Casey, Richard J. Davidson, Stephen M. Kosslyn,
Robert M. Rose, and Jonathan D. Cohen (2004), “Placebo-
Induced Changes in fMRI in the Anticipation and Experience of
Pain,Science, 303 (February), 1162–67.
Watson, David, Lee Anna Clark, and Auke Tellegen (1988),
“Development and Validation of Brief Measures of Positive and
Negative Affect: The PANAS Scales,” Journal of Personality
and Social Psychology, 54 (June), 1063–1070.
... In both cases, these brand enhancement benefits emerged despite no functional differences between products employed, akin to a marketing placebo effect. Marketing actions including price promotions (Shiv et al. 2005), logo placement (Garvey et al. 2016;Park and Roedder-John 2014), and anticounterfeiting messaging (Amar et al. 2018) have been found to drive placebo-like effects. ...
... Fundamental to the emergence of these effects are associations of product efficacythat is, the belief that the product has the capacity to alter performance (Amar et al. 2018;Garvey et al. 2016;Shiv et al. 2005). ...
... Importantly, exposure to an Under Armour advertisement should increase the attention to any worn branded gear, with an accompanying activation of Under Armour performance expectancies. In turn, consumer performance should be enhanced due to this heightened activation of performance expectancies during product use, consistent with extant research into marketing placebo effects (Amar et al. 2018;Garvey et al. 2016;Shiv et al. 2005). Hence, we anticipate that exposure to novel brand advertisement congruent with a performance branded product in use will activate the psychological presence of the brand and its performance expectancies, thereby eliciting brand enhancement benefits in the form of improved performance on a relevant task (e.g., increased work-out performance in the case of Under Armour). ...
... As argued earlier, most traditional marketing methods are ineffective due to a lack of understanding of consumers' mental processes before, during, and after being exposed to marketing stimuli (Chaney et al., 2017;Çora, 2019). Accordingly, Shiv et al. (2005) suggest that a neuroscientific approach can address these problems by (a) providing neural-level evidence of consumer behaviours towards marketing stimuli, (b) reconceptualising the operationalisation of consumers' underlying processes for successful marketing campaigns, (c) providing confirmatory evidence about behavioural generalisation through neural exploration, (d) providing more continuous stable data for longitudinal reanalysis and usage, and (e) providing an avenue for robust and rigorous methodological approaches to studying consumer behaviour. ...
... This approach eradicates systematic and methodological limitations common with most traditional means of exploring consumer responses by providing a neural understanding of how consumers will behave when deciding about consumption and what will make them behave or change behaviour in the same, similar, or different circumstances (Agarwal & Xavier, 2015;Shiv et al., 2005). The common variation found in traditional marketing approaches indicates that most obtained results are not ecologically generalisable due to disparity in consumer behavioural intention and actual behaviour. ...
Chapter
“Sponsorship: Practices and Benefits in Emerging Markets” details the strategic sponsorship activities and mechanisms that can be implemented to communicate with customers and other stakeholders.
... As argued earlier, most traditional marketing methods are ineffective due to a lack of understanding of consumers' mental processes before, during, and after being exposed to marketing stimuli (Chaney et al., 2017;Çora, 2019). Accordingly, Shiv et al. (2005) suggest that a neuroscientific approach can address these problems by (a) providing neural-level evidence of consumer behaviours towards marketing stimuli, (b) reconceptualising the operationalisation of consumers' underlying processes for successful marketing campaigns, (c) providing confirmatory evidence about behavioural generalisation through neural exploration, (d) providing more continuous stable data for longitudinal reanalysis and usage, and (e) providing an avenue for robust and rigorous methodological approaches to studying consumer behaviour. ...
... This approach eradicates systematic and methodological limitations common with most traditional means of exploring consumer responses by providing a neural understanding of how consumers will behave when deciding about consumption and what will make them behave or change behaviour in the same, similar, or different circumstances (Agarwal & Xavier, 2015;Shiv et al., 2005). The common variation found in traditional marketing approaches indicates that most obtained results are not ecologically generalisable due to disparity in consumer behavioural intention and actual behaviour. ...
... As argued earlier, most traditional marketing methods are ineffective due to a lack of understanding of consumers' mental processes before, during, and after being exposed to marketing stimuli (Chaney et al., 2017;Çora, 2019). Accordingly, Shiv et al. (2005) suggest that a neuroscientific approach can address these problems by (a) providing neural-level evidence of consumer behaviours towards marketing stimuli, (b) reconceptualising the operationalisation of consumers' underlying processes for successful marketing campaigns, (c) providing confirmatory evidence about behavioural generalisation through neural exploration, (d) providing more continuous stable data for longitudinal reanalysis and usage, and (e) providing an avenue for robust and rigorous methodological approaches to studying consumer behaviour. ...
... This approach eradicates systematic and methodological limitations common with most traditional means of exploring consumer responses by providing a neural understanding of how consumers will behave when deciding about consumption and what will make them behave or change behaviour in the same, similar, or different circumstances (Agarwal & Xavier, 2015;Shiv et al., 2005). The common variation found in traditional marketing approaches indicates that most obtained results are not ecologically generalisable due to disparity in consumer behavioural intention and actual behaviour. ...
... If the price rebate is given on the same product, the consumers buy more of that product. Significant focus was drawn to the relevance of marketing researchers' price promotions --- (Chandon et al., 2000;Chen et al., 1998;Shiv et al., 2005). The conventional price promotion provides safe or fixed price deals such as discounts and offers the most common promotional strategies used by retailers. ...
Article
Full-text available
The study explores how price promotions such as probabilistic and conventional fixed price promotion impact the pain of paying within the individual differences of tightwad consumers, unconflicted consumers and spendthrift consumers. An online experiment was done to see how the pain of paying might impact price promotion by simulating a supermarket shopping environment. We find that probabilistic pricing promotions successfully stimulate demand, as they minimise the pain of paying more effectively, compared with fixed price promotions. For tightwad consumers who are much more susceptible to pain, probabilistic price promotion has demonstrated greater efficacy in reducing the pain of paying compared to fixed price promotion. In the case of unconflicted consumers, the level of pain of paying showed a difference between fixed and probabilistic price promotion, but at a lower rate as compared to tightwad consumers. However, the probabilistic price promotion did not demonstrate any efficacy in reducing the pain of paying among spendthrift consumers, who are less susceptible to the pain of paying, than the fixed price promotion in contrast to tightwad consumers. The results elucidate a reduction in the pain of payment persists to be a path by which price promotions build customer value. Keywords: Fixed price promotion, Probabilistic price promotion, Pain of Paying, Spendthrift consumer, Tightwad consumer, Unconflicted consumer
... Notes. 1 Manipulated, 0 = no price increase, 1 = price increase, 2 measured on a sevenpoint scale, 3 manipulated, 0 = low CSR, 1 = high CSR SIPILÄ ET AL. | 5 fairness (see Aguilera et al., 2007 for a similar proposition), and that this signal extends to the sphere of expected price fairness through a "fairness halo." Such halo effects, in which a cue from one context affects perceptions in a related context, have been reported in various settings (e.g., Schuldt et al., 2012;Shiv et al., 2005), including a company's perceived CSR engagement (Chernev & Blair, 2015;Schuldt et al., 2012). For example, CSR claims focused on food manufacturing are associated with perceptions of the food's healthiness (Wei et al., 2018). ...
Article
Full-text available
Consumers care about the fairness of companies both in terms of corporate social responsibility (CSR) engagement and the fairness of prices. However, the interplay between these domains is not yet well understood. Therefore, this study examines how consumers’ perceptions of CSR engagement affect their perceived price fairness following a price increase. Drawing on cue‐utilization and expectancy disconfirmation theory, the authors propose that perceived CSR engagement exacerbates the negative effect of a price increase on perceived price fairness, because perceived CSR engagement increases consumers’ price fairness expectations which are violated through price increases. These propositions are tested in three experimental studies with samples consisting of approximately 3000 customers of a global furniture manufacturer and retailer (Study 1), as well as participants acquired through a self‐administered online consumer panel (Study 2) and Prolific (Study 3). The experiments yield support for the hypothesized effects and reveal CSR skepticism as a critical boundary condition. The findings extend existing literature on the effects of perceived CSR engagement and pricing on consumer reactions by examining the role of perceived CSR engagement in shaping consumer reactions to price increases and by establishing consumers’ price fairness expectations as the central psychological mechanism.
... Research has shown that even clinicians often forgo such evidence-based judgments, relying instead on pre-existing beliefs (Kim and Ahn 2002;De Kwaadsteniet and Hagmayer 2018). This is consistent with research showing that even when consumers physically experience products, they make judgments and decisions based on general knowledge such as categories, stereotypes, price, and brands (e.g., Allison and Uhl 1964;Gerstner 1985;Shiv, Carmon, and Ariely 2005;Lee, Frederick, and Ariely 2006;Berger, Draganska, and Simonson 2007). 4 Lacking sufficient medical knowledge and expertise to make evidence-based health judgments (Kahn et al. 1997;Keller and Lehmann 2008), consumers rely instead on lay beliefs when making decisions about medical products and treatments (Bolton et al. 2008;Wang, Keh, and Bolton 2010;Kramer et al. 2012;Ilyuk, Block, and Faro 2014;Scott, Rozin, and Small 2020), responding to health-related messages (Raghubir and Menon 1998;Agrawal, Menon, and Aaker 2007;Zhang, Mathur, and Block 2021), or supporting health-related causes (Kogut and Ritov 2005;Puntoni, Sweldens, and Tavassoli 2011). ...
Preprint
Full-text available
How do people assess the severity of health problems? How do they decide whether these merit medical attention? We investigate how beliefs about psychological and physical causes of medical signs affect their perceived severity. Three studies showed that people perceive objective and observable evidence of illnesses (i.e., medical signs) as less severe if they originate from psychological rather than physical causes. For instance, participants rated the same cough as less harsh and scratchy when they believe it was caused by anxiety rather than by drinking contaminated tap water. As a result, participants were less likely to recommend care for medical signs with psychological origins, less likely to prioritize their care among multiple health problems, and reluctant to financially support scientific research for their cure.
Chapter
Full-text available
La principal aportación de esta investigación se encuentra en la preferencia por productos ecológicos versus los contaminantes después de presentar su precio para tomar una decisión al momento de comprar, que es el punto central de este estudio. La evidencia encontrada demuestra que, ante una elección que incluye beneficios al medio ambiente, las personas en su primera elección durante sus compras (que podrían ser las que efectúan cotidianamente cada que compran), no necesariamente se deciden por los beneficios ambientales; al contrario, optan por productos que tienen consecuencias al medio ambiente. Sin embargo, la sensibilidad al precio y el umbral de comparación de precios bajos frente a precios altos, incide directamente en el cambio en su preferencia por productos ecológicos, con un menor precio y por consiguiente un bajo costo en sus finanzas personales.
Article
Full-text available
If there is still excess supply for a good at zero price, buyers can receive payment as an incentive to balance supply and demand. Overall, negative prices do not necessarily lead to losses if they serve only as a temporary solution for companies that generally generate high sales. Wenn für ein Gut auch bei einem Preis von Null noch ein Angebotsüberhang besteht, können die Käufer als Anreiz eine Zahlung erhalten, damit Angebot und Nachfrage zum Ausgleich kommen. Negative Preise führen insgesamt nicht zwangsläufig zu Verlusten, wenn sie den Unternehmen, die generell hohe Umsätze generieren, lediglich als temporäre Lösung dienen.
Article
Purpose – The purpose of this research is to determine the way functional magnetic resonance imaging (fMRI), as a research tool in neuromarketing, is used and to analyze important elements of conducting such research, as well as defining all relevant terms regarding neuromarketing and consumer neuroscience. Design/Methodology/Approach – The research study focused on conducting an fMRI experiment, using the observation method, survey, and interview methods. Findings and implications – By analyzing all steps in the fMRI experiment procedure and determining how to conduct fMRI research, obstacles in this kind of research were identified to establish what needs to be overcome for proceeding with further studies. Limitations – Numerous limitations included lack of adequate equipment and software, huge administrative barriers, and finding experts capable of conducting this kind of research. Originality – This research study combined the topics of sensory marketing, neuromarketing, and social marketing. Furthermore, it contributes to clearing the path for this kind of research approach in the future, representing a new trend that is here to stay.
Article
Full-text available
A Monte Carlo study compared 14 methods to test the statistical significance of the intervening variable effect. An intervening variable (mediator) transmits the effect of an independent variable to a dependent variable. The commonly used R. M. Baron and D. A. Kenny (1986) approach has low statistical power. Two methods based on the distribution of the product and 2 difference-in-coefficients methods have the most accurate Type I error rates and greatest statistical power except in 1 important case in which Type I error rates are too high. The best balance of Type I error and statistical power across all cases is the test of the joint significance of the two effects comprising the intervening variable effect.
Article
Full-text available
A consumer choosing a product must often wait before consuming it. In this article, we consider the consequences of waiting on consumption enjoyment. We propose that the effect of a delay on consumption enjoyment depends on both the negative utility of the wait itself and on the positive utility of anticipating a pleasant consumption experience. These factors exert different degrees of influence, depending on characteristics of the decision task. The results of three studies suggest that a delay increases consumption enjoyment for pleasurable products when actual consumption occurs, but decreases enjoyment for imagined consumption. Furthermore, the vividness of the awaited product moderates these effects.
Article
Though most observers acknowledge that consumers make inferences among product attributes, the influence of such inferences on product evaluation is much less clear. Study respondents evaluated products for which information on one of two attributes was systematically omitted. A general model is built to estimate the directional effect of inferences on product evaluation. The effect of inferences to a missing attribute is statistically significant and in the expected direction. In one case, the marginal value of the remaining attribute (price) reverses in sign because of an inference. Thus, inferences are theoretically important and a potentially troublesome issue in the modeling and measurement of consumer choice processes.
Article
As a company tries to find the factors accounting for strong and weak markets, typical consumer explanations for both tend to be in terms of the physical attributes of the product. Carling Brewing Company used a relatively inexpensive experiment to help dichotomize contributing influences as being either product or marketing oriented and, also, to indicate the magnitude of the marketing influence for various brands. The experiment involved the use of groups of beer drinkers that tasted (drank) and rated beer from nude bottles and from labeled bottles.
Article
Current thinking about Pavlovian conditioning differs substantially from that of 20 years ago. Yet the changes that have taken place remain poorly appreciated by psychologists generally. Traditional descriptions of conditioning as the acquired ability of one stimulus to evoke the original response to another because of their pairing are shown to be inadequate. They fail to characterize adequately the circumstances producing learning, the content of that learning, or the manner in which that learning influences performance. Instead, conditioning is now described as the learning of relations among events so as to allow the organism to represent its environment. Within this framework, the study of Pavlovian conditioning continues to be an intellectually active area, full of new discoveries and information relevant to other areas of psychology.
Article
The author reports some empirical results on the strength of the quality-price relation. For many products, the relation between quality and price appears to be very weak; hence, for many products, higher prices appear to be poor signals of higher quality.
Article
Identical experimental pain stimuli were applied to the right and left index fingers of 56 university students, following the application of a placebo in the guise of a topical anesthetic to one of the fingers The pain stimuli were administered simultaneously to treated and untreated fingers for half of the subjects and sequentially for the others Significant and equivalent reductions in pain were reported in both conditions as a function of placebo administration These data indicate that reductions of experimental pain produced by placebos presented in the guise of local anesthetics are not mediated by such global mechanisms as anxiety reduction or the release of endogenous opioids
Article
The authors integrate previous research that has investigated experimentally the influence of price, brand name, and/or store name on buyers' evaluations of prod¬uct quality. The meta-analysis suggests that, for consumer products, the relation¬ships between price and perceived quality and between brand name and perceived quality are positive and statistically significant. However, the positive effect of store name on perceived quality is small and not statistically significant. Further, the type of experimental design and the strength of the price manipulation are shown to significantly influence the observed effect of price on perceived quality.