ArticlePDF Available

Abstract and Figures

Despite assertions that coolness sells products, little is known about what leads consumers to perceive brands as cool. This research uses an experimental approach to examine the empirical relationship between consumers’ inferences of autonomy and perceived coolness. Six studies find that behaviors expressing autonomy increase perceived coolness, but only when the autonomy seems appropriate. Autonomy seems appropriate, and hence increases perceptions of coolness, when a behavior diverges from a norm considered unnecessary or illegitimate, the autonomy is bounded (i.e., deviations are small or occasional rather than large or perpetual), and when the consumer views social norms as being overly repressive. A final experiment further supports the connection between autonomy and coolness and illustrates that coolness is distinct from liking by showing that whether or not a consumer has a goal to express autonomy moderates preference for cool brands.
Content may be subject to copyright.
Journal of Consumer Research, Inc.
What Makes Things Cool? How Autonomy Influences Perceived Coolness
Author(s): Caleb Warren and Margaret C. Campbell
Source:
Journal of Consumer Research,
Vol. 41, No. 2 (August 2014), pp. 543-563
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/676680 .
Accessed: 19/08/2014 16:54
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
The University of Chicago Press and Journal of Consumer Research, Inc. are collaborating with JSTOR to
digitize, preserve and extend access to Journal of Consumer Research.
http://www.jstor.org
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
543
2014 by JOURNAL OF CONSUMER RESEARCH, Inc. Vol. 41 August 2014
All rights reserved. 0093-5301/2014/4102-0021$10.00. DOI: 10.1086/676680
What Makes Things Cool? How Autonomy
Influences Perceived Coolness
CALEB WARREN
MARGARET C. CAMPBELL
Despite assertions that coolness sells products, little is known about what leads
consumers to perceive brands as cool. This research uses an experimental ap-
proach to examine the empirical relationship between consumers’ inferences of
autonomy and perceived coolness. Six studies find that behaviors expressing au-
tonomy increase perceived coolness, but only when the autonomy seems appro-
priate. Autonomy seems appropriate, and hence increasesperceptions ofcoolness,
when a behavior diverges from a norm considered unnecessary or illegitimate,
when the autonomy is bounded (i.e., deviations are small or occasional ratherthan
large or perpetual), and when the consumer views social norms as being overly
repressive. A final experiment further supports the connection between autonomy
and coolness and illustrates that coolness is distinct from liking by showing that
whether a consumer has a goal to express autonomy moderates preference for
cool brands.
The marketplace values cool brands. A cool image
helped solidify Harley Davidson’s status as an iconic
brand (Holt 2004), rejuvenate sales of Pabst Blue Ribbon
(Walker 2003), and vault Apple into the ranking of “the best
global brand” of 2013 (Interbrand 2014). Coolness excites
consumers, adds symbolic currency to products, and drives
consumer trends (Frank 1997; Gladwell 1997; Heath and
Potter 2004; Leland 2004). Kerner and Pressman (2007, xii)
write, “our society is consumed with the trappings of cool.
. . . All across the psychographic spectrum everyone wants
it, even if they can’t define what ‘cool’ actually is.” This
quote synthesizes two interesting aspects of coolness. First,
coolness is often desired, both by consumers and by mar-
Caleb Warren (cwarren@mays.tamu.edu) is an assistant professor of
marketing, Mays Business School, Texas A&M University, 4112 TAMU,
College Station, TX 77843-4112. Margaret C. Campbell (meg.campbell@colorado
.edu) is a professor of marketing, Leeds School of Business, University of
Colorado, Boulder, UCB 419, Boulder, CO 80309. This article is based on
the first author’s dissertation and supported by a grant from the Center for
Research in Marketing and Services (CERMES) at Bocconi University.
The authors thank committee members Kent Grayson, John Lynch, Peter
McGraw, and Page Moreau for their invaluable advice and Martin Schreier
for helpful comments on a previous draft. Correspondence should be ad-
dressed to cwarren@mays.tamu.edu.
Ann McGill served as editor and James Burroughs served as associate
editor for this article.
Electronically published May 15, 2014
keters. Second, it is not clear exactly what, in addition to
being desirable, makes things cool.
In six experiments we demonstrate that consumers per-
ceive cultural objects, including brands and people, to be
cool when they infer that the object is autonomous (i.e.,
pursues its own motivations irrespective of the norms and
expectations of others) in an appropriate way. Consumers
infer that a brand (or person) is autonomous when its be-
haviors diverge from the norm. Autonomy seems appro-
priate, and thus leads to perceptions of coolness, when a
divergent behavior is perceived to be at least as effective
or valuable as the normative behavior, it diverges from a
norm that is not considered legitimate, and divergence is
bounded rather than extreme. Moreover, consumers with
countercultural values, who are more critical of societal in-
stitutions and more likely to consider norm divergence ap-
propriate than those without countercultural values, tend to
perceive a relatively higher level of autonomy cool. We
further show that although cool brands are typically desired,
coolness and desirability are not the same thing, as con-
sumers prefer cool brands only when they want to stand out
rather than fit in.
WHAT IS COOL?
Although researchers do not agree on a specific definition
of coolness (Dar-Nimrod et al. 2012; Kerner and Pressman
2007), a canvas of the literature reveals agreement on four
defining properties. One, coolness is socially constructed.
Cool is not an inherent feature of an object or person but
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
544 JOURNAL OF CONSUMER RESEARCH
is a perception or an attribution bestowed by an audience
(Belk, Tian, and Paavola 2010; Connor 1995; Gurrieri 2009;
Leland 2004). In this sense, coolness is similar to socially
constructed traits, like popularity or status (Hollander 1958);
objects and people are cool only to the extent that others
consider them cool.
Two, coolness is subjective and dynamic. The things that
consumers consider cool change both over time and across
consumers (Danesi 1994; MacAdams 2001; O’Donnell and
Wardlow 2000). Despite the subjective nature of coolness,
consumers have little difficulty recognizing coolness when
they see it (Belk et al. 2010; Leland 2004). Moreover, cool-
ness ranges on a continuum, and consumers with similar
backgrounds and interests tend to agree on what is more
and less cool within a particular social context (Leland
2004). Thus, coolness is best operationalized using a con-
sensual assessment technique (Amabile 1982) by asking a
group of consumers the extent to which they perceive some-
thing or someone to be cool or uncool. Amabile (1982)
originally developed the consensual assessment technique
to measure creativity, which, like coolness, subjectively de-
pends on the perceptions of an audience. Like judgments
of creativity, perceptions of coolness are continuous and
contextual. For example, clothing at WalMart seems more
or less cool relative to other clothes in the store, not relative
to designs at a fashion show. Likewise, consumers assess
coolness differently when evaluating the shoes people wear
in an office versus a nightclub.
Three, coolness is perceived to be a positive quality (Bird
and Tapp 2008; Heath and Potter 2004; Pountain and Robins
2000). The few quantitative empirical studies on the topic
confirm that cool people tend to possess personality traits
considered desirable by the audience evaluating coolness
(Dar-Nimrod et al. 2012; Rodkin et al. 2006). Qualitative
studies similarly describe coolness as having a positive va-
lence, noting that consumers even sometimes use the word
“cool” as a synonym for “I like it” (Belk et al. 2010).
Four, although coolness is a positive trait, coolness re-
quires more than the mere perception that something is pos-
itive or desirable (Leland 2004; MacAdams 2001). Pountain
and Robins (2000, 32) write, “Cool is not merely another
way of saying good. It comes with baggage.” Consumers
perceive some quality that sets cool things apart from other
things that they merely like or evaluate positively. However,
the literature is not clear as to what this additional quality
is. Our goal is to identify this quality and empirically val-
idate its influence on perceptions of coolness of cultural
objects, including brands and people.
WHAT DISTINGUISHES COOL
FROM GOOD?
What, in addition to being liked, makes things cool? De-
spite the absence of strong causal data, the literature has
made many claims about what leads to the perception that
a person or brand is cool. Some argue that people (and
brands) become cool by mimicking the behavior of other
cool people (Gladwell 1997) or by conforming to the norms,
standards, and ideals of a particular subculture or clique
(Danesi 1994; O’Donnell and Wardlow 2000; Thornton
1996). Others argue that a rebellious attitude (Pountain and
Robins 2000), nonconformity (Frank 1997; Heath and Potter
2004), individualism (Hebdige and Potter 2008), defiance
(MacAdams 2001), or an unwillingness to follow trends
(Connor 1995) leads to perceptions of coolness. Still others
point to factors like sexual permissiveness, hedonism, and
detachment as potential antecedents (Bird and Tapp 2008;
Connor 1995; Leland 2004).
Interestingly, a common theme is that all of these factors
are related to the extent to which one shows, or does not
show, autonomy. Autonomy refers to a willingness to pursue
one’s own course irrespective of the norms, beliefs, and
expectations of others. Conformity, mimicry, and belonging
suggest a lack of autonomy because they require following
or conceding to the will of others. Conversely, unconven-
tionality, rebellion, individuality, authenticity, and indepen-
dence show autonomy because they require doing one’s own
thing and going against what others expect or prescribe.
Because they similarly require parting from the normative
path, sexual permissiveness, hedonism, and detachment also
indirectly suggest autonomy. Some thus suggest that cool-
ness comes from factors associated with low autonomy
(Danesi 1994; Gladwell 1997; Thornton 1996), whereas oth-
ers point to antecedents associated with high autonomy
(Heath and Potter 2004; MacAdams 2001; Pountain and
Robins 2000). Indeed, Belk and colleagues (2010, 202) point
out that “there is a tension between standing-out cool and
fitting-in cool”; however, they do not elaborate on this ten-
sion, nor do they discuss when perceptions of coolness will
be more influenced by standing out and when they will be
more influenced by fitting in. In sum, the literature suggests
some relationship between autonomy and coolness, but it
does not provide a clear understanding of the nature of this
relationship.
Drawing from this work, we propose that the extra quality
that differentiates something from merely being liked to
being perceived as cool is inferred autonomy. However,
given the tension between whether coolness comes from
conforming or diverging, we hypothesize that there are con-
ditions that moderate how autonomy influences perceptions
of coolness. Specifically, autonomy will increase perceptions
of coolness only when it seems contextually appropriate.
Thus, we propose the following definition of coolness: cool-
ness is a subjective and dynamic, socially constructed pos-
itive trait attributed to cultural objects (people, brands, prod-
ucts, trends, etc.) inferred to be appropriately autonomous.
In the next section we develop this conceptualization, with
specific attention to what makes autonomy seem appropriate.
WHEN IS EXPRESSING
AUTONOMY COOL?
Autonomy refers to the extent to which the person or
brand follows its own character or motivations irrespective
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 545
of the norms, beliefs, and expectations of others. We propose
that this willingness to do one’s own thing, or be one’s own
person, regardless of the norm, is the extra quality that de-
fines coolness. Because autonomy is a function of internal
motivations, it cannot be directly observed but must be in-
ferred from behavior (Jones and Davis 1965). Typically con-
sumers infer autonomy from behaviors that diverge from a
norm (Bellezza, Gino, and Keinan 2014). For example, Har-
ley Davidson gained an autonomous image by associating
with “outlaw” biker gangs famous for rebelling against rules
and conventions associated with middle-class society (Holt
2004). Conversely, consumers infer a lack of autonomy from
behaviors that conform to the norm or imitate others. For
example, the “Mac vs. PC” ad campaign portrays the PC
character as lacking autonomy by having PC wear conven-
tional office attire and perform tasks that consumers asso-
ciate with mainstream corporate jobs (e.g., spreadsheets and
pie charts). Mac, in contrast, shows autonomy by diverging
from conventional office norms by wearing causal clothes
and tennis shoes.
The literature shows that norm divergence is often dis-
liked. Norms develop in order to help coordinate interactions
among people with different personal interests, and they
serve as standards for appropriate behavior to which people
are motivated and expected to conform (Cialdini and Trost
1998; Schultz et al. 2007). Norm violations are often pun-
ished (Fehr and Fischbacher 2003), and norm violators risk
creating negative impressions (Aronson 2008; Schachter
1951), losing social standing (Hollander 1958), and poten-
tially even becoming estranged (Hogan 2001). From this
perspective, it seems curious that autonomy based on di-
verging from the norm might make something seem cooler.
However, although divergent behaviors are often perceived
to be inefficient, harmful, or otherwise inappropriate, there
are some cases in which expressing autonomy leads to more
favorable impressions (Ariely and Levav 2000; Bellezza et
al. 2014).
We hypothesize that the effect of autonomy on percep-
tions of coolness depends on whether autonomy seems con-
textually appropriate. Autonomy should increase perceived
coolness when it seems appropriate but not when it seems
inappropriate. In order to understand when autonomy will
seem appropriate and, hence, increase coolness, there are
four important considerations: (1) whether a brand diverges
from a descriptive or an injunctive norm, (2) the perceived
legitimacy of the injunctive norm from which a brand di-
verges, (3) the extent to which a brand diverges from in-
junctive and descriptive norms, and (4) the extent to which
the observer or audience values autonomy.
Distinguishing between Descriptive and
Injunctive Norms
Understanding whether a divergent behavior seems ap-
propriate, and hence its likely effect on perceived coolness,
depends on the nature of the norm from which the behavior
diverges. Divergence can be from either a descriptive norm
(i.e., what most people typically do in a particular context)
or an injunctive norm (i.e., a cultural ideal or a rule that
people are expected to follow; Cialdini, Reno, and Kallgren
1990). People generally conform to both descriptive and
injunctive norms but for different reasons. People typically
conform to descriptive norms, especially when the optimal
course of action is uncertain or unclear, because the norm
provides an example of a behavior known to be generally
effective or valuable in a given context (Cialdini and Trost
1998). In contrast, people typically conform to injunctive
norms, which provide a cultural ideal or social obligation,
in order to build or maintain relationships or social esteem
(Cialdini and Trost 1998). Because there are different rea-
sons for conforming to (or diverging from) descriptive
norms than injunctive norms, the factors that influence
whether a divergent behavior seems appropriate differ for
descriptive and injunctive norms. In this section we discuss
when divergence from a descriptive norm seems appropriate.
In the next section, we discuss when divergence from an
injunctive norm seems appropriate.
Descriptive norms offer an example of how to behave
effectively. However, in many contexts, alternative behav-
iors exist that may be equally (or even more) effective or
valuable. Creative ideas and products, for example, diverge
from the norm in a way that seems both novel and functional
(Amabile 1982; Burroughs and Mick 2004; Moreau and
Dahl 2005). Creative advertisements similarly diverge from
the norm in a way that helps fulfill the strategic goals of
the advertising campaign (Koslow, Sasser, and Riordan
2003). Deviance regulation theory suggests that divergent
behavior that seems better than the norm can help people
reach their identity goals (Blanton and Christie 2003). We
thus propose that autonomy will seem appropriate in a par-
ticular social context when a behavior diverging from a
descriptive norm seems at least as effective or valuable as
the normative behavior. For example, an unusual water bot-
tle design would seem appropriate if it were to maintain
functional utility (e.g., holds and dispenses water, stands
upright) and avoid threatening consumers’ identity goals
(e.g., they would not be embarrassed to use it). Because the
creativity literature demonstrates that negative or valueless
divergence is not considered appropriate (Amabile 1982;
Burroughs and Mick 2004), our studies control for value
but do not directly test whether a lack of value detracts from
coolness.
Norm Legitimacy Moderates the Appropriateness
of Divergence from an Injunctive Norm
Injunctive norms prescribe cultural ideals by specifying
how people should behave in a given context. However, a
range of norms exists both within and across cultures (Cial-
dini et al. 1990), and some of these norms may seem ar-
bitrary, unnecessary, or potentially even harmful. Norms
often seem strange and unnatural to people unfamiliar with
a culture (Cialdini and Trost 1998; Sumner 1906). Opinions
of a norm vary even within a culture. Some norms may not
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
546 JOURNAL OF CONSUMER RESEARCH
seem legitimate because the norm is associated with a sub-
culture with different interests and goals (see articles in the
Autumn 2013 Research Curation—Thompson 2014). For
example, consumers who frequent dance clubs and raves
may reject the antidrug norms advocated by more main-
stream society (Thornton 1996). Thus, norms vary in the
extent to which people within a culture or subculture con-
sider them legitimate (e.g., necessary, socially beneficial,
just), and behaviors that part from an injunctive norm should
seem appropriate if the norm is seen as arbitrary or lacking
legitimacy. For example, Apple’s call to diverge from a
norm dictating the use of a particular type of computer and
operating system in its famous “1984” advertisement likely
seemed appropriate because many viewers considered the
policy of using IBM/Microsoft computers unnecessarily re-
strictive.
We thus hypothesize that the relationship between auton-
omy and perceived coolness depends on the legitimacy of
the norm from which an autonomous behavior diverges.
Specifically, we predict that autonomy should increase per-
ceptions of coolness when behavior diverges from a norm
considered illegitimate (i.e., unnecessary, arbitrary, or in-
correct) but not when behavior diverges from a norm con-
sidered legitimate.
Bounded Autonomy Is More Appropriate than
Extreme Autonomy
Regardless of whether they diverge from a descriptive or
an injunctive norm, divergent behaviors vary from being
similar to the norm (e.g., a product offered in a new color)
to being radically different (e.g., a completely new product
category). Cultural objects (people, brands, etc.) similarly
vary in the regularity with which their behavior diverges
from the norm, from never diverging to always diverging.
In both cases, autonomy occurs on a continuum from not
autonomous to extremely autonomous. Research suggests
that consumers often prefer cultural objects that signal a
moderate or bounded level of autonomy. Individuals attempt
to balance competing goals for uniqueness and belonging
by balancing divergent and normative behaviors (Brewer
1991) and avoiding brands that are either too popular (i.e.,
no divergence from the norm) or too unpopular (too much
divergence; Berger and Heath 2007). Consumers show
greater acceptance of products that diverge moderately from
the descriptive category norm, as extremely divergent prod-
ucts are often difficult to understand (Campbell and Good-
stein 2001; Jhang, Grant, and Campbell 2012; Meyers-Levy
and Tybout 1989). Consumers are also likely to be more
accepting of less extreme departures from injunctive norms.
For example, fans may consider a celebrity’s mild misdeeds
appropriate (e.g., driving above the speed limit, recreational
drug use) but are less likely to approve of more serious
norm deviations (e.g., animal cruelty, drug addiction).
Society likewise may benefit from bounded expressions
of autonomy. Absolute conformity is disadvantageous be-
cause it inhibits progress and innovation (Burroughs and
Mick 2004) and in extreme cases can result in blind obe-
dience, even to potentially harmful norms (Marcuse 1955;
Milgram 1963). Absolute autonomy, however, is problem-
atic because it can lead to antisocial behavior and general
lack of coordination between people (Heath and Potter 2004;
Hobbes 1651/1991). Bounded expressions of autonomy,
conversely, likely reduce the problems associated with either
extreme conformity or extreme divergence. Thus, autono-
mous behavior should seem more appropriate if it is bounded
rather than extreme. For example, Apple’s 1984 ad would
probably have seemed less appropriate if the ad encouraged
consumers to rebel against all of society rather than an office
policy.
Because bounded divergence should seem more appro-
priate than unconstrained divergence, we hypothesize that
autonomy increases perceptions of coolness when the level
of autonomy is moderate but not when it is too high. In
other words, autonomy should have a curvilinear effect on
perceived coolness, such that coolness first increases and
then decreases as autonomy increases.
Counterculturalism Moderates the Level of
Autonomy Considered Appropriate
Consumers differ in the extent to which they consider
autonomous behavior appropriate. Researchers have docu-
mented subcultures of consumers who generally consider
societal institutions and authority figures unjust, repressive,
and damaging (Heath and Potter 2004; Marcuse 1955;
Thompson and Coskuner-Balli 2007). Such consumers typ-
ically believe that societal institutions, including mass me-
dia, government, church, the economic elite, and even fam-
ily and friends, impose a repressive force that suffocates
creativity, individualism, and self-actualization and leads to
widespread conformity. We refer to this worldview as coun-
terculturalism. Because they are more likely to consider so-
cial norms unjust and overly repressive, consumers higher
in counterculturalism should perceive higher levels of au-
tonomy to be appropriate relative to consumers lower in
counterculturalism (Heath and Potter 2004). For example,
consumers high in counterculturalism will be more likely
to consider the highly autonomous behavior of the motor-
cycle gangs associated with Harley Davidson appropriate,
and therefore think of Harley Davidson as cool, than con-
sumers lower in counterculturalism.
Because consumers who are critical of authority and so-
cial norms are more likely to consider divergence appro-
priate, we hypothesize that counterculturalism will moderate
the level of autonomy that consumers perceive to be cool.
Specifically, consumers higher in counterculturalism should
perceive higher levels of autonomy as cool compared to
consumers lower in counterculturalism (see fig. 1).
Summary and Outline of Experiments
Our hypothesis that coolness stems from appropriate au-
tonomy suggests that the relationship between autonomy and
perceived coolness is more nuanced than previously dis-
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 547
FIGURE 1
HYPOTHESIZED RELATIONSHIP BETWEEN PERCEIVED
AUTONOMY, COUNTERCULTURALISM, AND
PERCEPTIONS OF COOLNESS
cussed in the literature. Specifically, we propose that the
relationship between autonomy and perceived coolness de-
pends on the legitimacy of the norm from which the au-
tonomous behavior diverges, the extremity of autonomy, and
the worldview of the perceiver.
We test the relationship between autonomy and perceived
coolness in six experiments. Study 1 shows that autonomous
product designs are perceived to be cooler than equally liked
normative designs. Study 2 illustrates that autonomy only
increases perceived coolness when the autonomous behavior
diverges from a norm seen as illegitimate, not when it di-
verges from a legitimate norm. Studies 3, 4a, and 4b show
that autonomy has a curvilinear relationship with percep-
tions of coolness such that coolness first increases and then
decreases as autonomy increases. Studies 4a and 4b also
reveal that consumers who are high in counterculturalism
perceive higher levels of autonomy to be cool compared to
consumers who are low in counterculturalism. Study 5 con-
firms that perceived coolness and preference are not the
same construct by showing that cool products are preferred
only when consumers want to show that they are indepen-
dent. Collectively, the studies affirm that autonomy increases
perceived coolness, but only when the expression of auton-
omy seems appropriate.
STUDY 1: COOL PRODUCTS DIVERGE
FROM THE NORM
Study 1 provided an initial test of the relationship between
autonomy and perceived coolness. The purpose of the study
was to examine whether consumers would perceive a water
bottle as being cooler when it diverged from, rather than
conformed to, the norm. Because we propose that autonomy
can increase coolness regardless of the source, we also ma-
nipulated whether the source was a well-known brand or an
unfamiliar brand. We expected the autonomous product to
seem cooler irrespective of brand familiarity.
Method
We first conducted a pretest (Np51) to identify bottle
designs for use in the study. The pretest, which asked par-
ticipants to rate 11 water bottle designs (without logos),
identified two designs perceived as differing in divergence
but not in value. Participants had equally positive attitudes
toward the two designs (M
lo div
p4.60, M
hi div
p4.39; tp
.63, NS) but perceived them as having different levels of
divergence from the norm (M
lo div
p3.24, M
hi div
p5.65;
tp5.98, p!.001; see table A1 for details).
Participants (Np190) recruited from Amazon’s Me-
chanical Turk completed the focal study in exchange for a
small payment (60% male; M
age
p30.92, range: 18–63;
38% college graduates; all in the United States). The study
used a 2 (divergence: low, high) #2 (brand: familiar, un-
familiar) between-subjects design. Participants read that a
coffee retailer, either Starbucks (familiar) or Sabbarrio (un-
familiar), was changing the design of its water bottles and
then viewed one of the two bottles from the pretest but with
the Starbucks or Sabbarrio logo added to the bottle. We
operationalized perceived coolness as a continuous variable
by having participants complete two 7-point scales anchored
by uncool/cool: “how cool or uncool do you consider the
design” and “how cool or uncool would your friends con-
sider the design” (rp.93, ap.96). Drawing from Ama-
bile’s consensual assessment technique, the individual par-
ticipants’ responses will show an overall consensus of what
is cool in the given context. Participants subsequently an-
swered an open-ended question: “Why do you think the
bottle design is cool or uncool?” Two coders, blind to the
participants’ conditions and coolness ratings, indicated
whether the ratings suggested that the participant inferred a
low (e.g., “the design is so typical,” “looks like any other
water bottle”) or a high level of autonomy (e.g., “it’s unique
and very different,” “it’s a little weird”). In the few instances
when the two coders did not agree (they agreed on 89%),
a third coder resolved the disagreement. Because the di-
vergence in this case was from a descriptive norm (i.e., a
typical bottle design), we next measured appropriate auton-
omy by assessing whether the divergent design seemed at
least as valuable as the normative design. Specifically, par-
ticipants indicated whether “the design is different in a good
way” and “the design is different in a bad way” as proxies
for appropriate and inappropriate autonomy, respectively,
on 7-point scales anchored by “strongly disagree” and
“strongly agree.” Participants finally reported their famil-
iarity with Starbucks and Sabbarrio (see table 1), age, gen-
der, and level of education.
Results and Discussion
A 2 (divergence: low vs. high) #2 (brand replicate:
familiar, unfamiliar) ANOVA revealed a highly significant
main effect of divergence. As predicted, the bottle that di-
verged from the norm was considered to be cooler than the
bottle that conformed to the norm (M
hi div
p4.98, M
lo div
p
3.47; F(1, 186) p34.04, p!.001). No other effects were
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
548 JOURNAL OF CONSUMER RESEARCH
TABLE 1
STIMULI AND RESULTS FOR STUDY 1
Familiar brand
Unfamiliar brand
Low divergence High divergence
a
Brand Unfamiliar Familiar Average Unfamiliar Familiar Average r
bcoolness
Pretest ratings of the unbranded bottles:
Divergence . . . . . . 3.24 (2.10) . . . . . . 5.65*** (1.73) . . .
Attitude . . . . . . 4.60 (1.84) . . . . . . 4.39 (1.97) . . .
Perceived coolness (mean) 3.49 (1.59) 3.46 (1.71) 3.47 (1.65) 4.77*** (1.90) 5.21*** (1.89) 4.98*** (1.93) . . .
Brand familiarity (mean) 1.74 (1.56) 5.73 (1.58) 3.74 1.45 (1.21) 6.16 (1.09) 3.81 .07
Appropriate autonomy (mean) 3.14 (1.89) 2.73 (1.90) 2.92 (1.89) 4.57*** (2.06) 5.07*** (1.94) 4.81*** (2.01) .85***
Inappropriate autonomy (mean) 3.35 (1.94) 3.27 (1.94) 3.31 (1.93) 3.65 (2.19) 2.96 (2.02) 3.32 (2.13) .44***
Proportion of open-ended inferences:
High autonomy 5% 4% 4% 76%*** 70%*** 73%*** .36***
Low autonomy 58% 60% 59% 2%*** 0%*** 1%*** .42***
N
OTE
.—Standard deviations in parentheses.
a
Asterisks indicate significant contrasts for the corresponding brand across the high- and low-divergence conditions.
b
Correlation between the response and perceptions of coolness.
***p!.01.
significant (all F!1), indicating that the effect of divergence
on coolness occurred both when the brand was familiar
(Starbucks) and when the brand was unfamiliar (Sabbarrio;
see table 1).
Next we assessed whether inferred autonomy, coded from
the open-ended responses by subtracting comments men-
tioning low autonomy from comments mentioning high au-
tonomy, mediated participants’ perceptions of coolness. We
tested for mediation using a model with the divergence ma-
nipulation (low autonomy coded as 1, high autonomy
coded as 1) as the independent variable, perceived coolness
as the dependent variable, and inferred autonomy as the
mediating variable (Preacher and Hayes 2008). As predicted,
participants inferred more autonomy when the bottle design
diverged rather than conformed to the norm (bp.63, tp
16.42, p!.001), and they perceived the bottle as cooler
when they inferred a higher level of autonomy (bp.85, t
p3.60, p!.001). Importantly, inferred autonomy mediated
the effect of the divergence manipulation on perceived cool-
ness (indirect effect p.53; 95% confidence interval [CI] p
.21–.87). Thus, the open-ended responses confirmed that the
divergent bottle design seemed cooler because participants
inferred that the bottle was more autonomous when its de-
sign diverged from rather than conformed to the norm.
Finally, we conducted a preliminary test of whether con-
sumers perceived the divergent design to be cooler because
they considered the expression of autonomy appropriate. To
do so, we ran a second mediation analysis with the measures
of appropriate and inappropriate autonomy as potential me-
diating variables. Participants rated the bottle that diverged
from the norm as expressing a higher level of appropriate
autonomy (bp.95, tp6.69, p!.001) but a similar level
of inappropriate autonomy (bp.01, tp.04, NS) as the
bottle that conformed. Moreover, ratings of appropriate au-
tonomy significantly increased perceptions of coolness (bp
.70, tp16.36, p!.001), whereas ratings of inappropriate
autonomy had the opposite effect (bp.10, tp2.56,
p!.05). Importantly, the effect of the divergence manip-
ulation on perceived coolness was significantly mediated by
appropriate autonomy (indirect effect p.66; 95% CI p
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 549
.46–.89) but not by inappropriate autonomy (indirect effect
p.00; 95% CI p.04 to .03).
A product whose design differed from the norm seemed
cooler than an equally liked product whose design con-
formed to the norm. Moreover, the effect of diverging from
the norm on perceived coolness occurred because partici-
pants inferred that the brand was autonomous in an appro-
priate way. However, the data also tentatively suggest that
consumers do not always perceive autonomous behavior as
cool, as the perception that the product diverged from the
norm in a bad way was negatively related to perceived cool-
ness. Our subsequent studies follow up on this idea by more
directly examining when autonomous behavior increases
perceived coolness.
STUDY 2: DIVERGING FROM
LEGITIMATE AND
ILLEGITIMATE NORMS
Study 2 examined when expressing autonomy by diverg-
ing from an injunctive norm is likely to increase perceptions
of coolness. Injunctive norms indicate rules or ideals that
people are expected to follow. People have different goals,
values, and interests and, thus, different views about which
social norms are necessary and beneficial and which are
expendable. For example, although most consumers con-
sider norms that forbid stealing music CDs legitimate, some
perceive norms that forbid the piracy of digital music files
illegitimate (Giesler 2008).
To test our hypothesis that expressing autonomy is only
likely to increase perceptions of coolness when behavior
diverges from a norm that is not considered legitimate, we
asked participants to read an advertisement for a foreign
clothing brand that advocates wearing blue. In order to ma-
nipulate autonomy and norm legitimacy while controlling
for the brand’s behavior, we described the culture in which
the advertisement ostensibly aired. Depending on the de-
scription, the brand’s behavior either conformed to or di-
verged from the norm, which varied in legitimacy. We pre-
dicted that diverging from an illegitimate norm would
increase perceptions of coolness but that diverging from a
legitimate norm would have the opposite effect.
Method
We randomly assigned participants from Amazon’s Me-
chanical Turk (Np196, 43% female, M
age
p32.5, all in
the United States) to one condition in a 2 (autonomy: high,
low) #2 (norm: legitimate, illegitimate) between-subjects
experiment. Participants read that they would complete a
survey about a foreign retailer’s advertising strategy.
Participants read about a retailer, Roiku, that advertises
and sells clothing in foreign markets, including the city-state
of Ballai. Citizens of Ballai were described as celebrating
a holiday called Masakha Day by wearing a certain color.
Participants in the low-autonomy condition read that the
norm was to wear blue, whereas participants in the high-
autonomy condition read that the norm was to wear white.
We manipulated the legitimacy of the norm by describing
the reason for wearing blue (or white). Participants in the
legitimate norm condition read that citizens of Ballai wear
blue (white) in order to honor fallen soldiers. Conversely,
participants in the illegitimate norm condition read that the
citizens wear blue (white) in order to pay tribute to a corrupt
dictator (see complete manipulations in table A2). In order
to check that they understood the fashion norm for Masakha
Day and the reason behind it, participants next answered
two open-ended questions: (1) What color do citizens typ-
ically wear on Masakha Day? (2) Why do most citizens
wear this color? Next, participants had the option of re-
reading the description of the fashion norms or of continuing
with the study.
Subsequently, participants in the low (high) autonomy
condition read, “Roiku recently created an advertising cam-
paign in which the brand endorsed (parted from) the norm
by encouraging consumers to wear blue on Masakha Day.”
Next, all participants viewed an advertisement with the
headline “Roiku is blue” and the subtext “Masakha Day
Collection” (see fig. 2). In sum, all participants viewed the
same advertisement in which a brand endorsed wearing blue;
however, the brand’s behavior either conformed to or di-
verged from the norm, and the norm seemed either legiti-
mate or illegitimate depending on the background descrip-
tion of the culture.
After viewing the advertisement, participants rated the
perceived coolness of the brand on two 7-point scales an-
chored by uncool/cool: “To what extent do you personally
consider the brand cool or uncool,” and “To what extent do
you think your close friends would consider the brand cool
or uncool” (ap.95, rp.90). Next, participants completed
the manipulation checks (see table A1) for legitimacy (ap
.94) and autonomy (ap.94). Because autonomy is a central
construct in our research, we conducted several pretests in
order to develop a scale to measure inferred autonomy. The
pretests identified six items (e.g., “doesn’t do things just to
fit in”) that consistently show a unidimensional factor struc-
ture and high reliability. Additionally, to examine whether
expressing autonomy from an illegitimate norm increases
coolness because the autonomy seems appropriate but that
expressing autonomy from a legitimate norm decreases cool-
ness because the autonomy seems inappropriate, we mea-
sured whether participants considered the brand autonomous
in a positive way (“It does its own thing, but in a good
way,” and “It is different in a positive way”; ap.84, rp
.73) and autonomous in a negative way (“It does its own
thing, but in a bad way,” and “It is different in a negative
way”; ap.90, rp.82). Finally, participants reported their
demographic information, including age and gender.
Results
Both manipulations worked as intended. A 2 (autonomy:
high, low) #2 (norm: legitimate, illegitimate) ANOVA
revealed the intended main effect of the autonomy manip-
ulation on the perceived divergence of the brand (M
lo aut
p
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
550 JOURNAL OF CONSUMER RESEARCH
FIGURE 2
ADVERTISEMENT PARTICIPANTS VIEWED IN STUDY 2
2.99, M
hi aut
p5.46; F(1, 192) p142.47, p!.001). The
legitimacy manipulation had neither a main (F(1, 192) p
1.18, NS) nor an interacting (F(1, 192) p2.39, NS) effect.
The legitimacy manipulation was also successful. Partici-
pants perceived the norm as being more legitimate when
citizens wore the color to honor fallen soldiers rather than
a dictator (M
legit
p6.00, M
illegit
p3.04; F(1, 192) p250.45,
p!.001). Consistent with the finding that observing an act
of deviance can make a norm seem less legitimate (Cialdini
et al. 1990; Keizer, Lindenberg, and Steg 2008), the norm
also seemed slightly more legitimate when the brand con-
formed to it than when it expressed autonomy from it (M
lo aut
p4.64, M
hi aut
p4.23; F(1, 192) p5.95, p!.05). Im-
portantly, however, the norm legitimacy manipulation did
not interact with the autonomy manipulation (F(1, 192) p
1.08, NS), and the norm seemed more legitimate in the
legitimate condition both when the norm was to wear blue
(i.e., low autonomy; M
legit
p6.13, M
illegit
p3.35; F(1, 192)
p112.57, p!.001) and when the norm was to wear white
(i.e., high autonomy; M
legit
p5.86, M
illegit
p2.69; F(1, 192)
p138.30, p!.001).
Next, we assessed the effects of the manipulations on
perceived coolness using a 2 (autonomy: high, low) #2
(norm: legitimate, illegitimate) ANOVA. There was a mar-
ginal main effect of autonomy (F(1, 192) p2.97, pp.09);
the main effect of norm legitimacy was not significant (F(1,
192) p.04, NS). Importantly, the interaction between au-
tonomy and norm legitimacy was highly significant (F(1,
192) p42.30, p!.001). As predicted, expressing autonomy
increased the perceived coolness of the brand only when
the norm from which the brand diverged seemed illegitimate
(M
lo aut
p3.96, M
hi aut
p5.10; F(1, 192) p12.02, p!
.001). Expressing autonomy from a legitimate norm instead
decreased the perceived coolness of the brand (M
lo aut
p
5.56, M
hi aut
p3.59; F(1, 192) p32.24, p!.001).
We next examined whether diverging from an illegitimate
norm increases coolness because the autonomy is seen as
appropriate but diverging from a legitimate norm decreases
coolness because the autonomy is seen as inappropriate. If
so, then the increase in perceived coolness when the brand
diverged from an illegitimate norm should be mediated by
the perception that the brand is autonomous in a positive
way. Indeed, mediation tests (Preacher and Hayes 2008)
revealed that the positive effect of expressing autonomy on
perceived coolness in the illegitimate norm condition was
mediated by the perception that the brand was autonomous
in a positive way (indirect effect p.47; 95% CI p.24–
.76). Diverging from an illegitimate norm increased the per-
ception that the brand was autonomous in a positive way
(M
lo aut
p3.03, M
hi aut
p5.03; bp1.00, tp5.92, p!
.001). Moreover, the perception that the brand was auton-
omous in a positive way increased perceived coolness (bp
.47, tp5.34, p!.001). The direct effect of the autonomy
manipulation was no longer significant after accounting for
the indirect effects of positive and negative autonomy (bp
.08, tp.49, NS). If only appropriate displays of autonomy
lead to coolness, then we might find that the decrease in
perceived coolness when a brand diverges from a legitimate
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 551
norm would be mediated by the perception that the brand
is autonomous in a negative way. Consistent with this ex-
planation, the indirect effect of autonomy on perceived cool-
ness in the legitimate norm condition was significant (in-
direct effect p.59; 95% CI p.91 to .36). Diverging
from a legitimate norm increased the perception that the
brand was autonomous in a negative way (M
lo aut
p2.04,
M
hi aut
p4.40; bp1.18, tp7.45, p!.001). Moreover,
the perception that the brand was autonomous in a negative
way decreased perceived coolness (bp.50, tp5.07,
p!.001). Again, the direct effect of the autonomy manip-
ulation was no longer significant after accounting for the
indirect effects of positive and negative autonomy (bp
.27, tp1.51, NS).
Discussion
The effect of diverging from a norm on perceived cool-
ness depends on norm legitimacy. Participants perceived a
brand to be cooler when it diverged from an illegitimate
norm but less cool when it diverged from a legitimate norm.
Although both acts of divergence increased perceptions of
autonomy, diverging from an illegitimate norm seemed ap-
propriate, whereas diverging from a legitimate norm did not.
Thus, the study supports our hypothesis that expressing au-
tonomy increases perceived coolness, but only when the
expression of autonomy seems appropriate.
STUDY 3: BOUNDED AUTONOMY IS
COOLER THAN EXTREME AUTONOMY
We propose that whether autonomy seems appropriate
depends on the extent to which a cultural object diverges
from the norm. Because divergent behavior is less likely to
harm others and disrupt the social order—and hence, seem
appropriate—when it is bounded rather than extreme, we
predict that bounded autonomy will increase perceptions of
coolness, relative to low autonomy, but extreme autonomy
will not. Study 3 thus examined whether perceived coolness
first increases but then decreases as autonomy increases. The
study also investigated the relationship between coolness
and product choice. Although research has assumed that
coolness leads to sales (e.g., Gladwell 1997; Kerner and
Pressman 2007), we could not find any quantitative inves-
tigations of the relationship between perceived coolness and
actual behavior. In the current study, a diverse sample of
participants read interviews with and could choose to down-
load and keep a few songs by three rock bands. The inter-
views portrayed one band as having a low level of autonomy,
a second band as having a higher but bounded level of
autonomy, and a third band as having an extreme level of
autonomy. In addition to expecting a curvilinear relationship
between autonomy and coolness, we expected a similar cur-
vilinear relationship between autonomy and actual choice,
mediated by perceptions of coolness.
Method
Participants. We offered US workers from Amazon’s
Mechanical Turk (Np133) a small payment and their
choice of four song downloads for participating in a study
on rock music. The sample was heterogeneous in terms of
gender (60% female), age (range: 18–62, Mp30.3, SD p
10.6), annual household income (31%, !$25,000; 24%,
$25,000–$50,000; 20%, $50,000–$75,000; 11%, $75,000–
$100,000; 16%, 1$100,000), and education (highest degree:
38%, high school; 15%, associate’s degree; and 47%, bach-
elor’s degree or higher).
Procedure. Participants read three interviews with mu-
sicians in up-and-coming rock bands, answered questions
about each, and chose four songs to download. Through a
personal connection, we identified three professional mu-
sicians who had recently started bands that were not yet well
known (no more than 5% of the sample had previously heard
of any of the bands) and were willing to be in the study
and to sell us four song files each to use as choice options:
Nick Campbell of Wages, Andy Herod of Electric Owls,
and Matt Rumley of Her Marigold. Each interview in-
cluded four questions: (1) “How would you describe your
band?” (2) “What motivates you when writing your songs?”
(3) “What is your favorite song to play live?” and (4) “What
songs would you recommend to someone who has never
heard your band?” The answers to questions 1, 3, and 4
were provided by each band and held constant across par-
ticipants. With permission from the musicians, we created
the answers to the second interview question to manipulate
autonomy at three levels (low, bounded, extreme), within
subjects, as follows:
Low autonomy: “We make an effort to write songs that
appeal to a mass audience. By following the current
trends and sticking to popular sounds, we create songs
that we hope everyone will love. Our goal is to match
our sound to mainstream tastes so as many people as
possible enjoy our music.”
Bounded autonomy: “We don’t try to write a bunch of
hits or records that go triple platinum. We just write
songs that feel right to us and reflect what we are
experiencing at the time. We see what is happening
around us and try to incorporate these observations into
songs that we hope some people can relate to.”
Extreme autonomy: “We write what we feel like writ-
ing, which usually means completely ignoring typical
conventions and doing something totally different. We
do what we want and if people don’t like it, that’s their
problem. Honestly, we couldn’t care less what others
think of us or our music.”
While all participants read about the bands in the same
order (Wages first, Electric Owls second, and Her Marigold
third), we counterbalanced which band’s second answer in-
dicated low, bounded, or extreme autonomy such that au-
tonomy was orthogonal to the band and order. Each band
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
552 JOURNAL OF CONSUMER RESEARCH
FIGURE 3
PERCEIVED COOLNESS AND SONG
DOWNLOADS IN STUDY 3
expressed low autonomy for a third of the participants,
bounded autonomy for a third of the participants, and ex-
treme autonomy for a third of the participants. After reading
the interviews, participants rated the extent to which they
considered the bands cool, rated perceived autonomy of the
band, and chose four songs to download and keep. We coun-
terbalanced the order of the measures of band perceptions
and song downloads. Order did not have main or interacting
effects on perceptions of autonomy, coolness, or download-
ing behavior and is not discussed further. We measured cool-
ness by asking, “Do you think this band is cool?” Partici-
pants indicated their response on a 7-point scale anchored
by not cool/cool. We measured perceived autonomy of the
band using the same six-item scale used in the previous
study (ap.91; see table A1). Participants also selected
four out of 12 songs (four by each band) to download. After
reporting demographic information, participants were de-
briefed and directed to a web page to download their song
selections and learn more about the bands.
Results
Effects of Order and Band. First, we calculated the main
effect of band/order on perceptions of autonomy, percep-
tions of coolness, and the number of songs downloaded
using a repeated-measures ANOVA with band as a within-
subjects factor. Band did not have a significant main effect
on perceived autonomy (M
Wages
p3.19, M
Owls
p3.14,
M
Marigold
p3.30; F(2, 264) p.56, NS) or perceived coolness
(M
Wages
p4.62, M
Owls
p4.46, M
Marigold
p4.77; F(2, 264)
p1.42, NS). There was, however, a main effect of band
on downloads (M
Wages
p1.41, M
Owls
p1.02, M
Marigold
p
1.56; F(2, 264) p7.93, p!.001). Subsequent analysis
adjusts the number of downloads to control for the main
effect of band in order to separate it from the effects of the
autonomy manipulation.
Manipulation Check. We assessed the effectiveness of
the autonomy manipulation by entering perceived autonomy
as the dependent variable in a 3 (autonomy: low, bounded,
extreme) #3 (band-autonomy pairing: Wages/bounded,
Electric Owls/bounded, Her Marigold/bounded) mixed
ANOVA model with autonomy as a within-subjects factor
and band-autonomy pairing as a between-subjects factor.
Only the main effect of autonomy was significant (F(2, 260)
p123.77, p!.001). Paired sample t-tests revealed that
participants considered the extreme-autonomy band (Mp
4.02) more autonomous than the bounded autonomy band
(Mp3.27; tp7.99, p!.001), which they considered
more autonomous than the low-autonomy band (Mp2.34;
tp9.78, p!.001). Thus, the autonomy manipulation was
successful.
Coolness. We tested the relationship between autonomy
and coolness by running the same analysis with perceptions
of coolness as the dependent variable. The analysis revealed
a significant main effect of autonomy (F(2, 260) p10.15,
p!.001; see fig. 3). We tested the nature of the relationship
between autonomy and perceived coolness by assessing the
effects of the linear and quadratic trends of autonomy. As
hypothesized, there was a significant quadratic effect of au-
tonomy on coolness (F(1, 130) p28.96, p!.001). The
band displaying bounded autonomy was considered cooler
(Mp5.07) than both the band displaying low autonomy
(Mp4.30; tp4.56, p!.001) and the band displaying
extreme autonomy (Mp4.48; tp3.76, p!.001). The
linear effect of autonomy was not significant (F(1, 130) p
.64, NS), indicating that the main effect of the autonomy
manipulation was driven by the aforementioned curvilinear
relationship between autonomy and coolness. Finally, the
effect of autonomy on coolness did not depend on the band-
autonomy pairing (F(4, 260) p.77, NS).
Download Choice. If people are more likely to choose
products that they consider cool, we would expect a similar
curvilinear relationship between autonomy and participants’
download choices; this is what we found. There was a mar-
ginally significant main effect of autonomy (F(2, 260) p
2.84, pp.06) driven by a significant quadratic trend (F(1,
130) p4.62, p!.05; see fig. 3). Participants chose more
songs from the band showing bounded autonomy (Mp
1.51) than both the band showing low autonomy (Mp1.18;
tp2.48, pp.01) and, directionally but not significantly,
the band showing extreme autonomy (Mp1.31; tp
1.37, NS). The effect of autonomy on choice did not de-
pend on the band-autonomy pairing (F(4, 260) p.82, NS).
Mediation. We examined whether perceptions of cool-
ness mediated choice behavior using a procedure recom-
mended for within-subjects designs (Judd, Kenny, and Mc-
Clelland 2001). First, we calculated a score representing the
quadratic effect of autonomy on downloads (DL
quad
) for each
participant, by subtracting the number of songs downloaded
from the low-autonomy band, adding twice the number of
songs downloaded from the bounded autonomy band, and
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 553
subtracting the number of songs downloaded from the ex-
treme-autonomy band (i.e., coded 1, 2, 1). Consistent
with the quadratic trend reported above, on average DL
quad
was significantly greater than zero (Mp.53; tp2.17, p
!.05). We then regressed DL
quad
on mean-centered average
ratings of coolness, and two contrast-coded variables rep-
resenting the linear (coded 1, 0, 1) and quadratic (coded
1, 2, 1) effects of autonomy on coolness. The quadratic
effect of autonomy on coolness explained significant vari-
ance in the quadratic effect of autonomy on downloads (b
p.35, tp4.45, p!.001); however, the linear effect of
autonomy on coolness did not (bp.11, tp1.23, NS).
Importantly, the intercept in the regression equation was not
significant (bp.08, tp.32, NS), suggesting that differ-
ences in perceived coolness fully mediated the curvilinear
effect of autonomy on song choice.
Discussion
The data revealed a curvilinear relationship between au-
tonomy and perceived coolness: bands were considered
cooler when they showed bounded autonomy rather than a
low or extremely high level. Additionally, song choice
showed a similar curvilinear pattern that was mediated by
perceived coolness, indicating that coolness can influence
consumer choice. The study provided additional evidence
that autonomy increases perceptions of coolness when the
deviation seems appropriate, in this case because autonomy
was bounded rather than too extreme. However, the study
did not address how this boundary differs across consumers.
Understanding how perceptions of coolness differ is im-
portant in order to gain a deeper theoretical understanding
of the antecedents of coolness.
STUDIES 4A AND 4B: MORE AUTONOMY
SEEMS COOL TO COUNTERCULTURALS
Consumers differ in the extent to which they consider
autonomy appropriate. Because they are skeptical of society
and are less likely to consider its norms legitimate, con-
sumers high in counterculturalism are likely to consider
higher levels of autonomy appropriate than consumers low
in counterculturalism. Therefore, we hypothesize that more
countercultural consumers perceive a higher level of auton-
omy to be cool than less countercultural consumers.
We conducted two studies to test this hypothesis. Both
studies manipulated autonomy and measured countercultur-
alism. Study 4a asked participants to evaluate the coolness
of four fashion brands, which varied by showing a low,
moderate, high, or extreme level of autonomy, respectively.
We manipulated autonomy at four (rather than three) levels
in order to increase the ability to detect differences in the
level of autonomy considered the coolest for high- and low-
countercultural participants. Study 4b examined the gener-
ality of the findings by asking participants to evaluate the
coolness of another person, whose autonomy varied from
low to extreme (manipulated between subjects). Understand-
ing what makes people cool is relevant for consumer re-
search because many consumers buy products and engage
in activities that they believe will help make them cool (Belk
et al. 2010; Heath and Potter 2004). Additionally, research-
ers have suggested that cool people exert a larger social
influence and are more likely to be imitated by other con-
sumers (Belk et al. 2010; Gladwell 1997). Therefore, un-
derstanding whether the same factors that influence the cool-
ness of brands also influence the coolness of people could
help inform consumers who want to be cool as well as
marketers who want to target cool consumers.
In both study 4a and study 4b we expected to replicate
the curvilinear effect of autonomy on perceived coolness,
but we also expected that consumers higher in countercul-
turalism would perceive higher levels of autonomy to be
cool compared to consumers lower in counterculturalism.
Thus, we predicted a curvilinear pattern for all consumers
but with coolness peaking at a relatively higher level of
autonomy for consumers higher in counterculturalism.
Study 4a
Method. Undergraduates (Np58, 29% female) at the
University of Colorado participated in the study for course
credit. The study crossed four levels of autonomy (low,
moderate, high, extreme) manipulated within subjects with
a continuous measure of counterculturalism. Participants
read descriptions of four fictitious brands: (1) Baraccio coats,
(2) Ero sunglasses, (3) Setia shoes, and (4) Solle watches.
We counterbalanced which brand expressed a low, moderate,
high, or extreme level of autonomy, such that the autonomy
manipulation was orthogonal to the brand and presentation
order. We manipulated autonomy from low to extreme levels
by indicating whether the brand “follows the market,” “usu-
ally conforms to popular styles,” “defies industry standards,”
or is “rebellious and controversial” (see table A3 for full
descriptions).
We measured the perceived coolness of each brand using
two 7-point measures—“Is this brand cool?” and “Is this
brand hip?”—anchored by not cool/cool and not hip/hip (all
r1.7). Participants then read, “Brands are often seen as
possessing human characteristics and personalities. Next,we
are going to ask you a few questions about the personality
of these same four brands.” Participants indicated the extent
to which the six items in the autonomy scale (all a1.73)
described each brand. Finally, as part of an ostensibly sep-
arate survey, we measured counterculturalism using an orig-
inal Likert-type six-item scale (e.g., “rules and conventions
often overly restrict people’s freedom”; see table A1). We
conducted several pretests to select and validate the items
in the scale. The pretests confirmed that the scale consis-
tently shows a unidimensional factor structure, high reli-
ability (a1.8), and discriminant validity from other po-
tentially related measures, including our six-item measure
of autonomy (r!F.2F), measures of individualism (r!
F.3F), and need for uniqueness (.2 !r!.5).
Results. First, we examined the main effect of brand,
which was confounded with product and order, on percep-
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
554 JOURNAL OF CONSUMER RESEARCH
FIGURE 4
EFFECTS OF AUTONOMY AND COUNTERCULTURALISM ON
PERCEPTIONS OF COOLNESS IN STUDIES 4A AND 4B
N
OTE
.—Autonomy refers to the level of autonomy of the brand (a,
study 4a) or person (b, study 4b) evaluated in the study, whereas
counterculturalism refers to individual differences in the participants
evaluating the brand or person.
tions of autonomy and coolness. If the brand/product/order
factor were significant, we would want to account for this
effect in subsequent analyses in order to separate it from
the effects of the autonomy manipulation. Brand did not
have a significant effect on perceived autonomy (M
coat
p
3.33, M
sungl
p2.94, M
shoes
p3.21, M
watch
p3.05; F(3, 168)
p.76, NS) but did on perceived coolness (M
coat
p4.19,
M
sungl
p4.80, M
shoes
p4.34, M
watch
p4.88; F(3, 171) p
3.22, p!.05). Therefore, we adjusted the ratings of per-
ceived coolness in the subsequent analysis to control for this
effect.
Next, we tested the success of the autonomy manipulation
by entering perceived autonomy as the dependent variable
in a 4 (autonomy: low, moderate, high, extreme) #4 (order:
low autonomy first, moderate autonomy first, high autonomy
first, extreme autonomy first) repeated-measures ANOVA
that also included counterculturalism and its interactions
with the manipulated factors as continuous predictor vari-
ables. Only the main effect of autonomy was significant
(F(3, 156) p316.43, p!.001). Paired sample t-tests re-
vealed that participants perceived each level of brand au-
tonomy to be significantly different in the expected direction
(M
low
p1.43, M
mod
p2.60; tp12.26, p!.001; M
high
p
3.60; tp8.85, p!.001; M
extreme
p4.67; tp10.38, p!
.001). Thus, the manipulation worked as intended.
To test our primary hypothesis, we conducted the same
analysis with perceptions of coolness as the dependent var-
iable. Analysis revealed significant main effects of auton-
omy (F(3, 159) p15.07, p!.001) and counterculturalism
(F(1, 53) p4.87, p!.05), qualified by a significant inter-
action (F(3, 159) p3.41, p!.05; see fig. 4a). To interpret
these effects, we split the autonomy factor into its linear,
quadratic, and cubic components. Consistent with the pre-
vious study, there was a significant main effect of the qua-
dratic trend of autonomy on coolness (F(1, 53) p41.38, p
!.001). Participants perceived brands showing moderate
autonomy (Mp4.83) to be cooler than brands showing
low autonomy (Mp3.64; tp5.98, p!.001) but brands
showing extreme autonomy (Mp4.50) as less cool than
brands showing high autonomy (Mp5.13; tp3.01, p
!.01). The curvilinear pattern held throughout the sample,
as indicated by an insignificant interaction between coun-
terculturalism and the quadratic trend of autonomy (F(1, 53)
p.63, NS). Unlike in the previous study, we also observed
a main effect of the linear trend of autonomy on coolness
(F(1, 53) p7.97, p!.01); perceived coolness increased
with autonomy, although, as previously noted, this increas-
ing pattern did not hold at extremely high levels of auton-
omy. Furthermore, the increasing relationship between au-
tonomy and perceived coolness was truer of participants
high in counterculturalism, as indicated by the significant
interaction between counterculturalism and the linear effect
of autonomy (F(1, 53) p5.35, p!.05). Consistent with
our prediction, participants higher in counterculturalism con-
sidered higher levels of autonomy cooler than did partici-
pants lower in counterculturalism (see fig. 4a).
Study 4b
Method. Undergraduates (Np132, 35% female) at the
University of Colorado participated in exchange for a candy
bar. The study design was the same as study 4a with two
exceptions: (1) we manipulated autonomy between subjects
instead of within subjects, and (2) participants read a de-
scription of a hypothetical target person instead of descrip-
tions of brands (see table A4).
Results. Analysis confirmed the successful manipulation
of autonomy. Participants considered the target person less
autonomous in the low- (Mp2.36) than the moderate-
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 555
autonomy condition (Mp3.86; tp5.89, p!.001), more
autonomous in the high-autonomy condition (Mp5.00; t
p4.42, p!.001), and the most autonomous in the extreme-
autonomy condition (Mp5.50; tp1.95, pp.05). Coun-
terculturalism did not influence perceived autonomy (F(1,
123) p1.12, NS), nor did it interact with the autonomy
manipulation (F(3, 123) p.57, NS). The autonomy ma-
nipulation did not influence the counterculturalism measure
(F(3, 127) p1.33, NS).
To test the hypothesis that autonomy would have a cur-
vilinear effect on perceived coolness and the hypothesis that
higher levels of autonomy would seem cooler to consumers
higher in counterculturalism, we analyzed coolness as the
dependent variable in a regression equation with three or-
thogonal, contrast-coded variables representing the auton-
omy manipulation, the mean-centered measure of counter-
culturalism, and three variables representing the interaction
between the autonomy manipulation and counterculturalism.
As predicted, there was a significant main effect of the qua-
dratic trend of autonomy on coolness (bp.31, tp2.36,
p!.05). Overall, perceived coolness increased as autonomy
increased from a low to a moderate level but began to de-
crease as the level of autonomy became more extreme. The
interaction between the quadratic trend of autonomy and
counterculturalism was not significant (bp.27, tp1.51,
NS), indicating that the overall relationship between auton-
omy and coolness was curvilinear across the entire sample
irrespective of level of counterculturalism. Importantly,
however, the predicted interaction between the linear trend
of autonomy and counterculturalism was significant (bp
.21, tp2.66, p!.01). Relative to participants lower in
counterculturalism, those higher in counterculturalism per-
ceived higher levels of autonomy to be cool (see fig. 4b).
Discussion
Replicating study 3, both brands (study 4a) and people
(study 4b) were considered the coolest when perceived as
having bounded autonomy, as evidenced by a quadratic re-
lationship between autonomy and perceived coolness. Im-
portantly, the studies also illustrated one reason why per-
ceptions of coolness vary across consumers. Perceptions of
coolness peaked at a higher level of autonomy for partici-
pants higher as compared to lower in counterculturalism, a
result that provides additional support for the hypothesis
that autonomy increases perceived coolness but only when
the expression of autonomy seems appropriate. Consumers
high in counterculturalism are critical of societal institutions
and thus tend to perceive higher levels of autonomy to be
cooler than do consumers lower in counterculturalism, who
are more likely to consider autonomous behavior inappro-
priate and, therefore, uncool.
Our studies thus far indicate that perceptions of coolness
increase when a brand or person seems autonomous in an
appropriate manner. An alternative hypothesis, however, is
that coolness does not require autonomy but is merely an-
other way of saying that something is liked. Although the
results of study 1 suggest that autonomy influences per-
ceived coolness independent of liking, identifying contexts
in which consumer preferences diverge from perceptions of
coolness would provide strong evidence that coolness and
desirability are distinct.
STUDY 5: WHEN DO CONSUMERS
PREFER COOL BRANDS?
Preferences depend on goals (Van Osselaer et al. 2005),
and coolness seems more likely to facilitate symbolic, iden-
tity goals rather than practical, utilitarian goals (Leland
2004). Therefore, it is likely that consumers use cool brands
as a means for pursuing symbolic goals, such as signaling
a desired identity trait (Berger and Heath 2007). But what
identity do cool brands signal? Because coolness comes
from expressing autonomy in an appropriate way, consumers
could use cool brands to show that they are autonomous
individuals. If so, then consumers should desire cool brands
more when they want to signal an autonomous identity.
Study 5 examined whether a goal to signal autonomy versus
conformity would lead to greater preference for cool brands.
Support for the hypothesis would (1) provide a better un-
derstanding of when consumers prefer cool brands, (2) sup-
plement earlier studies by providing complementary evi-
dence that coolness is rooted in autonomy in addition to
desirability, and (3) verify the discriminant validity between
perceived coolness and desirability.
We examined when consumers prefer cool brands by ma-
nipulating coolness and classiness. Depending on condition,
participants thought of a real brand that they perceived to
be cool, they did not perceive to be cool, they considered
classy, or they did not consider classy. Participants then
indicated their preference for using the brand in different
social contexts, which we varied to manipulate participants’
desire to express autonomy. Although Western consumers,
like our participants, often want to express an autonomous
identity (Brewer 1991; Markus and Schwartz 2010), there
are social contexts that increase their desire to fit in. We
predicted that although cool brands will be liked better than
uncool brands in general, this will not hold in contexts in
which consumers want to fit in. We hypothesized that con-
texts that encourage conformity over autonomy will reduce
preferences for cool relative to uncool brands but will not
influence preferences for classy versus not classy brands.
Method
Seventy-four undergraduates (19% female) at the Uni-
versity of Colorado participated in an online study for course
credit. All participants named one real shoe brand. We asked
participants to name a brand that they considered high qual-
ity and, depending on randomly assigned condition, cool,
not cool, classy, or not classy (we solicited only high-quality
brands in order to control for the perceived value of the
product). Participants next rated their attitude toward wear-
ing the brand in different social contexts that varied in terms
of desirability of standing out. The study used a 2 (trait
valence: positive, negative) #2 (trait type: cool, classy) #
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
556 JOURNAL OF CONSUMER RESEARCH
2 (context autonomy: desired more, desired less) #2 (con-
text replicate: job, dinner) mixed design. We manipulated
the trait factors between subjects and the context factors
within subjects.
After naming a shoe brand that was high quality and
(depending on random assignment) classy, not classy, cool,
or not cool, participants listed their general attitude toward
the brand on 7-point scales anchored by bad/good and un-
favorable/favorable (rp.93, ap.96). Next, participants
indicated their attitude toward wearing the brand in four
specific contexts on a 7-point scale anchored by bad/unfa-
vorable and good/favorable. On the basis of a pretest, we
identified two contexts in which participants have a higher
desire to express autonomy, “a networking event hosted by
a small advertising boutique” and “an informal dinner at a
local cafe´,” and two contexts in which participants have a
lower desire to express autonomy, “a job interview with a
large, traditional corporation” and “a formal dinner at a
fancy restaurant.” Finally, we conducted a manipulation
check of desire to express autonomy in each of the different
contexts by asking, “In the following situations, would you
rather show that you are independent or that you fit in?”
Participants indicated their responses on a 7-point scale an-
chored by “prefer to show that I fit in” and “prefer to show
that I am independent.” We randomized the order in which
we presented the different contexts for both measures.
Results and Discussion
Manipulation Check. To check the effectiveness of the
autonomy manipulation, we ran a 2 (trait valence: positive,
negative) #2 (trait type: cool, classy) #2 (context au-
tonomy: desired more, desired less) #2 (context replicate:
job, dinner) repeated-measures ANOVA on participants’ de-
sire to express autonomy. As expected, there was only a
significant main effect of the context-autonomy factor (Mp
4.48 vs. 3.18; F(1, 70) p21.29, p!.001). Paired-samples
t-tests revealed that participants wanted to show indepen-
dence more at an event with an ad boutique than at an
interview with a corporation (Mp4.88 vs. 3.93; tp4.17,
p!.001) and at a local cafe´ than at a fancy restaurant (M
p4.69 vs. 4.16; tp2.84, p!.01). Neither the replicate
factor nor the between-subjects trait manipulations had any
main or interacting effects. In sum, the manipulation worked
as intended.
General Brand Attitude. To test whether, in general, par-
ticipants preferred cool brands to uncool brands, we ana-
lyzed the effects of the manipulations on general brand at-
titude using a 2 (trait valence: positive, negative) #2 (trait
type: cool, classy) ANOVA, which revealed only a signif-
icant main effect of trait valence (Mp6.65 vs. 4.48; F(1,
70) p32.53, p!.001). Consistent with the downloading
preferences in study 2, participants expressed a higher gen-
eral attitude toward a brand they considered cool than a
brand they did not consider cool (Mp6.69 vs. 4.39; F(1,
70) p26.44, p!.001). Not surprisingly, they also expressed
a higher general attitude toward a brand they considered
classy than a brand they did not consider classy (Mp6.61
vs. 5.34; F(1, 70) p8.38, p!.01).
Context-Specific Brand Attitude. To test whether partic-
ipants’ preference for cool brands depends on the extent to
which they want to express autonomy, we analyzed their
attitude toward wearing the brand in the four different social
contexts, using a 2 (trait valence: positive, negative) #2
(trait type: cool, classy) #2 (context autonomy: desired
more, desired less) #2 (context replicate: job, dinner) re-
peated-measures ANOVA. The analysis revealed main ef-
fects of trait valence (overall, participants had a higher at-
titude toward wearing brands that have a positive trait than
brands that do not have a positive trait; F(1, 70) p28.18,
p!.001), the context-autonomy factor (overall, participants
had a higher attitude toward wearing all of the shoe brands
in contexts in which they were more motivated to express
autonomy; F(1, 70) p29.87, p!.001), and the context
replicate (overall, participants had a higher attitude toward
wearing the shoes to a dinner than to a job-related event;
F(1, 70) p26.46, p!.001). Importantly, the main effects
of trait valence and autonomy were qualified by the pre-
dicted three-way interaction between trait valence, trait type,
and context autonomy (F(1, 70) p6.86, pp.01). No other
interactions were significant. Because the replicate factor
did not interact with any of the other factors, we simplified
the subsequent analysis of the three-way interaction by col-
lapsing across replicates.
To test the hypothesis that consumers are more likely to
prefer cool brands when they want to signal autonomy, we
investigated whether the contrast between participants in the
cool and the uncool condition differed depending on the
social context. A 2 (coolness: yes, no) #2 (context auton-
omy: desired more, desired less) repeated-measures analysis
retaining only data from participants in the cool trait con-
dition revealed the predicted interaction (F(1, 34) p4.52,
p!.05). In situations in which they want to express au-
tonomy, participants had a higher attitude toward a cool than
an uncool brand (Mp5.53 vs. 3.33; F(1, 70) p18.61, p
!.001). When participants were not motivated to express
autonomy, they had a similar attitude toward a cool as an
uncool brand (Mp3.22 vs. 2.36; F(1, 70) p1.70, NS).
Next, we investigated whether desire to express autonomy
similarly moderated attitudes toward brands seen as pos-
sessing a different valued identity trait: class. We compared
attitudes toward using the brand considered classy with at-
titudes toward using the brand not considered classy with
a 2 (class: yes, no) #2 (context autonomy: desired more,
desired less) repeated-measures analysis retaining only data
from participants in the classy trait condition. The interaction
in this analysis was not significant (F(1, 36) p2.69, NS),
indicating that the preference for a classy versus a nonclassy
brand did not depend on participants’ desire to express au-
tonomy. As expected, participants preferred the classy brand
to the nonclassy brand, both in situations in which they
would want to express autonomy (Mp5.24 vs. 3.82; F(1,
70) p8.24, p!.01) and in situations in which they would
not (Mp4.84 vs. 2.24; F(1, 70) p16.40, p!.001). Thus,
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 557
a desire to express autonomy influenced preferences for cool
brands but not for classy brands.
Discussion. Study 5 demonstrated that consumers prefer
cool brands only when they want to express an autonomous
identity. In contexts that encourage autonomy expression,
consumers preferred a cool brand more than an uncool brand
and as much as a classy brand. Conversely, in contexts that
encourage conformity, consumers did not prefer a cool brand
to an uncool brand and preferred a classy brand to both.
Thus, cool is not merely another way of saying something
is desirable or liked. Although brands seen as cool are often
preferred to brands seen as uncool, consumers are less likely
to show this preference when they want to fit in.
GENERAL DISCUSSION
Understanding what makes things cool has puzzled ac-
ademics and marketers alike. We address this question by
empirically examining the relationship between autonomy
and perceived coolness, finding that brands and people that
diverge from the norm in a way that seems appropriate are
perceived to be cool. Study 1 illustrates that consumers per-
ceive a product whose design diverges from the norm to be
cooler than a product whose design conforms to the norm
both for a familiar, real brand (Starbucks) and for an un-
known, fictitious brand (Sabbarrio). Studies 2, 3, 4a, and 4b
illustrate that autonomous behavior only increases percep-
tions of coolness when the autonomy seems appropriate.
Whether expressing autonomy seems appropriate, and thus
cool or uncool, depends on the norm from which an object
diverges. Study 2 shows that diverging from an illegitimate
norm increases perceived coolness, but diverging from a
legitimate norm has the opposite effect. Whether autonomy
seems appropriate also depends on the extent to which an
object diverges. Our studies reveal a curvilinear effect of
autonomy on perceptions of coolness such that rock bands
(study 3), fashion brands (study 4a), and people (study 4b)
expressing moderate, “bounded” autonomy are considered
cooler than both those expressing low autonomy and those
expressing very high levels of autonomy. Further, the level
of autonomy considered the coolest depends on the con-
sumers evaluating the object. Studies 4a and 4b illustrate
that countercultural consumers, who are critical of main-
stream social institutions and thus more likely to consider
autonomous behavior appropriate, perceive higher levels of
autonomy to be cool compared to less countercultural con-
sumers. Finally, study 5 affirms that cool is not merely another
way of saying something is “good.” Although cool brands
are often preferred to uncool brands, this is moderated by a
consumer’s goal to signal autonomy versus to fit in. Collec-
tively, the studies empirically support a conceptualization of
coolness as a subjective, socially constructed positive trait
attributed to cultural objects (e.g., people, brands, products,
trends) perceived to be appropriately autonomous.
How Do Perceptions of Coolness Change?
One limitation of the research presented is that it does
not directly investigate the dynamic nature of coolness.
However, the finding that perceptions of appropriate auton-
omy lead to coolness and that countercultural consumers
consider a higher level of autonomy appropriate may help
explain where coolness originates and how cool trends dif-
fuse and change over time. Cool trends typically begin when
people and subcultures seen as removed from mainstream
society enact a behavior that deviates from a mainstream
norm (Belk et al. 2010). A broad reading of the literature
suggests that such behaviors may originate in any subculture
perceived to be autonomous from mainstream norms, in-
cluding African Americans (Connor 1995; Leland 2004),
teens (Danesi 1994; O’Donnell and Wardlow 2000), bohe-
mians (Brooks 2000), musicians (Bird and Tapp 2008); beats
(Mailer 1957), hippies (Frank 1997), punks (Hebdige 1979),
ravers (Thornton 1996), mythologized communities like the
imagined Western frontier (Holt 2004), or other subcultures
in which people ignore or resist the constraints imposed by
mainstream society.
Because consumers high in counterculturalism consider
high levels of autonomy appropriate, we believe that they
often ignite cool trends by adopting behaviors that are too
divergent or too obscure to seem appropriate to the less
countercultural masses. As consumers who are high in coun-
terculturalism but outside the subculture begin to adopt the
behavior, its autonomy seems less extreme, making it more
appropriate and, hence, cooler in the eyes of the less coun-
tercultural masses. Eventually, the masses then begin adopt-
ing the behavior in an attempt to become cool themselves
(Gladwell 1997). However, the widespread adoption causes
the behavior to seem less autonomous (Berger 2008) and,
consequently, lose its coolness, first among more counter-
cultural consumers and eventually among less countercul-
tural consumers. The originally cool behavior thus becomes
the new norm and then is replaced by the next cool trend.
For example, the band Green Day initially seemed highly
autonomous due to its relative obscurity and close associ-
ation with a punk subculture in the Bay Area. However, as
Green Day’s critically acclaimed album Dookie spread to
listeners outside of the local punk scene, they became cooler
to a wider audience, including eventually even consumers
lower in counterculturalism. At this time, consumers higher
in counterculturalism, including many of the band’s original
fans, accused Green Day of “selling out.” By the time
Dookie sold 20 million copies, Green Day had completed
a transformation from unknown to cool to pop.
Other times cultural objects have some characteristic or
attribute that most consumers have difficulty considering ap-
propriate. For example, in contrast to Green Day’s polished
and catchy songs, punk band Minor Threat’s radical ideology,
lo-fi recording, and rough vocal style prevented it from ever
catching on with mainstream audiences. In cases like these,
behaviors often remain cool within a small subculture that
considers the divergent behavior appropriate without spread-
ing to a more general population that considers the behavior
too outrageous. For example, even 30 years after breaking
up, Minor Threat remains cool in the eyes of punk subcultures
but relatively unknown to the general public.
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
558 JOURNAL OF CONSUMER RESEARCH
Implications
An important implication of the subjective and dynamic
nature of coolness is that cultivating a cool image may not
be an optimal strategy for brands that want to sustain a
position as a market leader rather than a niche player. Be-
cause coolness prompts diffusion but widespread adoption
makes brands seem less autonomous, it is difficult for mass-
market brands to sustain a high level of perceived coolness
over time. It is also difficult, if not impossible, to seem cool
to everybody simultaneously, as consumers differ in terms
of which norms they consider legitimate and the level of
autonomy they consider appropriate. Maintaining percep-
tions of coolness among more countercultural consumers
requires acts that seem too autonomous to less countercul-
tural consumers. For example, one possible reason why Har-
ley Davidson motorcycles continue to be perceived as cool
by more countercultural consumers is because most main-
stream consumers think Harleys are inappropriately loud and
obnoxious. Conversely, brands that portray a level of au-
tonomy considered cool by less countercultural consumers,
like Starbucks or the Gap, often seem too mainstream to more
countercultural consumers (Thompson and Arsel 2004).
Our research also has interesting implications for public
policy attempting to reduce risky behaviors, such as un-
derage drinking, smoking, and drug use. The most common
strategy to curb these behaviors is to tell people not to do
them. For example, teens are told that drinking will get them
arrested or that smoking cigarettes will give them cancer or
that drugs are bad. This “just say no” strategy may actually
exacerbate risky behaviors by making them seem autono-
mous and, thus, cool. A more effective strategy may be to
position risky behaviors as mainstream or to associate them
with conformity, as done in the successful “Truth” anti-
smoking campaign.
Limitations and Opportunities for Future Research
We limited the scope of our research to the relationship
between autonomy and perceptions of coolness at a specific
point in time. However, as previously discussed, perceptions
of coolness change over time. A better understanding of
how perceptions of coolness change is an important topic
for future research. Additionally, the literature discusses a
number of factors other than autonomy that potentially in-
fluence perceptions of coolness, including concealing emo-
tion, cultural knowledge, narcissism, hedonism, or excite-
ment (Leland 2004; Nancarrow, Nancarrow, and Page 2002;
Pountain and Robins 2000; Southgate 2003). Another im-
portant objective for future research could be to empirically
test whether and how such factors influence perceived cool-
ness.
An additional opportunity for future research could be to
more explicitly identify the different ways in which brands
can express appropriate autonomy. Our studies and the lit-
erature suggest at least three behaviors that prompt infer-
ences that a brand is autonomous. Because brand person-
alities are often inferred from physical features of the
product (Aggarwal and McGill 2007), one way for a brand
to express autonomy is by creating products with uncon-
ventional product design, packaging, or attribute combina-
tions. A second way a brand could express autonomy is
through marketing communications, like the ad for the
Roiku brand in study 2. There are many similar examples
of real brands that have expressed autonomy through ad-
vertising headlines and slogans: Apple (“Think Different”),
Adidas (“Celebrate Originality”), Jack Daniels (“We never
follow the crowd. But they’re always welcome to stop by”),
and Tabasco (“We don’t bother keeping up with the Joneses”).
Because meaning often transfers from people to associated
brands (McCracken 1986), a third way that brands can be-
come autonomous is through association with people. Thus,
the perceived autonomy of a brand can be influenced by the
behavior of the brand’s visible employees (e.g., Richard
Branson and Virgin), endorsers (e.g., Snoop Dogg and Mon-
ster Energy Drink), and customers (e.g., snowboarders and
Burton). Future research could explore whether expressing
autonomy through one of these ways (e.g., an unusual prod-
uct design) may be more likely to seem appropriate and cool
than another (e.g., hiring a rebellious spokesperson).
Additionally, it would be helpful for future research to
investigate whether coolness and the pursuit of coolness is
limited to certain product and service categories or if cool-
ness is potentially a relevant trait for all brands. For example,
consumers may care more about acquiring cool brands for
products consumed publicly (e.g., a car) rather than privately
(e.g., a dishwasher) or for hedonic products (e.g., music)
rather than utilitarian products (toothpaste). A related pos-
sibility is that coolness plays a more important role for prod-
ucts that consumers use to signal their identity (e.g., cloth-
ing) than products that are not seen as identity signals (e.g.,
dish soap; Berger and Heath 2007). Because our studies
typically used public, hedonic products in categories that
consumers frequently use as identity signals (e.g., fashion,
music, beverages), the extent to which our results generalize
to other product and service categories remains unclear.
Another limitation is that our research investigates per-
ceptions of coolness exclusively in Western consumers, who
tend to be individualistic and hold a model of agency that
suggests it is better to control the environment than to try
to adjust to fit within it (Markus and Schwartz 2010; Oys-
erman, Coon, and Kemmelmeier 2002). In contrast, con-
sumers from collectivistic cultures are more likely to believe
that it is better to adjust one’s self to fit within the envi-
ronment than to try to control it (Markus and Schwartz 2010;
Oyserman et al. 2002). It is unclear whether such cultural
differences will merely alter consumers’ perceptions of what
is normal and which norms are legitimate or whether they
will fundamentally alter the relationship between autonomy
and perceived coolness or between perceived coolness and
preference. These too are important questions that could be
addressed in future research.
Another interesting question, which our research only tan-
gentially addresses, is why has coolness become so ubiq-
uitous? Cool has been part of the popular vocabulary in the
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 559
United States since at least the 1950s (Belk et al. 2010;
Pountain and Robins 2000), and, although the objects con-
sidered cool vary, the term has spread beyond the United
States and is now used similarly in a variety of cultures
(Allison 2009; Belk et al. 2010; Gurrieri 2009). The per-
sistence and proliferation of coolness suggests that it may
serve some kind of social function. One possibility, consis-
tent with the finding that appropriate autonomy leads to
perceptions of coolness, is that coolness serves as a reward
to incentivize socially beneficial change. By bringing status
and esteem to people and brands that diverge from the norm
in an appropriate way, the pursuit of coolness may encourage
them to innovate, take risks, and challenge potentially out-
dated norms. In this sense, coolness may offer an alternative
social hierarchy providing status to those whose behavior
offers an appropriate alternative to the status quo rather than
exclusively to those with wealth or a prestigious family
background (Heath and Potter 2004). The notion that cool-
ness encourages socially acceptable change may help ex-
plain why the growth of coolness in public discourse has
coincided with an increasing demand for independence and
creativity in the marketplace (Brooks 2000; Florida 2002).
It may also explain why cool brands appear to be especially
valued in product categories characterized by high levels of
innovation (e.g., high tech) and stylistic change (e.g., fash-
ion, music).
Grossman (2003, 2) describes coolness as “an invisible,
impalpable substance that can make a particular brand of
an otherwise interchangeable product . . . fantastically val-
uable.” Despite the recognized importance of cultivating
cool brands, quantitative research on antecedents of coolness
has been limited. By showing that brands (and people) be-
come cool by being different in an appropriate way, our
research provides a better understanding of this important
aspect of consumer behavior. We hope that our work inspires
continued research on the important question of what makes
things cool.
DATA COLLECTION INFORMATION
We collected the data for each study as follows. Study 1
pretest: collected online by a survey panel service under the
supervision of both authors in August and September 2012.
The first author analyzed the data. Study 1: collected online
through Amazon’s mTurk under the supervision of both
authors in August 2013. The first author analyzed the data.
Study 2: collected online through Amazon’s mTurk under
the supervision of the first author in August 2013. The first
author analyzed the data. Study 3: collected online through
Amazon’s mTurk under the supervision of the first author
in December 2010 and January 2011. The first author an-
alyzed the data. Study 4a: collected at the Behavioral Re-
search Lab at the University of Colorado, Boulder, in April
2010 by research assistants under the supervision of the
second author. The first author analyzed the data. Study 4b:
collected in a classroom at the University of Colorado, Boul-
der, by the first author in April and May 2009. The first
author analyzed the data. Study 5 (both pretest and focal
study): collected at the Behavioral Research Lab at the Uni-
versity of Colorado, Boulder, in April 2010 by research
assistants under the supervision of the second author. The
first author analyzed the data.
APPENDIX
TABLE A1
SCALE ITEMS
Item
Attitude: study 1 pretest (1–7) I like the design.
I would want to drink water from a bottle like this.
Divergence: study 1 pretest (1–7) The design is different from the norm.
The design is unique.
The design shows independence.
Autonomy: studies 2 (1–7), 3 (1
5), 4a (1–5), and 4b (1–7) Lives how it/they/she (hereafter, “it”) wants to live whether or not it pleases others.
Doesn’t do things just to fit in.
Pays little attention to established social norms or conventions.
Rarely caves into social pressure.
Doesn’t change who it is to suit others.
Breaks rules when it feels like it.
Counterculturalism: studies 4a
(1–5) and 4b (1–5) Rules and conventions often overly restrict people’s freedom.
Society traps people by restricting individual autonomy and independence.
Large institutions like corporations and the government exert too much control over our everyday lives.
Many of the problems in society are caused by overly restrictive norms and conventions.
Authority is usually a bad thing.
Society has far too many rules and conventions.
Legitimacy: study 2 (1–7) The norm seems reasonable.
There is a good reason for people to follow this norm.
The norm seems beneficial for the Ballai society.
In general, the norm seems legitimate.
N
OTE
.—Scale range in parentheses.
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
TABLE A2
STUDY 2 STIMULI
Description
Low autonomy, legitimate norm Ballai is a small, foreign city-state. Most citizens of Ballai wear white clothes on Masakha Day, the
city’s primary holiday. Masakha Day commemorates the men and women who sacrificed lives in or-
der to protect Ballai from foreign invaders. Citizens are expected to wear blue in order to honor fallen
soldiers. The act of wearing blue helps the country remember and thank those who sacrificed them-
selves for the safety of others.
Roiku recently created an advertising campaign in which the brand endorsed the norm by encouraging
consumers to wear blue on Masakha Day.
Low autonomy, illegitimate norm Ballai is a small, foreign city-state. Most citizens of Ballai wear white clothes on Masakha Day, the
city’s primary holiday. Masakha Day celebrates the glory and the power of the city’s corrupt dictator.
Citizens are expected to wear white in order to pay tribute to the dictator. The act of wearing white
helps bolster the ego and pride of the city’s corrupt leader.
Roiku recently created an advertising campaign in which the brand endorsed the norm by encouraging
consumers to wear blue on Masakha Day.
High autonomy, legitimate norm Ballai is a small, foreign city-state. Most citizens of Ballai wear white clothes on Masakha Day, the
city’s primary holiday. Masakha Day commemorates the men and women who sacrificed lives in or-
der to protect Ballai from foreign invaders. Citizens are expected to wear blue in order to honor fallen
soldiers. The act of wearing blue helps the country remember and thank those who sacrificed them-
selves for the safety of others.
Roiku recently created an advertising campaign in which the brand parted from the norm by encourag-
ing consumers to wear blue on Masakha Day.
High autonomy, illegitimate norm Ballai is a small, foreign city-state. Most citizens of Ballai wear white clothes on Masakha Day, the
city’s primary holiday. Masakha Day celebrates the glory and the power of the city’s corrupt dictator.
Citizens are expected to wear white in order to pay tribute to the dictator. The act of wearing white
helps bolster the ego and pride of the city’s corrupt leader.
Roiku recently created an advertising campaign in which the brand parted from the norm by encourag-
ing consumers to wear blue on Masakha Day.
TABLE A3
STUDY 4A STIMULI
Description
Low autonomy Baraccio/Ero/Setia/Solle [hereafter designated as X] follows the market to design coats/sunglasses/shoes/watches
[hereafter designated as products] that fit mainstream tastes. There is nothing atypical or controversial about X
products.Xis concerned with gaining the approval of mainstream consumers and tries hard to follow the norm so
it will be liked and accepted.
Moderate autonomy Xusually conforms to popular styles. That said, the brand is not a slave to convention. Xoccasionally does its own
thing and designs products that are slightly unusual. However, the brand is careful never to stray too far from the
expectations and tastes of the average consumer.
High autonomy Xoften defies industry standards by creating products that are edgy and unique. Xis irreverent and sometimes re-
bellious. Although it usually does its own thing, Xis careful not to produce products that are so deviant that they
seem strange or extremely odd.
Extreme autonomy Xis a rebellious and controversial brand. Their products are radically different than other brands. Xshows contempt
for rules and a complete disregard for marketplace opinion. Xand its employees do what they want whether or not
it pleases others.
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 561
TABLE A4
STUDY 4B STIMULI
Description
Low autonomy Amber Johnson understood that society expects people to display “typical” manners, engage (and not engage) in
certain behaviors, and pursue particular types of goals. Amber was well aware of society’s code, and she always
conformed to it. She rarely would assert her independence or do her own thing. For example, Amber never wore
unusual hairstyles or dressed differently than others. Although there were times she disagreed with the govern-
ment, her boss, or her parents, she didn’t protest against these or other authorities. After she finished college,
Amber moved to the city where she has become part of a larger community and regularly interacts with other
people.
Moderate autonomy Amber Johnson understood that society expects people to display “typical” manners, engage (and not engage) in
certain behaviors, and pursue particular types of goals. Amber was well aware of society’s code, and she usually
conformed to it. Although there were times she disagreed with the government, her boss, or her parents, she
didn’t protest against these or other authorities. However, Amber occasionally would assert her independence and
do her own thing. For example, she often wore unusual hairstyles and dressed differently than others. After she
finished college, Amber moved to the city where she has become part of a larger community and regularly inter-
acts with other people.
High autonomy Amber Johnson understood that society expects people to display “typical” manners, engage (and not engage) in
certain behaviors, and pursue particular types of goals. Amber was well aware of society’s code, and she rarely
conformed to it. She usually would assert her independence and do her own thing. For example, Amber often
wore unusual hairstyles and dressed differently than others. There were times she disagreed with the government,
her boss, or her parents and would engage in protests against these or other authorities. After she finished col-
lege, Amber moved to the city where she has become part of a larger community and regularly interacts with other
people.
Extreme autonomy Amber Johnson understood that society expects people to display “typical” manners, engage (and not engage) in
certain behaviors, and pursue particular types of goals. Amber was well aware of society’s code, and she never
conformed to it. She always would assert her independence and do her own thing. Amber often wore unusual
hairstyles and dressed differently than others. There were times she disagreed with the government, her boss, or
her parents and would engage in protests against these or other authorities. After she finished college, Amber
moved into the wilderness where she now lives in isolation and avoids interaction with other people.
REFERENCES
Aggarwal, Pankaj, and Ann L. McGill (2007), “Is That Car Smiling
at Me? Schema Congruity as a Basis for Evaluating Anthro-
pomorphized Products,” Journal of Consumer Research,34
(December), 468–79.
Allison, Anne (2009), “The Cool Brand, Affective Activism and
Japanese Youth,” Theory, Culture and Society, 26 (March),
89–111.
Amabile, Teresa M. (1982), “Social Psychology of Creativity: A
Consensual Assessment Technique,” Journal of Personality
and Social Psychology, 43 (December), 997–1013.
Ariely, Dan, and Jonathan Levav (2000), “Sequential Choice in
Group Settings: Taking the Road Less Traveled and Less En-
joyed,” Journal of Consumer Research, 27 (December), 279–
90.
Aronson, Elliot (2008), The Social Animal, New York: Worth.
Belk, Russell W., Kelly Tian, and Heli Paavola (2010), “Consum-
ing Cool: Behind the Unemotional Mask,” in Research in
Consumers Behavior, Vol .12, ed. Russel W. Belk, Bingley:
Emerald, 183–208.
Bellezza, Silvia, Francesca Gino, and Anat Keinan (2014), “The
Red Sneakers Effect: Inferring Status and Competence from
Signals of Nonconformity,” Journal of Consumer Research,
41 (June), 35–54.
Berger, Jonah (2008), “Identity Signaling, Social Influence, and
Social Contagion,” in Peer Influence Processes among Youth,
ed. M. J. Prinstein and K. A. Dodge, New York: Guilford.
Berger, Jonah, and Chip Heath(2007), “Where Consumers Diverge
from Others: Identity Signaling and Product Domains,” Jour-
nal of Consumer Research, 34 (August), 121–34.
Bird, Sarah, and Alan Tapp (2008), “Societal Marketing and the
Meaning of Cool,” Social Marketing Quarterly, 14 (March),
18–29.
Blanton, Hart, and Charlene Christie (2003), “Deviance Regula-
tion: A Theory of Action and Identity,” Review of General
Psychology, 7 (June), 115–49.
Brewer, Marilynn B. (1991), “The Social Self: On Being the Same
and Different at the Same Time,” Personality and Social Psy-
chology Bulletin, 17 (October), 475–82.
Brooks, David (2000), Bobos in Paradise: The New Upper Class
and How They Got There, New York: Simon & Schuster.
Burroughs, James E., and David Glen Mick (2004), “Exploring
Antecedents and Consequences of Consumer Creativity in a
Problem-Solving Context,” Journal of Consumer Research,
31 (September), 402–11.
Campbell, Margaret C., and Ronald C. Goodstein (2001), “The
Moderating Effect of Perceived Risk on Consumers’ Evalu-
ations of Product Incongruity: Preference for the Norm,”Jour-
nal of Consumer Research, 28 (December), 439–49.
Cialdini, Robert B., Raymond R. Reno, and Carl A. Kallgren
(1990), “A Focus Theory of Normative Conduct: Recycling
the Concept of Norms to Reduce Littering in Public Places,”
Journal of Personality and Social Psychology, 58 (6), 1015–
26.
Cialdini, Robert B., and Melanie R. Trost (1998), “Social Influence:
Social Norms, Conformity and Compliance,” in The Hand-
book of Social Psychology, Vols. 1 and 2, 4th ed., ed. D. T.
Gilbert, S. T. Fiske, and G. Lindzey, New York: McGraw-
Hill, 151–92.
Connor, Marlene K. (1995), What Is Cool? Understanding Black
Manhood in America, New York: Crown.
Danesi, Marcel (1994), Cool: The Signs and Meanings of Ado-
lescence, Toronto: University of Toronto Press.
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
562 JOURNAL OF CONSUMER RESEARCH
Dar-Nimrod, Ilan, I. G. Hansen, T. Proulx, and D. R. Lehman
(2012), “Coolness: An Empirical Investigation,” Journal of
Individual Differences, 33 (3), 175–85.
Fehr, Ernst, and Urs Fischbacher (2003), “The Nature of Human
Altruism,” Nature, 425 (October), 785–91.
Florida, Richard (2002), “Bohemia and Economic Geography,”
Journal of Economic Geography, 2 (1), 55–71.
Frank, Thomas (1997), The Conquest of Cool: Business Culture,
Counterculture, and the Rise of Hip Consumerism, Chicago:
University of Chicago Press.
Giesler, Markus (2008), “Conflict and Compromise: Dramain Mar-
ketplace Evolution,” Journal of Consumer Research,34
(April), 739–53.
Gladwell, Malcolm (1997), “The Coolhunt,” New Yorker, March
17, 78–87.
Grossman, Lev (2003), “The Quest for Cool,” Time, September 8,
162.
Gurrieri, Lauren Kate (2009), “Cool Brands: A Discursive Identity
Approach,” in ANZMAC 2009: Sustainable Management and
Marketing Conference Proceedings, ed. Dewi Tojib, Mel-
bourne: Australia and New Zealand Marketing Academy.
Heath, Joseph, and Andrew Potter (2004), Nation of Rebels:
Why Counterculture Became Consumer Culture, New York:
HarperCollins.
Hebdige, Dick (1979), Subculture, the Meaning of Style, London:
Methuen.
Hebdige, Dick, and Andrew Potter (2008), “A Critical Reframing
of Subcultural Cool and Consumption,” in European Ad-
vances in Consumer Research, Vol. 8, ed. Stefania Borghini,
Mary Ann McGrath, and Cele Otnes, Duluth, MN: Associ-
ation for Consumer Research, 527–28.
Hobbes, Thomas (1651/1991), Leviathan, New York: Cambridge
University Press.
Hogan, Patrick Colm (2001), The Culture of Conformism: Un-
derstanding Social Consent, Durham, NC: Duke University
Press.
Hollander, E. P. (1958), “Conformity, Status, and Idiosyncrasy
Credit,” Psychological Review, 65 (March), 117–27.
Holt, Douglas B. (2004), How Brands Become Icons: The Prin-
ciples of Cultural Branding, Cambridge, MA: Harvard Busi-
ness School Press.
Interbrand (2014), “Best Global Brands of 2013,” http://www.inter
brand.com/en/best-global-brands/2013/top-100-list-view.aspx.
Jhang, Ji Hoon, Susan Jung Grant, and Margaret C. Campbell
(2012), “Get It? Got It. Good! Enhancing New Product Ac-
ceptance by Facilitating Resolution of Extreme Incongruity,”
Journal of Marketing Research, 49 (April), 247–59.
Jones, Edward E., and Keith E. Davis (1965), “From Acts to Dis-
positions: The Attribution Process in Person Perception,” in
Advances in Experimental Social Psychology, Vol. 2, ed. L.
Berkowitz, New York: Academic Press, 219–66.
Judd, Charles M., David A. Kenny, and Gary H. McClelland
(2001), “Estimating and Testing Mediation and Moderation
in Within-Subject Designs,” Psychological Methods, 6 (June),
115–34.
Keizer, Kees, Siegwart Lindenberg, and Linda Steg (2008), “The
Spreading of Disorder,” Science, 322 (December), 1681–85.
Kerner, Noah, and Gene Pressman (2007), Chasing Cool: Standing
out in Today’s Cluttered Marketplace, New York: Atria.
Koslow, Scott, Sheila L. Sasser, and Edward A. Riordan (2003),
“What Is Creative to Whom and Why? Perceptions in Ad-
vertising Agencies,” Journal of Advertising Research,43
(March), 96–110.
Leland, John (2004), Hip, the History, New York: Ecco.
MacAdams, Lewis (2001), Birth of the Cool: Beat, Bebop, and
the American Avant-garde, New York: Free Press.
Mailer, Norman (1957), The White Negro, San Francisco: City
Lights.
Marcuse, Herbert (1955), Eros and Civilization: A Philosophical
Inquiry into Freud, Boston: Beacon.
Markus, Hazel Rose, and Barry Schwartz (2010), “Does Choice
Mean Freedom and Well-Being?” Journal of Consumer Re-
search, 37 (August), 344–55.
McCracken, Grant (1986), “Culture and Consumption: A Theo-
retical Account of the Structure and Movement of the Cultural
Meaning of Consumer Goods,” Journal of Consumer Re-
search, 13 (June), 71–84.
Meyers-Levy, Joan, and Alice M. Tybout (1989), “Schema Con-
gruity as a Basis for Product Evaluation,” Journal of Con-
sumer Research, 16 (June), 39–54.
Milgram, Stanley (1963), “Behavioral-Study of Obedience,” Jour-
nal of Abnormal Psychology, 67 (October), 371–78.
Moreau, C. Page, and Darren W. Dahl (2005), “Designing the
Solution: The Impact of Constraints on Consumers’ Creativ-
ity,” Journal of Consumer Research, 32 (June), 13–22.
Nancarrow, Clive, Pamela Nancarrow, and Julie Page (2002), “An
Analysis of the Concept of Cool and Its Marketing Impli-
cations,” Journal of Consumer Behaviour, 1 (June), 311–22.
O’Donnell, Kathleen A., and Daniel L. Wardlow (2000), “A Theory
on the Origins of Coolness,” in Advances in Consumer Re-
search, Vol. 27, ed. Stephen J. Hoch and Robert J. Meyer,
Provo, UT: Association for Consumer Research, 13–18.
Oyserman, Daphna, Heather M. Coon, and Markus Kemmelmeier
(2002), “Rethinking Individualism and Collectivism: Evalu-
ation of Theoretical Assumptions and Meta-analysis,” Psy-
chological Bulletin, 128 (January), 3–72.
Pountain, Dick, and David Robins (2000), Cool Rules: Anatomy
of an Attitude, London: Reaktion.
Preacher, Kristopher J., and Andrew F. Hayes (2008), “Asymptotic
and Resampling Strategies for Assessing and Comparing In-
direct Effects in Multiple Mediator Models,” Behavior Re-
search Methods, 40 (3), 879–91.
Rodkin, Philip C., Thomas W. Farmer, Ruth Pearl, and Richard
Van Acker (2006), “They’re Cool: Social Status and Peer
Group Supports for Aggressive Boys and Girls,” Social De-
velopment, 15 (2), 176–204.
Schachter, Stanley (1951), “Deviation, Rejection, and Communi-
cation,” Journal of Abnormal and Social Psychology,46
(April), 190–207.
Schultz, P. Wesley, Jessica M. Nolan, Robert B. Cialdini, Noah J.
Goldstein, and Vladas Griskevicius (2007), “The Construc-
tive, Destructive, and Reconstructive Power of Social Norms,”
Psychological Science, 18 (5), 429–34.
Southgate, Nick (2003), “Coolhunting, Account Planning and the
Ancient Cool of Aristotle,” Marketing Intelligence and Plan-
ning, 21 (7), 453–61.
Sumner, William Graham (1906), Folkways, Boston: Ginn.
Thompson, Craig J. (2014), “The Politics of Consumer Identity
Work,” Journal of Consumer Research, 40 (5), iii–vii.
Thompson, Craig J., and Zeynep Arsel (2004), “The Starbucks
Brandscape and Consumers’ (Anticorporate) Experiences of
Glocalization,” Journal of Consumer Research, 31 (Decem-
ber), 631–42.
Thompson, Craig J., and Gokcen Coskuner-Balli (2007), “Coun-
tervailing Market Responses to Corporate Co-optation and
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
WARREN AND CAMPBELL 563
the Ideological Recruitment of Consumption Communities,”
Journal of Consumer Research, 34 (August), 135–52.
Thornton, Sarah (1996), Club Cultures: Music, Media, and Subcultural
Capital, Hanover, NH: University Press of New England.
Van Osselaer, Stijn M. J., Suresh Ramanathan, Margaret C. Camp-
bell, Joel B. Cohen, Jeannette K. Dale, Paul M. Herr, Chris
Janiszewski, Arie W. Kruglanski, Angela Y. Lee, Stephen J.
Read, J. Edward Russo, and Nader Tavassoli (2005), “Choice
Based on Goals,” Marketing Letters, 16 (December), 335–
46.
Walker, Rob (2003), “The Marketing of No Marketing,” New York
Times, June 22.
This content downloaded from 128.138.73.68 on Tue, 19 Aug 2014 16:54:02 PM
All use subject to JSTOR Terms and Conditions
... Perceived brand statussuch as either niche or mainstreamaffects retail sales performance (Hoskins and Griffin, 2020) and brand love, since a meaningful set of perceptions influences how consumers relate to brands and inspires what they desire to own (Ahuvia et al., 2020). In consumption settingssuch as observed with the choice of touristic destinationhigh technology and socially visible categories, brand love and attitudinal outcomes are related to a sense of coolness (Apaolaza et al., 2021;Kock, 2021;Loureiro and Blanco, 2021;Tiwari et al., 2021), which means a socially constructed concept which constitutes a favorable feature by itself (Warren and Campbell, 2014). ...
... Over 70 different ways to define the term coolness are available (Warren and Campbell, 2014). Conceptually, the definition of coolness refers to a dynamic and positive trait, attributed to autonomous cultural objects (Warren et al., 2019), which diverge from (illegitimate) norms. ...
... Conceptually, the definition of coolness refers to a dynamic and positive trait, attributed to autonomous cultural objects (Warren et al., 2019), which diverge from (illegitimate) norms. When applied to brands, coolness is defined by a set of characteristics used to assess a brand's attractiveness, encompassing four essential features: assuming a positive valence, being subjective, autonomous and dynamic (Warren and Campbell, 2014). Often employed as an indicator of success of a brand (Swaminathan et al., 2020), the cool factor is frequently associated with product/brand adoption, as described in literature dedicated to consumer-brands relationships (e.g. ...
Article
Full-text available
Purpose The current paper aims to explore how brand coolness can mediate the relationship between tactical green marketing orientation (GMO) and willingness to pay (WTP), by exploring the differences between two global brands with opposite green marketing perceptions. Design/methodology/approach Based on the stimuli-organism-response (S-O-R) framework, the relation between tactical GMO, brand coolness and consumer's WTP is examined through a survey with 272 participants, who gave their perceptions about two different brands regarding their green orientations: British Petroleum (BP) and L'Oréal. The variable set was adapted and validated through focus group sessions. Findings Brand coolness is found to mediate the impact GMO on WTP and, for both brands, green marketing does affect the extent to which brands are perceived as cool by consumers. More importantly, evidence shows that only in the case of the “green brand” (e.g. L'Oréal), the impact on WTP is significant, which offers new implications regarding the outcomes of companies' pro-environmental policies. Originality/value This study is the first to investigate the outcomes of GMO over consumer's intentions (WTP) and the role of brand perceptions (coolness). The effects are compared between two global brands, with significantly different perceptions on their environmental sustainability.
... This feature, although pursued, does not have a clear meaning. However, it has some properties that already seem clear (Warren & Campbell, 2014). The first one lies on the social construction of this meaning, i.e., it is a feature attributed by a given audience (Belk et al., 2010). ...
... Social objects (which can be either people or things) considered cool change over time and between groups. Consumers with similar backgrounds and interests tend to agree on the perception of coolness (Warren & Campbell, 2014). The third one regards the positive (desirability) nature of the coolness feature (Belk et al., 2010). ...
... The third one regards the positive (desirability) nature of the coolness feature (Belk et al., 2010). Finally, cool is not a different way of saying "good", since it has an implicit cultural background (Warren & Campbell, 2014). Warren and Campbell (2014) advocated that the autonomy over other individuals' standards, beliefs and expectations is what differentiates "good" from "cool". ...
Article
Full-text available
Objective: This paper reports the results of a research that aimed to understand the articulation between sociocultural practices and symbolic processes in the construction of a cool identity by young people inserted in the funk environment. Method: Fieldwork focused on participant observation at funk events and dances and in-depth interviews with 3 MCs and 26 young people. To analyze the data, the content analysis methodology proposed by Bardin (1995) was used. Results: Among the results, it is noteworthy that, because funk is part of the daily lives of young people, their diverse experiences with funk enhance the creation of a cool identity that makes these individuals find a place in this universe of signs and meanings. Originality: The article discusses the concept of cool identity in order to understand the (re)construction of the identity of a group of urban young people supported by and in consumption.
... 2.3 | Intrinsic attributes of AR experience: The mediating role of perceived coolness In modern consumption, perceived coolness has increasingly become an important dimension of user experience (Raptis et al., 2017). It is a positive perception based on users' multidimensional judgment, meaning the user's perception of uniqueness, attractiveness, and subcultural appeal (Sundar et al., 2014;Warren & Campbell, 2014). ...
... Perceived coolness is an immediate and overall judgment of consumers' technological experience, representing whether they are attracted in the first place and perceive uniqueness and desirability (Raptis et al., 2017). It may be triggered by novel experiences, and it covers richer emotional connotations, including autonomy (using AR to control the presentation of products autonomously) and uniqueness (personalized display interface) (Chen & Chou, 2019;Warren & Campbell, 2014). According to Javornik et al. (2016) in situ study of "magic mirror" (providing AR-based virtual try-on of cosmetics), more than 70% of the participants exclaim how cool it is. ...
Article
Full-text available
Augmented reality (AR) is increasingly coming into the spotlight for its potential to improve the consumer experience through the creation of presence. This research aims to explore the theoretical mechanisms through which AR‐based product presentation influences online store attractiveness and whether the effects differ in the purchasing contexts of hedonic and utilitarian product types. From the perspective of consumers’ bi‐dimensional experience, we find that AR increases online store attractiveness by creating perceived coolness (intrinsic attribute) and spatial presence experience (extrinsic attribute), which would further have a positive impact on consumers’ purchase intention. The mediating mechanisms are different in purchasing contexts of various product types: for hedonic products, perceived coolness and spatial presence are parallel mediating factors leading to the improvement of online store attractiveness; while for utilitarian products, only the mediation effect of spatial presence presents. Our findings enrich the literature on AR marketing by proposing an insightful mediating force (i.e., perceived coolness) to complement the effect of presence, and explore the different purchasing contexts. We also provide managerial guidance for e‐retailers to differentiate AR interface design for diverse product types to apply AR technology effectively.
... Cool has majorly been associated with subcultural and countercultural values (Ferguson, 2011) which promote rebelliousness, independence, autonomy and moral deviance (Warren and Campbell, 2014;Dar-Nimrod et al., 2018;Bird and Tapp, 2008). Cool has been found to determine consumers' music preferences (Dar-Nimrod et al., 2012), fashion choices (Dalton and Wang, 2014), technology orientations (Bruun et al., 2016) and even socially undesirable behaviors such as smoking and drugging (Bird and Tapp, 2008;Dalton and Wang, 2014). ...
... Thus, cool can be conceptualized as an identity-based construct that signals group memberships (Oyserman, 2009). Countercultural groups often tend to question and rebel against the institutional authority (Pountain and Robins, 2000;Bird and Tapp, 2008) and, thus, are likely to engage in cool consumption to signal divergence and autonomy (Warren and Campbell, 2014). Cool is, thus, a rebellious attitude that seeks to challenge the conventions of society (Pountain and Robins, 2000). ...
Article
Purpose This paper aims to theorize counterfeit luxury consumption among millennials from a generational identity perspective. Design/methodology/approach The paper proposes and tests a model of counterfeit buying behavior using an online survey of 467 millennial respondents. The study uses multi-item measures from the extant literature and uses the structural equation modeling technique to test the proposed hypotheses. Findings The findings reveal when millennials have a self-defining relationship with their generation, they tend to internalize the generational norm pertaining to counterfeit luxury consumption. Millennials’ counterfeit related values: market mavenism, postmodernism, schadenfreude and public self-consciousness contribute to their generational identity. Moreover, market mavenism, cool consumption and public self-consciousness establish counterfeit luxury consumption as a generational norm. Practical implications The findings of this paper suggest that the expertise and influence of market mavens can be used to deter counterfeit consumption. Moreover, luxury brands must communicate a cool image to offset the rebellious image of counterfeits. Further, from a standardization versus adaption standpoint, the generational perspective allows for the standardization of anti-counterfeiting campaigns. Originality/value The paper makes a novel contribution to the counterfeiting literature by demonstrating that millennials pursue counterfeit luxury brands when they pledge cognitive allegiance to their generation. The paper, thus, extends the identity perspective of counterfeit luxury consumption to group contexts. The authors also test and validate the role of descriptive norms in group contexts by introducing the construct generational norm to counterfeiting literature.
... The meaning of coolness is seen as subjective, of positive valence, autonomous and dynamic (Warren and Campbell 2014;Anik et al. 2017). The brand coolness concept shows a positive multi-attribute association with ten-dimension (Warren et al. 2019;Loureiro, Jimenez, and Romero 2020), namely (1) useful/extraordinary, the brand offers high quality and tangible benefits (e.g., Belk, Tian, and Paavola 2010;Dar-Nimrod et al. 2012); (2) aesthetically appealing, the brand offers attractive and designs different from its competitors (Bruun et al. 2016); (3) energetic, the brand is perceived as being active, outgoing and youthful (Runyan, Noh, and Mosier 2013); (4) original, an original brand is a creative, unique brand (Runyan, Noh, and Mosier 2013); (5) authentic, the brand has a set of true authentic values from production to customer care (Biraglia and Brakus 2020); (6) rebellious, the brand strives to set apart by being nonconformist (Potter and Heath 2004); (7) high-status, the brand is associated with prestige and sophistication (Nancarrow, Nancarrow, and Page 2003;Loureiro, Jimenez, and Romero 2020);(8) subcultural, cool brands are often associated with groups of people who are perceived to work independently from mainstream society (Warren et al. 2019;Sundar, Tamul, and Wu 2014); (9) iconic, the brand reflects the individuals' values and beliefs, being recognized as a cultural symbol (Holt 2004); and (10) popular, a cool brand is perceived to be fashionable, trendy, and liked by most people (Dar-Nimrod et al. 2012;Warren et al. 2019). ...
Article
This study aims to explore the moderator role of popular and iconic coolness dimensions on the relationship between hedonic versus utilitarian beauty product brands and high-status perceptions, using internet memes as stimuli. An experimental study was conducted to analyse whether two dimensions of brand coolness (popular and iconic) moderate the relationship between type of internet meme (utilitarian versus hedonic) and high-status. After conducting a pre-test, two internet memes were created for each condition, utilitarian and hedonic. In total, 428 completely answers were collected from an online MTurk panel, and the hypotheses were tested using moderation analysis. The results indicate that (i) hedonic brands are perceived as being high-status in the presence of both moderators (iconic and popular); (ii) utilitarian brands can be associated with high-status perceptions, if moderated by the popular dimension. Findings demonstrate that the popularity of the brand plays an important role in consumers perceptions. This study contributes to the marketing literature by analysing the relationship between three core dimensions of brand coolness, namely, iconic, popular, and high-status, regarding brands associated with hedonics and utilitarian products.
Article
The absence of anthropology from cognitive science may be related to the fact that the Western-dominated cognitive approach is not human-centered. From an anthropological perspective, we propose a possible social cognitive model—the human-centered social cognitive chain. It is a theory of the human cognitive process that includes two phenomena— anthropomorphism and dehumanization; has two directions—upward and downward; involves two dimensions—morality and competence; and produces two results—cuteness and coolness. It is a way for humans to understand the whole world with themselves as starting point and end point, governing social cognition. The human-centered social cognitive chain involves gods, humans, animals, plants, artificial objects, pure objects, and even abstract concepts, on the basis of which upward and downward anthropomorphism and upward and downward dehumanization arise. The existing research on anthropomorphism and dehumanization is slightly biased in that it only emphasizes upward anthropomorphism and downward dehumanization. Theoretically, both anthropomorphism and dehumanization should be two-stage processes based on essentialism, but humans engaged in cognition tend to adopt a dimensional view that simplifies them. The morality-competence dichotomy reflects this tendency and results in the cognitive outcomes of cuteness, where morality is valued higher than competence and coolness, where competence is valued higher than morality. The human-centered social cognitive chain is the paradigm by which humans understand themselves, their culture, and the world.
Article
Purpose This paper aims to explore how motivations to stand out and fit in through consumption affect consumers’ perceptions of subcultural and popular brand coolness. Importantly, how do perceptions of brand coolness affect consumers’ formations of hot, emotional brand attachments and their willingness to pay more? Design/methodology/approach This study incorporates survey data from consumers regarding cool brands. A structural equation modeling approach is used to assess the relationship between the variables of interest. Findings Susceptibility to influence is positively related to desire for unique consumption. While this desire may be fulfilled by both subcultural and popular perceptions of brand coolness, only subcultural coolness has a positive relationship with the willingness to pay more. The importance of an emotional brand attachment is established between both dimensions of brand coolness and price premiums. Research limitations/implications This study is based on cross-sectional survey data. As brand coolness is often transitory, longitudinal research on trends focusing on different elements of brand coolness may shed light on the cool brand lifecycle. Practical implications Firms wanting to position brands as cool should emphasize how the brand can help consumers stand out. If a cool brand is already well-known, resources should be allocated to building hot, emotional attachments to command price premiums. Originality/value This research contributes to a nascent body of literature empirically exploring brand coolness. It builds on past literature that notes the tension between standing out and fitting in conceptualizations of coolness by assessing individual differences. Significantly, it examines specific attributes of cool brands to explore the differences in how subcultural and popular perceptions of brand coolness relate to important marketing outcomes.
Article
Purpose This study aims to investigate the effect of color temperature on consumption. Color is one of the most powerful elements of sensory marketing. However, how warm and cool colors drive consumer indulgence and interact with other visual cues is minimally understood. Design/methodology/approach This research conducts six experiments to investigate the effect of eight warm and cool colors and the effect of warm/cool color’s interaction with reflectance on indulgent consumption/use in various retail environments. Findings Studies 1a and 1b support the contrasting effects of warm vs cool colors on consumers’ indulgent consumption. Studies 2a and 2b establish the serial mediating role of arousal and self-reward focus in the color-indulgence relationship. Study 3a demonstrates the interactive effect of warm (vs cool) colors and glossy (vs matte) reflectance on consumer indulgence, and Study 3b confirms how glossy (vs matte) reflectance moderates the serial mediating effect of arousal and self-reward focus in the color-indulgence relationship. Research limitations/implications This research contributes to the growing stream of research on the visual aspect of sensory marketing, especially color, and advances the theoretical knowledge of how color could be used effectively to influence consumer indulgence. Practical implications This research provides actionable managerial implications on the effective use of warm and cool colors and glossy and matte reflectance to influence consumer indulgence. Originality/value This research advances the theoretical and empirical knowledge of color’s interaction with other visual sensory cues and the underlying psychological processes shaping consumer indulgence.
Article
In the past decades, as traditional luxury goods and conspicuous consumption have become more mainstream and lost some of their signaling value, new alternative signals of status (e.g., vintage, inconspicuous consumption, sustainable luxury) have progressively emerged. This research applies the grounded theory method to establish a novel framework that systematically unifies existing conceptualizations, findings, and observations on alternative signals of status. The proposed framework organizes alternative signals in terms of their distance from traditional status symbols and categorizes them along six focal dimensions: time (new/old), quantity (many possessions/few possessions), conspicuousness (conspicuous/inconspicuous), aesthetics (beautiful/ugly), culture (highbrow/lowbrow), and pace of life (slow/fast). This parsimonious framework captures various consumption phenomena related to status signaling, unifies existing theories, and generates a fruitful agenda for future research.
Article
Purpose This study aims to investigate coolness – a hedonic attribute – in online consumer reviews (OCRs) by exploring the relationship between OCR framing, presence of pictorial images and OCR-perceived coolness. It also demonstrates how reviewers and brands can create or identify cool OCRs. Design/methodology/approach Two studies – one experimental and the other using archival data from an OCR platform – were used to test three hypotheses on the effect of OCR framing and pictorial images on perceived OCR coolness. Findings The results reveal that OCRs framed with negative words but reflecting positive views about products are perceived as cooler than OCRs framed with positive words and reflecting positive views. OCRs with pictorial images are perceived to be cooler than those without pictorial images. Originality/value Studies on coolness have focused on people and products but not message content. This research links a message’s framing and pictorial images to its perceived coolness in the OCR context. It provides practical suggestions to marketers, coolhunters and individuals interested in creating cool message content.
Article
Full-text available
Creativity is an underresearched topic in consumer behavior, yet integral in many instances of consumer problem solving. Two experiments were conducted to investigate antecedents and consequences of creativity in a consumption context. The results indicate that both situational factors (i.e., time constraints, situational involvement) and person factors (i.e., locus of control, metaphoric thinking ability) affect creative consumption and that some of these variables have interactive influences. The results also show that acting creatively enhances positive affect.
Article
Predicting the future is hard work. Ask any professional trend spotter: it takes insight, dedication--and secret armies of superhip teenagers
Book
“Reflecting on the principles of government, David Hume was struck by the ‘implicit submission’ of people to the rulers in all societies even though ‘force is always on the side of the governed,’ a surprising outcome that he attributes to control of opinion. Patrick Hogan greatly enriches the Humean concept, unraveling an intricate web of ‘modes of action and thought’ that ensnare the governed in forms of ‘implicit submission,’ contrary to their deeper motives and ideals and long-term interests. His goal is not merely to explain, but to provide tools of understanding that will be of practical value to those who struggle for justice and freedom. Drawing from an impressive array of sources, his valuable study advances both ends considerably, no mean accomplishment.” Noam Chomsky, Massachusetts Institute of Technology “This is a sophisticated examination of why masses of people allow injustice (in the form of racism, sexism, homophobia, and economic oppression, among others) to occur . . . . Hogan takes a multilayered approach (political, social, economic, and cognitive) in his thoughtful, well-documented discussions of the wild disparities that exist and will continue to exist if left unchecked among different elements of the world’s populations. The prognosis is not entirely bleak, however; Hogan is hopeful that readers will tap into the source list of advocates for change included at the end of this excellent book. Highly recommended.” Ellen D. Gilbert, Rutgers University, New Brunswick
Article
Marketers face an enormous challenge these days reaching finicky, ad-zapping consumers. In this fifth annual ranking of the 100 most valuable brands, Businessweek and Interactive Corp. show which ones are succeeding.
Article
Marketing practitioners have long understood the importance of identifying emerging trends for products which are "cool." This paper posits a theory that "coolness" originates in the fluctuating discrepency between actual and ideal selves in early adolescence (narcissistic vulnerability), which motivates teens to reduce this drive through strategies of peer-group affiliation. Within-group semiotic codes evolve (signifying osmosis) which function to maintain a group identity. These within-group codes determine what is "cool" and "uncool" within the group. Aggregating commonalities across groups results in a metacode of coolness which is amenable to diffusion through the general population.