ABSTRACT
This study examines the behavioral intentions of heterosexual and lesbian, gay, and bisexual (LGB) consumers following exposure to advertising featuring heterosexual or LGB imagery. Results indicated that viewing advertisements congruent with participants' sexuality increased the advertisements' appeal, particularly for LGB consumers. Advertising appeal significantly predicted purchase intentions and likelihood of recommending the advertised brands. Incongruent advertisements generally still were appealing for all participants. The indirect effect of advertising imagery on behavioral intentions through advertising appeal was stronger for LGB participants than heterosexual participants. Advertising appeal was related to behavioral intentions three weeks postexposure, suggesting potential lasting effects of advertisement exposure.
MANAGEMENT SLANT
Lesbian, gay, and bisexual (LGB) consumers are highly responsive to LGB-inclusive advertising with respect to both purchase intentions and likelihood to recommend.
Consumers' perception of an advertisement's appeal was the strongest predictor of lasting effects on behavioral intentions, regardless of the sexual orientation of the models depicted in the advertisement or the congruency between the advertisement and sexual orientation of the consumer.
Heterosexual consumers no longer may experience the same aversion to LGB-inclusive advertisements that consistently was found in previous studies. Heterosexual consumers place more weight on the advertisement being interesting, trustworthy, and a good match with the brand than on the sexuality of the models depicted within the advertisement.
Recognition as a corporate ally provides a significant benefit toward building relationships with LGB consumers.
INTRODUCTION
As part of its “Never Hide” campaign, the sunglasses brand Ray-Ban ran a print advertisement depicting two well-dressed men holding hands as they walked down a busy street. The Las Vegas hotel Luxor portrayed two men lying on a bed seductively looking into one another's eyes in an advertisement from the “Get Your Lux On” campaign. Tylenol featured two gay men and their adopted children in a television commercial connected to its “How We Family” campaign.
These advertisements exemplify a growing trend of depicting lesbian, gay, and bisexual (LGB) individuals in advertising. LGB-inclusive advertising likely is driven by the belief that LGB individuals are a valuable segment of the consumer market to target. LGB consumers' buying power is predicted to extend well beyond $900 billion a year (Nelson and Mitchell, 2018), and LGB consumers are fiercely loyal to brands that project LGB acceptance and inclusion (Oakenfull and Greenlee, 2004). The perception that LGB consumers have significant spending power and strong brand loyalties has cultivated their “dream market” status and rationalized LGB-inclusive advertising.
Although LGB-inclusive advertisements are increasing in frequency, scholarship within this domain is still in its infancy. Much of the marketing literature on LGB-inclusive advertising focuses on how to target the LGB market segment without alienating heterosexual consumers. The extant literature acknowledges that heterosexual consumers generally dislike LGB-inclusive advertisements and that LGB consumers respond favorably (Bhat et al., 1998; Hester and Gibson, 2007; Oakenfull and Greenlee, 2004, 2005; Um, 2014).
Although it seems that LGB imagery in advertising can influence consumers' attitudes, a thorough review of the literature uncovered a limited number of articles exploring how LGB-inclusive advertisements affect the behavioral intentions of both heterosexual and LGB consumers. Recent research measuring behavioral intentions has relied primarily on advertisements created specifically for the experimental design, thereby lessening the ecological validity of the study (Pounders and Mabry-Flynn, 2016), or has included only heterosexual participants, thereby eliminating the ability to make comparisons between heterosexual and LGB consumers (Hooten et al., 2009).
The cultural fabric of American society has evolved since a majority of the prior literature in this area was published. Since the publication of important research on LGB-inclusive advertising (Oakenfull and Greenlee, 2004, 2005), for example, the United States repealed the anti-LGB “Don't Ask, Don't Tell” military policy, legalized same-sex marriage, and has seen public favorable opinion about LGB acceptance shift from 42 percent in 2004 to 67 percent in 2018 (Gallup, 2018). Although earlier studies found that LGB consumers were an insular group focusing their attention on LGB-specific media outlets (Burnett, 2000), recent scholarship has found greater variance among the media consumed by the LGB market segment (Um, 2014). The changing sociocultural environment that consumers and marketers must navigate justifies a robust and continued reevaluation of LGB-inclusive advertising among both heterosexual and LGB consumers (Pounders and Mabry-Flynn, 2016).
The current study contributes to marketing research on LGB-inclusive advertising in several novel and notable ways. First, it moves beyond attitudes by measuring behavioral intentions related to both purchasing and recommending the advertised brands. Second, the researchers included both heterosexual and LGB consumers to make cross-orientation comparisons. Third, the researchers used real marketing campaigns and presented them to the participants in the context of a magazine to maximize ecological validity. Fourth, the researchers contacted participants three weeks postexposure to examine effects of advertising exposure after a temporal delay.
LITERATURE REVIEW
LGB-Inclusive Advertisements
The majority of scholarship exploring LGB imagery in advertising seeks to understand how heterosexual consumers interpret and perceive LGB-inclusive advertisements. Heterosexual consumers, particularly men, consistently have been displeased, made uncomfortable, or offended by LGB imagery in advertisements (Hester and Gibson, 2007; Hooten et al., 2009; Oakenfull, McCarthy, and Greenlee, 2008; Um, 2014). The general aversion to LGB-inclusive advertising is not surprising given the history of homonegativity and LGB exclusion in the United States.
Corporations, including Visa and Disney, have come under fire from conservative, anti-LGB groups after promoting LGB-inclusive marketing campaigns (Gill, 1998). Case studies of companies such as Visa and Disney have created fears of alienating core heterosexual customer bases over LGB inclusion in advertisements and promotional materials. These fears have led some brands to shy away from LGB imagery in their marketing campaigns in spite of the perceived disposable income and brand loyalty attributed to LGB consumers.
Other brands, however, have earned the label of corporate ally by embracing the LGB market segment. Corporate allies are defined as corporations, brands, or products that LGB consumers perceive to be “gay friendly,” or supportive of the LGB community (Oakenfull and Greenlee, 2004). Infusing LGB inclusiveness into marketing materials and corporate actions facilitates the development of a gay-friendly brand persona that bridges a company to LGB cultures, communities, and lifestyles.
Corporate allies often have taken on social utility within the LGB community, playing a role in the political landscape and social experiences of LGB individuals (Granovetter, 1985). Levi-Strauss, for example, was able to acquire the perception of being gay friendly when it severed profitable ties with an antigay nonprofit organization for which it previously had made uniforms. Nearly instantly, Levi-Strauss became a corporate ally (Kates, 2000). Having gay-friendly brand status implies not only that the brand is open and inclusive but that the brand proactively tends to the needs of the LGB community (Tuten, 2006).
Of importance to advertising scholars and practitioners, the implications of ally status extend beyond reputation enhancement. Studies have shown that LGB consumers more likely will buy products from companies they consider allies, pay a higher price for those products, and even accept lower quality goods (Kates, 1998, 2000). More than 90 percent of LGB participants in one survey noted that they would purchase brands that sponsored the Gay Games, an international LGB-oriented sporting event (Pitts, 1998). Corporations have taken notice. In just a decade, the number of companies achieving top status as allies in the Human Rights Campaign Foundation's (2017) annual report increased from 13 to 512.
HYPOTHESIS DEVELOPMENT
Distinctiveness Theory
Advertisements featuring social-minority groups, such as LGB individuals, historically have been explored in the marketing literature through the lens of distinctiveness theory. Distinctiveness theory predicts that individuals respond particularly favorably to advertisements that feature their distinctive traits, defined as traits that place individuals in unique or niche segments of larger populations (McGuire, 1985). Such a prediction is based on the assumption that individuals are mindful and aware of the attributes that make them distinctive (Brumbaugh and Grier, 2006; Johnson and Grier, 2011).
Hispanic individuals living in a city with a majority white population, for example, responded significantly more favorably to Hispanic models in advertisements than did Hispanic individuals living in a city with a Hispanic majority, arguably because their distinctiveness was tied more closely to their ethnicity (Deshpande and Stayman, 1994). Individuals with a heightened perception of their own cultural distinctiveness even will come to prefer brands originally deemed unfavorable if those brands are advertised within the context of the individuals' own culture, because these brands cultivate identification for the consumers (Torelli et al., 2017).
Although distinctiveness theory predicts attitudinal outcomes, marketing scholars have illustrated a strong connection between heightened attitudes toward a product or brand and resulting expected purchase intention and likelihood to recommend the product or brand to others (Batra and Ray, 1986; Fishbein and Ajzen, 1975). Researchers have used the relationship between attitudes and behavioral intentions to better understand advertising tactics and techniques (Droge, 1989; Mizerski, Allison, and Calvert, 1980) and to develop brand relationships (Alpert and Kamins, 1995; Machleit, Allen, and Madden, 1993). The premises of distinctiveness theory thus could be applied not just to attitudinal outcomes but also to behavioral outcomes, such as purchase intention and likelihood to recommend.
LGB identities can be considered distinctive traits because they distinguish LGB individuals from the heterosexual majority. LGB-inclusive advertisements are deemed particularly persuasive, credible, and trustworthy by LGB consumers, given the congruency between the consumers' distinctiveness and the advertisements' messages of inclusion (Aaker, Brumbaugh, and Grier, 2000). Sexual orientation is not a distinctive trait for heterosexual consumers, given their majority status. Heterosexual imagery in advertisements is less salient and less meaningful for heterosexual consumers, conforms to societal expectations, and likely would not influence independently the persuasive appeal of advertisements in the same way that LGB imagery would for LGB consumers.
Individuals also may respond more negatively toward advertisements that feature distinctive traits with which they do not identify than toward advertisements that feature nonidentified, nondistinctive traits. When consumers are faced with advertising messaging that is unfamiliar or does not match with their preconceived expectations, they report unfavorable attitudes toward the advertisement (Brumbaugh, 2002; Dimofte, Forehand, and Deshpande, 2003). LGB imagery in advertisements likely is both unfamiliar and unexpected for heterosexual consumers (i.e., incongruent), potentially provoking negative reactions to the advertisements (Oakenfull and Greenlee, 2005).
Consumers likely have more positive reactions to congruent advertisements (those that depict their own sexuality) than to incongruent advertisements (those that depict other sexualities; Oakenfull and Greenlee, 2005). Distinctiveness theory likely would predict that LGB consumers will have the strongest positive reactions to advertisements depicting their own sexuality (i.e., congruent advertisements) and that heterosexual consumers will have the strongest negative reactions to advertisements depicting LGB sexualities (i.e., incongruent advertisements) because sexual identity is a distinctive marker for LGB consumers. LGB imagery has salience as a distinctive trait, whereas heterosexual imagery does not. H1 and H2 are based on the congruency assumptions of distinctiveness theory.
H1: There will be a direct effect of advertising congruency on behavioral intentions, such that participants in congruent conditions will report stronger behavioral intentions than participants in incongruent conditions immediately following exposure.
H2: The direct effect of advertising congruency on behavioral intentions will be stronger for LGB participants than for heterosexual participants immediately following exposure.
Advertising Appeal
The relationship between advertising congruency and behavioral intentions may be explained by the perceived appeal of the advertisement. Consumers' perceptions of advertisements have been dissected extensively in the marketing literature, given the abundance of research suggesting that advertising appeal predicts consumer behaviors (Mitchell and Olson, 1981; Spears and Singh, 2004). Heterosexual individuals with more accepting attitudes toward LGB sexualities tend to respond more favorably to LGB-inclusive advertisements than those with less accepting attitudes (Bhat, Leigh, and Wardlow, 1998; Hester and Gibson, 2007; Oakenfull and Greenlee, 2004). Heterosexuals who are accepting of LGB individuals may not perceive sexual identity to be as distinct as those who are less accepting, which results in more acceptance of LGB-inclusive advertisements (Um, 2014).
Similarly, LGB individuals who perceive advertisements as appealing may patronize the advertised brand, particularly if the advertisements include LGB imagery (Oakenfull, 2007). Ultimately, individuals who find advertisements appealing will patronize the brands advertised (Brown, 1991; Olney, Holbrook, and Batra, 1991). The researchers thus expected that advertising appeal would mediate the relationship between advertising congruency and behavioral intentions.
Sexual identity may moderate the relationship between advertising congruency and advertising appeal as well as the relationship between advertising congruency and behavioral intentions. Distinctiveness theory would posit that the relationship between advertising congruency and the mediator (advertising appeal) and dependent variables (purchase intention and likelihood to recommend) may be stronger for LGB participants than for heterosexual participants given the salience of LGB imagery to the identity of LGB consumers. Distinctiveness theory and the extant literature suggest that the indirect effect of advertising congruency on behavioral intentions through advertising appeal may be moderated by sexual identity. The relationships among advertising congruency, advertising appeal, and behavioral intentions fit a moderated mediation model (See Figure 1).
H3: Advertising appeal will mediate the relationship between advertising congruency and behavioral intentions.
H4: Sexual identity will moderate the mediated relationship among advertising congruency, advertising appeal, and behavioral intentions.
Behavioral Intentions after a Temporal Delay
No previous research has examined effects of exposure to LGB-inclusive advertisements beyond the time frame immediately following exposure. The marketing literature generally concludes that advertising effects fade over time as consumers are exposed to competing advertisements and alternative brands (Burke and Srull, 1988). Advertising appeal, however, may explain potentially lasting effects of exposure. Advertisements perceived as likable are cognitively durable; their influence fades slower than advertisements perceived as unlikable (Brown, 1991).
The motivation of the consumer to engage with an advertisement may explain the longevity of effects. Individuals who are motivated highly to process an advertisement expend more cognitive effort on the advertisement and, in turn, more likely will retain and recall the advertisement's message with greater ease (Petty, Brinol, and Priester, 2009). When individuals are motivated to process an advertisement, the models in the advertisement may be elaborated on as substantive components of the persuasive message, which thereby increases the persistence of the effect (Petty and Cacioppo, 1986). Participants who report more positive attitudes toward advertisements spend more time viewing advertisements than those who report more negative attitudes (Olney et al., 1991).
In the current study, individuals who find the advertisements appealing should be more motivated to invest cognitively in the persuasiveness of the advertisements, particularly the models in the advertisements. Advertising appeal reported immediately following exposure should predict behavioral intentions following a temporal delay.
H5: Advertising appeal will mediate the relationship between advertising congruency and behavioral intentions measured three weeks postexposure.
METHOD
Participants
The study was conducted with an online data panel facilitated by Precision Sample, a market-research company that recruits vetted survey participants. The researchers used Precision Sample purposefully to oversample individuals from its database who identify as LGB to allow for statistical comparisons to be made between heterosexual and LGB participants (See Table 1). Individuals were sent recruitment messages to encourage participation and were offered $4 compensation for their time.
Participants who voluntarily clicked on the recruitment link were sent to the home page of the study, which contained the consent form. Given the nature of participant recruitment and the monetary incentive, the researchers took extra caution to ensure that only participants who had followed instructions were included in the final dataset reported here. A total of 67 participants were eliminated from the study for failing to answer accurately an attention check placed in the measures.
Procedure
After providing consent, participants were instructed to review a new digital magazine using the following prompt:
We are currently in the beginning stages of developing a new magazine and we are seeking some initial input! Please take a few minutes to carefully browse the magazine below, reading through the articles and checking out the advertisements. After reading through the magazine, you will be asked specific questions regarding its contents.
The online survey software randomly assigned participants to view a digital magazine containing advertisements featuring either same-sex couples (49 percent, n = 108) or other-sex couples (51 percent, n = 112; see the Appendix). Participants spent as much time as needed viewing the digital magazine.
After review, participants were shown each advertisement that appeared in the digital magazine individually and were asked to respond to scales measuring the advertisements' appeal, fit, and trustworthiness and their intent to purchase and likelihood to recommend the brand being advertised. Participants also responded to the appeal of the cover and each story in the digital magazine to maintain the deception that they were judging the digital magazine as a comprehensive product.
Three weeks after exposure to the digital magazine, participants were contacted to complete a brief follow-up questionnaire. Participants were not prompted to remember the digital magazine, nor were they shown the advertisements from the digital magazine. They only reported their purchase intention and likelihood of recommending the brands that were featured in the digital magazine's advertisements.
Participants who completed the follow-up questionnaire (n = 139; 63 percent response rate) were provided $2.50 compensation. No significant differences existed between those who participated in the follow-up survey and those who did not in terms of demographics, mean scores on attitudes toward any of the advertisements or stories in the digital magazine, or intent to purchase or recommend the brands advertised. The lack of difference suggests that the study design had no effect on likelihood of participating in the follow-up.
Experimental Stimulus
The decision to create a digital magazine as the experimental stimulus was made in light of recent research suggesting that magazine producers and consumers increasingly are interested in and accustomed to reading digital content (Guenther, 2011). The magazine was created with the digital-publishing platform Issuu, which presents content in a high-definition magazine-like reader that allows users to flip through pages virtually like they would with a paper magazine. The study utilized a magazine format for two reasons:
to increase the ecological validity of advertisement viewing, and
to shield the participants from an explicit awareness of the intended hypotheses.
The cover of the magazine displayed a photograph of sunflowers in the background, teasers for the articles contained within in the foreground, and the title of the magazine (THIS) in large font. The authors chose the generic sunflower photo and the title THIS to minimize the possibility of participants gleaning any information from the cover. The magazine contained three stories and five advertisements.
The decision was made to use stories and advertisements that had been published previously in magazines to increase the ecological validity of the study; none of the content in THIS magazine was created by the researchers. The stories and two of the advertisements served as fillers; as such, they were the same regardless of condition. Three of the advertisements varied by condition (i.e., other-sex or same-sex). The advertisements used for the manipulation were identical, with the exception of the sexual orientation of the couples depicted. The advertisements in the other-sex version portrayed heterosexual couples; the advertisements in the same-sex version depicted gay male couples (see the Appendix).
Pretest
The researchers conducted a pretest to determine the stories and advertisements that would compose both versions of the magazine. A research assistant collected neutral stories covering non-controversial topics. These stories stood to serve the purpose of filler content. As such, ideal stories would not distract or heighten the emotions of the reader. A total of seven stories were chosen for the pretest.
The advertisements were collected purposefully from campaigns that provided contextually identical same-sex and other-sex models to encourage the focus of the study's findings to be on model effects versus brand effects. The only meaningful difference was the sexual orientation of the individuals depicted. A total of 12 advertisements featuring six unique brands were chosen for the pretest.
Pretest participants collected through Amazon's mTurk (N = 84) reported their enjoyment (“This article is interesting, enjoyable, or entertaining”) and perceived offensiveness (“This article is offensive or controversial”) of seven magazine articles on a scale ranging from 1 (“strongly disagree) to 5 (“strongly agree”). The stories that were most enjoyable and least offensive were included in the digital magazine. These included articles about the birth of a baby penguin, the worst foods to eat before bed, and the friendship between a police officer and a child running a lemonade stand.
Pretest participants also were asked to describe the relationship between the two people featured in the advertisements (i.e., “In as few words as possible, describe the relationship between the two individuals pictured in this advertisement”). A research assistant coded the open-ended responses describing the relationship depicted in the LGB-inclusive advertisements as either platonic or intimate. Intercoder reliability, which the researchers calculated by having an author double-code 20 percent of the responses, was acceptable (Krippendorf's α = .86).
The advertisements that participants most likely perceived as portraying a gay couple were included in the same-sex version of the magazine; their other-sex counterparts were included in the other-sex version of the magazine. These included advertisements for Target (97 percent rated as intimate), Gap (95 percent), and Tiffany (91 percent). This pretest increased the likelihood that participants in the main study perceived the relationship between the two men in the advertisements as intimate rather than platonic. Finally, the researchers included in the digital magazine two filler advertisements for Tylenol and Sherwin-Williams depicting no people.
Measures
Immediately following exposure to the digital magazine, participants were asked to report the advertisements' appeal, their intentions to purchase the advertised brands, and their likelihood of recommending the advertised brands. Age, gender, race, and sexual orientation were collected with single-item measures in the questionnaire. Other measures of story appeal and overall magazine appeal are not detailed here, because they only appeared on the questionnaire to maintain deception and were not used in analyses. Three weeks postexposure, participants responded to the same measures of purchase intention and likelihood to recommend for the experimental manipulation brands advertised (i.e., Target, Gap, and Tiffany).
Advertising Congruency
Advertising congruency was a dichotomized variable. LGB participants viewing same-sex advertisements and heterosexual participants viewing other-sex advertisements were categorized as congruent. LGB participants viewing other-sex advertisements and heterosexual participants viewing same-sex advertisements were categorized as incongruent.
Advertising Appeal
Advertising appeal is a complex, multidimensional construct (Olney et al., 1991) and was measured in the present study with three scales:
a 5-item measure of attitude toward the advertisement (Lee, 2000),
a 3-item measure of advertisement trustworthiness (Lawrence, Fournier, and Brunel, 2013), and
a 3-item measure of advertisement fit (Putrevu, 2008).
Example items included,
“The ad is interesting to me” (attitude),
“The ad feels authentic” (trustworthiness), and
“The ad is a good match with the company depicted” (fit).
Participants responded to advertising appeal items on a 7-point scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). The 11 items measuring these components of advertising appeal were highly consistent (α = .98). The mean scores for attitude, trustworthiness, and fit also were correlated highly among the three stimulus advertisements (r = .55 to .86). A factor analysis suggested that the attitude, trustworthiness, and fit scores for each of the stimulus advertisements factored onto one item (Eigenvalue = 6.60), accounting for 73.74 percent of the variance. The authors thus combined attitude, trustworthiness, and fit to create a composite score representing appeal, ranging from 1 to 7 (M = 5.18, SD = 1.37).
Purchase Intention
The authors measured purchase intention by asking participants how likely they would be to shop at Tiffany, Target, and Gap on 7-point scales ranging from 1 (“very unlikely”) to 7 (“very likely”). The three measures of purchase intention immediately following exposure were consistent (α = .80) and correlated (r = .46 to .68). Results from a factor analysis reinforced that the purchase intention for all three brands should be collapsed into one measure, given that they factored onto one item (Eigenvalue = 2.15), accounting for 71.74 percent of the variance. Mean scores then were calculated for purchase intention; higher scores equated to stronger purchase intention.
Likelihood to Recommend
Participants reported how likely they were to recommend Tiffany, Target, and Gap to others on a 10-point scale ranging from 1 (“not at all likely”) to 10 (“extremely likely”). Much like with purchase intention, the researchers collapsed the items measuring likelihood to recommend the three brands immediately following exposure into one measure because of consistency (α = .84), correlation (r = .60 to .69), and a one-factor outcome (Eigenvalue = 2.28) accounting for 75.86 percent of the variance. The researchers then calculated mean scores to represent likelihood to recommend advertised brands; higher scores equated to stronger likelihood to recommend.
The researchers measured both purchase intention and likelihood to recommend to gauge a deeper brand connection than attitudes toward the advertisement. With respect to statistics, analyzing each advertised brand separately would not have added insight into consumers' intention, given the high correlations and single factor loadings. As such, purchase intention and likelihood to recommend were composite measures of all experimental advertisements in the digital magazine.
RESULTS
Preliminary analyses revealed a strong correlation between purchase intention and likelihood to recommend (r = .90). To avoid problems with multicollinearity, the authors conducted separate analyses to examine all predicted differences in the two dependent variables (Leech, Barrett, and Morgan, 2005). They tested the hypotheses with the PROCESS macro (Model 8; Hayes, 2013) to investigate the moderated mediation models, with bootstrapping set to 10,000 (See Table 2).
H1 predicted a direct effect of advertising congruency on behavioral intentions immediately following exposure. Advertising congruency predicted both purchase intentions (B = 0.96, p < .05) and likelihood to recommend (B = 1.46, p < .05) immediately following exposure. H1 was supported.
H2 predicted moderation, such that the direct effect of advertising congruency on behavioral intentions would be stronger for LGB participants. Although the interaction between condition and sexual identity was a significant predictor of behavioral intentions, PROCESS analyses suggested no conditional direct effects of advertising congruency on either purchase intention or likelihood to recommend. H2 was not supported.
H3 and H4, when taken together, proposed moderated mediation models in which sexual identity moderated the indirect effect of advertising congruency on behavioral intentions through advertising appeal immediately following exposure (See Figure 1). Sexual identity significantly contributed to advertising appeal, such that LGB participants found the advertisements more appealing than heterosexual participants. The significant interaction between advertising congruency and sexual identity on advertising appeal suggested that the relationship between advertising congruency and advertising appeal varied by sexual identity.
To facilitate interpretation, the researchers ran an analysis of variance with Advertising Congruence × Sexual Orientation as the independent variable (i.e., four nominal categories: LGB congruent, LGB incongruent, heterosexual congruent, heterosexual incongruent) and advertising appeal as the dependent variable. The researchers employed Bonferroni post hoc tests to examine mean differences. The model was significant, F(3, 215) = 22.68, p < .001. LGB participants in the congruent condition found the advertisements significantly more appealing than any other group; heterosexual participants in the congruent condition found the advertisements significantly more appealing than either incongruent-condition group. Participants in the incongruent conditions found the advertisements equally appealing (See Figure 2).
In tests for moderated mediation, the key indicator is the indirect effect of the interaction term on the dependent variables through the mediator (Preacher, Rucker, and Hayes, 2007). The indirect effect of advertising congruency on purchase intentions through advertising appeal was stronger for LGB participants (B = −1.71, SE = 0.20, 95 percent CI [−2.14, −1.29]) than for heterosexual participants (B = −0.89, SE = 0.28, 95 percent CI [−1.47, −0.37]), indicating moderated mediation.
The indirect effect of congruency on likelihood to recommend through advertising appeal was also stronger for LGB participants (B = −2.85, SE = 0.36, 95% CI [−3.56, −2.14]) than for heterosexual participants (B = −1.49, SE = 0.47, 95% CI [−2.45, −0.62]). The index of moderated mediation was significant for the model testing purchase intention as the dependent variable (B = −0.81, SE = 0.35, 95 percent CI [−1.49, −0.13]) and for the model testing likelihood to recommend (B = −1.35, SE = 0.58, 95 percent CI [−2.49, −0.21]). The moderated mediation model supports H3 and H4.
The final hypothesis predicted that advertising appeal would mediate the relationship between advertising congruency and behavioral intentions measured three weeks postexposure. The results suggest that only advertising appeal measured immediately following exposure significantly contributed to purchase intention and likelihood to recommend measured three weeks postexposure (See Table 2). Neither the index of moderated mediation for purchase intention (B = −0.38, SE = 0.22, 95% CI [−0.87, 0.01]) nor likelihood to recommend (B = −0.72, SE = 0.42, 95 percent CI [−1.62, 0.04]) was significant. H5 was supported.
DISCUSSION
Perceptions of LGB consumers' strong buying power and brand loyalty have increased the importance of courting the LGB market segment. Although many companies strive to be designated corporate allies in the eyes of LGB consumers, they often do so while attempting to appease the larger heterosexual market segment. This study contributes to the literature on LGB-inclusive advertising by examining purchase intention and likelihood to recommend immediately following exposure to advertisements in an ecologically valid setting and after a temporal delay of three weeks. The current study makes several unique contributions that advance scholarship within the domain of LGB advertising, highlighting a potentially decreasing disparity between LGB and heterosexual consumers' perceptions of LGB inclusivity, as well as the salience of LGB imagery in advertising for LGB consumers.
Consistent with prior studies and in line with distinctiveness theory, advertising congruency was related to behavioral intentions, such that individuals who saw advertisements congruent with their own sexual identity reported higher purchase intentions and likelihood to recommend. Advertising appeal, however, mediated the relationship between advertising congruency and behavioral intentions. The fact that advertising appeal acted as an important mediating mechanism explaining the relationship between advertising congruency and behavioral intentions is also in line with previous research suggesting that attraction to an advertisement is correlated with consumers' behaviors (Mitchell and Olson, 1981).
The indirect effect of advertising congruency on behavioral intentions through advertising appeal also was moderated by sexual identity, such that the conditional effect was stronger for LGB participants than heterosexual participants immediately following exposure to the advertisements. LGB participants reported stronger advertising appeal following exposure to congruent advertisements, and advertising appeal was associated strongly with purchase intention and likelihood to recommend. These findings illustrate the potential power and influence that being a corporate ally has for reaching LGB consumers. LGB-inclusive advertising seems to have a strong influence on LGB consumers; a brand actively increasing LGB visibility sends a potent message of inclusion that can affect LGB consumers' intended behaviors.
Advertising appeal mediated the relationship between advertising congruency and behavioral intentions, and the interaction between advertising congruency and sexual orientation predicted advertising appeal. The appeal of incongruent advertisements was practically identical for heterosexual and LGB participants. All participants, regardless of sexual orientation, reported mean appeal scores above the midpoint, which suggests that the sexuality of the models depicted did not create strong negative feelings toward the advertisements, regardless of the participants' sexual orientation. This is in contrast to more dated research that found heterosexuals' perceptions of LGB imagery in advertisements to be below the midpoint on seven-point scales (Oakenfull and Greenlee, 2004). The present findings are in line with more recent research suggesting that heterosexual consumers are not as adverse to LGB imagery as they once were (Um, 2014). These findings could be used to argue that advertising appeal is more important in predicting consumer behavior than the sexual orientation of the advertisement models.
The current study found that heterosexual participants' perceptions of the advertisements as interesting, trustworthy, and a good fit predicted behavioral intentions more than did advertising congruency. These findings therefore largely contradict previous research suggesting that heterosexual consumers respond negatively to LGB-inclusive advertisements (Bhat et al., 1998; Hester and Gibson, 2007; Hooten et al., 2009). Heterosexual consumers no longer may experience the same aversion to LGB-inclusive advertising campaigns as they once did. This finding could be a result of social progress.
In an experimental study examining attitudes toward LGB-inclusive advertisements, the author predicted that “the negativity effect likely to be derived from advertisements with gay-themed imagery is expected to diminish as public opinion of homosexuality shows a trend of increasing approval” (Um, 2014, p. 829). Indeed, attitudes toward LGB individuals have improved since previous studies were published on this topic (Gallup, 2018). One could argue that the “othering” of LGB individuals has waned, which has allowed heterosexual consumers to perceive LGB sexualities as less distinct. It is possible that heterosexual consumers find LGB imagery in mainstream advertisements more palatable as a result of increased societal acceptance of LGB sexualities and lifestyles.
Advertising appeal continued to contribute significantly to purchase intention and likelihood to recommend three weeks after exposure, without prompts for participants to think about the advertisements they had viewed previously. Although the effects of advertisements may fade over time because of competing resources (Burke and Srull, 1988), consumers' perceptions of advertisements in the current study were related to behavioral intentions after a temporal delay. Such a finding could be used to argue that original reactions to the advertisements were prompted not solely by pre-established brand experience or preferences but by distinct and independent reactions to the advertisements themselves. Feelings toward the advertisements then contributed to behavioral intentions even when participants were not engaged with the experimental stimuli. No other variables in the model significantly contributed to the behavioral intentions three weeks post-exposure, however, which suggests that advertising appeal might have been influenced by other external factors beyond advertising congruency or participant sexual identity. Future research should study the long-term effects of exposure to LGB-inclusive advertising.
Managerial Implications
The findings can be of assistance to the advertising community and broader marketing industry. First, because LGB consumers experienced the most notable influence from appealing, sexually congruent advertising, it is important for marketing managers to identify and target their LGB contingent through channels and placements that are the most frequented and utilized. Clearly delineated targeting and delivery of congruent advertisements to the LGB consumer group will serve to elevate the effectiveness of the advertisements as related to purchase intentions and likelihood to recommend. Marketing managers should place particular effort and attention on the strategic development and placement of LGB-inclusive advertising, which may include an increased emphasis on product offerings most utilized by LGB consumers or placement within media outlets popular with LGB audiences.
Similarly, the findings also serve to help brands that are seeking to cater better to their current LGB customer base or to attract new consumers in the LGB market. As the LGB market continues to grow in size and purchasing power (Nelson and Mitchell, 2018), large brands should feel more confident in incorporating LGB-inclusive imagery into their comprehensive marketing campaigns. As illustrated by Target, Tiffany, and Gap in the current study, LGB consumers expressed favorable attitudes toward the advertisements as well as positive intentions to purchase and recommend the product. The utilization of same-sex models in appropriately targeted media may help to attract a broader LGB consumer base.
Finally, it is important to note that, broadly, heterosexual consumers no longer may experience the same aversion to LGB-inclusive advertisements found in previous studies. Although heterosexual participants found congruent advertisements significantly more appealing than incongruent advertisements, the appeal of incongruent advertisements depicting LGB individuals was above the midpoint. This is important for brands that likely have shied away from LGB-inclusive marketing strategies in fear of retribution or potential loss of sales.
Advertisers should focus on developing advertisements that consumers find interesting, authentic, and a good match with the respective brand—established through proper marketing research. Because the results indicate that increased advertising appeal leads to higher intentions to purchase and recommend the respective brand, focus should be placed on the design and appropriateness of the advertisement, regardless of the sexualities depicted through the models. Increased appeal demonstrated lasting effects three weeks postexposure. This highlights the necessity of developing appealing advertisements above all else. Because this effect was found for both LGB consumers and heterosexual consumers, the effects of appeal do not discriminate by sexual orientation.
Limitations and Future Research
Although this study advances the field's understanding of advertising effects, it is not without limitations. In an effort to preserve power, the researchers did not examine differences by participant gender, and they analyzed sexual orientation in a dichotomous fashion: heterosexual and LGB. Sexual orientation was measured with a single item that might not have grasped the nuanced nature of participants' sexual orientations. Future research should reconsider the operationalization of sexual orientation and collect a larger sample of gay men, lesbians, and bisexual individuals to examine potential differences within LGB sexual orientations.
The sample also was skewed in terms of education (more educationally diverse samples would be more externally generalizable), and the three-week follow-up sample was smaller than the original sample, which limited the power of the over-time analyses. Additionally, the same-sex advertisements used in the current study only depicted gay male couples; no images of lesbian women were included in the present study. Depictions of gay men in advertisements provoke stronger contempt than depictions of lesbians (Oakenfull and Greenlee, 2005). Future research may draw more nuanced conclusions if it uses both images of gay men and lesbian women in experimental stimuli.
The current study also used stimuli from the lens of stores as brands. That is, Target, Gap, and Tiffany are larger entities that often contain many subbrands within their overall offerings. Target carries several unique and independent brands. Future studies should look into the effects that LGB-inclusive advertising may have on distinctly specific brands, versus stores that encompass a collection of brands. Brands such as Ray-Ban, which are sold at a variety of retailers, may not illustrate the same effects as were found in this study. There is the potential for important implications for advertisers utilizing LGB-inclusive advertising for either stores or brands separately.
Similarly, additional research should look into differences in context presented within the advertisement. Two of the advertisements used in this study were wedding related, which could influence the distinctiveness of the models' sexual orientation for the viewer. More expansive research analyzing the effects of advertisements through a variety of contexts may serve to either increase or dampen the effects found herein. Possible examples may include advertisements focusing on family togetherness, adventure and travel, friendship, or fashion and lifestyle. Understanding how both LGB and heterosexual consumers respond to congruent and incongruent advertising across a multitude of contexts would provide advertisers with more in-depth guidance across a broader base of product offerings.
Finally, most studies examining LGB imagery in advertisements have exposed participants to advertisements without any media context (Hester and Gibson, 2007). The present study required participants to browse a fictitious magazine for purposes of increasing the ecological validity of the study. Although the neutral nature of the magazine was desirable for the study design, the advertisements might have been interpreted differently if they were included in a magazine focused on sports, home improvement, or fashion.
Future research should continue to investigate how media context can influence audiences' perceptions of advertisements, as it relates to the broader marketing context experienced by the consumer. Such findings would help inform advertisers and media buyers on appropriate advertisement placement within specific media channels or, more broadly, across advertising domains.
ABOUT THE AUTHORS
Bradley J. Bond is an associate professor of communication studies at the University of San Diego. His research examines the cognitive, affective, and behavioral effects of media exposure. His recent work focuses on the representation of sex and sexuality in the media, particularly depictions of LGBTQ individuals, and effects of exposure on audiences. He has published in the Journal of Communication, Media Psychology, Mass Communication & Society, and Psychology of Popular Media Culture, among others.
Justine Rapp Farrell is an associate professor of marketing at the University of San Diego's School of Business. Her research interests focus on the domain of consumer welfare, in particular underrepresented and at-risk population groups. Her research has implications for a variety of domains, including consumer behavior, public policy, and transformative consumer research, and she has published articles in journals including Journal of Consumer Research, Journal of Public Policy & Marketing, and Journal of Business Ethics.
Appendix Advertisements Used in Experimental Design
- Received June 1, 2017.
- Received (in revised form) November 26, 2018.
- Accepted December 19, 2018.
- Copyright© 2020 ARF. All rights reserved.