Creating Persuasive Environmental Communicators: Spokescharacters as Endorsers in Promoting Sustainable Behaviors

Abstract

Effective communication strategies are vital to educating consumers about sustainable behaviors given that environmental problems such as climate change are a result of excessive human activities. This study examined the suitability of using anthropomorphic characters (i.e., spokescharacters) in environmental communication and investigated traits that make them persuasive communicators in promoting sustainable behaviors. Using the Source Credibility Model and Matchup Hypothesis as theoretical lenses, this study explored spokescharacter traits that can enhance their persuasiveness, particularly their ability to influence consumers’ recycling attitudes and behavioral intentions. An online questionnaire was administered to 314 U.S consumers recruited through Amazon’s MTurk platform. Research hypotheses were tested using Structural Equation Modelling. The results show that spokescharacter traits such as likability, expertise, and relevance positively influence spokescharacter trustworthiness, which in turn enhances the recycling attitudes and recycling intentions of consumers. Overall, the results suggest that spokescharacters can be effective communicators to promote sustainable behaviors like recycling. Theoretical and practical implications are offered, and future research directions are outlined.

1. Introduction

Significant scientific research has established that environmental issues such as climate change are attributable to human activity and pose a grave threat to mankind [1]. Still, a large segment of the public denies anthropogenic climate change [2] owing to the polarization and politicization of environmental issues [3] and low levels of scientific knowledge among the public [4], which may inhibit the adoption of sustainable behaviors such as recycling. As a result, the public tends to rely on trusted sources to inform their views on environmental issues [5]. In this paper, we explore the suitability of spokescharacters, who are trusted by public [6], in enhancing the persuasiveness of environmental communication. While spokescharacters have been used extensively to persuade consumers in commercial marketing [7], less is known about whether they could be as effective in promoting sustainable behaviors such as recycling.

Assessing spokescharacters’ effectiveness in environmental communication is warranted as they can be persuasive communicators owing to their high perceived trustworthiness [8] and as such, may not face backlash from consumers as in the case of human celebrities promoting their own brand image by supporting environmental causes [9]. However, at the same time, being fictional in nature and not affected by environmental problems may render spokescharacters less effective. The presence of spokescharacters can make environmental messages humorous [10,11] which can distract the audience and potentially dilute the perceived seriousness of the environmental issues [12,13]. Still, Borden and Suggs [14] reported positive outcomes of using humor in environmental campaigns. Further, research is emerging that explores the effectiveness of using cute stimuli in environmental appeals [15] and as spokescharacters are considered cute by consumers [16,17], it is reasonable to explore if spokescharacters can be effective communicators in an environmental context. Specifically, this paper empirically explores the spokescharacter traits that can enhance their persuasiveness in environmental communication.

We adopted the Source Credibility Model [18] and the Matchup Hypothesis [19] as theoretical lenses in this study for three reasons. First, source trust is an important variable influencing environmental communication effectiveness [8] and spokescharacters command a high level of trust among consumers [6,20]. Second, as marketers can control the appearance and personality of a spokescharacter, it is possible to achieve a high matchup (or relevance) between the spokescharacter and the advertised product [21], such as environmental cause. In persuasion literature, research has consistently shown that source credibility positively influences the persuasiveness of the message. Source credibility is conceptualized to mainly comprise of two dimensions, viz., expertise (comprising experience, knowledge, or qualification of communicator) and trustworthiness (comprising honesty, integrity, and intention of communicator to tell the truth) [18]. Hence, a source with high expertise and trustworthiness can make environmental communication more persuasive, resulting in adoption of sustainable behaviors.

Finally, a number of meta-analyses have confirmed that message persuasion depends on source credibility as well as perceived endorser-product congruence or fit [22,23,24]. We chose recycling as the context within which to test spokescharacter persuasiveness in environmental communication for two reasons. First, despite ongoing efforts, recycling rates in the world remain low [25,26] and second, recycling is an important climate change mitigation behavior with wide ranging positive impacts on environmental sustainability [5].

This study contributes to extant literature in four ways. First, even though studies have established the positive outcomes of anthropomorphizing nature itself in promoting sustainable behaviors [25,27], our research focuses on the suitability of anthropomorphic characters as communicators in environmental appeals by exploring traits that support their persuasiveness. By doing so, we add to the limited literature that focuses on exploring anthropomorphic characters’ role in environmental communication [28,29]. Second, by empirically showing that spokescharacters can enhance message persuasion, this study adds to the environmental communication literature by answering research calls to explore (1) additional trusted environmental communicators [30] as well as (2) the outcomes of using trusted communicators [31]. Next, this research offers a theoretical understanding of how the relevance of anthropomorphic characters enhances message persuasion to promote sustainable behaviors, thus adding to the Matchup Hypothesis literature that predominantly reports outcomes of matchup using human spokespersons in the commercial marketing contexts. Finally, by testing likability as a separate source credibility dimension, our study adds to the limited literature that explores how source likability influences consumers’ attitudes and intentions [23]. In the next section, we review relevant literature to develop the conceptual framework tested in this study.

2. Literature Review

2.1. Spokescharacters

Phillips [32] (p. 146) defined a spokescharacter as a “fictional, animate being or animated object that has been created for the promotion of a product, service, or idea”. These fictional characters were introduced in the late nineteenth century for the purposes of product differentiation, to give personality to, and generate trust for manufacturing companies [33]. Spokescharacters are used in advertisements for products targeting both children and adults (males and females), and promote both low involvement (e.g., food products) and high involvement products (e.g., prescription drugs, electronics, insurance, banking, and travel solutions) [7,34,35,36]. Studies have consistently demonstrated the positive influence of using spokescharacters in communication on consumers’ brand and ad attitudes, brand trust, and behavioral intentions [21,36,37,38]. Hence, irrespective of product type, age, and gender of audiences, spokescharacters have been found effective in commercial marketing. However, the literature examining whether spokescharacters are effective in promoting sustainable behaviors is nascent (but see Hayden & Dills [39] for an example of anthropomorphic characters in environmental campaigns), as evidenced by the fact that two recent articles focusing on exploring the prevalence of spokescharacters in advertising and anthropomorphism in environmental contexts did not report the use of spokescharacters in environmental marketing [7,29].

Spokescharacters possess unambiguous qualities, remain predictable, and always embody the same meaning [39]. Therefore, they appear trustworthy and reliable to consumers which positively influences consumers’ attitudes and behaviors [20,21]. In fact, cartoon characters create an aura of trust [40], which may result in higher environmental communication persuasiveness [41]. Compared to a human spokesperson, consumers are more likely to believe a spokescharacter [42,43]. A large-scale survey with 1630 U.S households found that consumers generally like and trust spokescharacters [6]. Since marketers can alter the appearance and personality of spokescharacters, it is possible to create highly trusted spokescharacters by manipulating traits that influence trustworthiness [21,43].

Hence, it is reasonable to examine whether, owing to the high trust placed in spokescharacters by consumers, spokescharacters are effective environmental communicators [8]. This is also supported by recent research which demonstrates the effectiveness of using cute stimuli in promoting sustainable behaviors like recycling [15]. However, being fictional in nature and not affected by environmental problems and priming humor in environmental messages may render them ineffective. Therefore, research to test the persuasiveness of spokescharacters in environmental communication promoting sustainable behaviors like recycling is warranted.

2.2. Source Credibility Model

The Source Credibility Model emerged as the result of pioneering work by Carl Hovland and his colleagues during the mid-twentieth century [18], who found that communicator’s credibility positively influences message persuasiveness. According to this model, source credibility comprises of two dimensions: expertise and trustworthiness. Whereas expertise is defined as “the extent to which a communicator is perceived to be a source of valid assertions”, trustworthiness represents “the degree of confidence in the communicator’s intent to communicate the assertions he/she considers most valid” [18] (p.21). In other words, whereas source expertise refers to the ability to tell the truth, trustworthiness refers to the willingness to tell the truth [44]. In environmental communication, source trustworthiness is key to the persuasiveness of the message [5] and research supports the notion that using credible sources to communicate about sustainability practices is an effective strategy [45].

Attribution theory [46] can explain why highly credible sources are perceived as more persuasive communicators. When the source is perceived to have high levels of expertise, consumers perceive the recommendation to be genuine rather than encouraged by external rewards (such as monetary compensation). That is, a credible source may appear internally motivated to inform, rather than persuade, consumers. Marketers are able to portray spokescharacters as experts of the product they endorse as consumers view them repeatedly in several campaigns promoting and/or using the same product (e.g., Tony the Tiger for Kellogg’s breakfast cereal), thus developing a strong mental association between the spokescharacter and the product. Such repetitive exposure (in advertisements, or on product packages) gives validity to the perception that spokescharacters have knowledge and experience of the product [18]. Hence, consumers may attribute higher trustworthiness to the spokescharacters owing to their perceived expertise [21].

The positive link between a spokesperson’s expertise and trustworthiness is well established in literature [47,48,49,50]. Sundar [51] reported an ‘expertise heuristic’ phenomenon, wherein information coming from an expert source appears more credible. Further, expertise is identified as one of the main dimensions influencing source trust in risk communication [52]. In the spokescharacter context, Garretson and Niedrich [21] empirically showed that perceived expertise predicted spokescharacter trustworthiness. Hence, the following hypothesis is offered:

H1: 

Spokescharacter expertise will positively influence spokescharacter trustworthiness in a recycling message.

Extant literature consistently suggests that source credibility enhances message persuasion. Reviewing literature since the 1950s that examined the role of source credibility in enhancing message persuasion, Pornpitakpan [53] found that overall, a highly credible (vs. less credible) source is more persuasive in terms of attitude change and enhancing the behavioral intentions of consumers. A number of systematic literature reviews have consistently established the positive influence of source trustworthiness on consumers’ attitude towards product as well as their behavioral intentions [22,24,44,54]. Therefore, it is reasonable to hypothesize that:

H2: 

Spokescharacter trustworthiness will positively influence the (a) recycling attitudes and (b) recycling intentions of consumers in a recycling message.

As per the Theory of Planned Behavior, attitude towards behavior is an important predictor of behavioral intentions [55,56]. In their meta-analysis of factors influencing recycling, Geiger et al. [57] found that recycling attitudes was a significant predictor of recycling intentions. Hence, the following hypothesis is offered:

H3: 

Recycling attitudes will positively influence the recycling intentions of consumers in a recycling message.

2.3. Spokescharacter Likability

Likability is defined as an ‘affection for the source as a result of the source’s physical appearance and behavior’ [44]. In their comprehensive endorsement framework, Schimmelpfennig and Hunt [24] have suggested that source likability was linked to positive affect among consumers, which gets transferred to the advertised product, leading to higher persuasion. Source likability, unlike expertise or a source’s congruence or fit with the advertised product, is context independent [23]. Hence, a likable source such as a spokescharacter can be persuasive whether promoting commercial brands or sustainable behaviors like recycling.

While attractiveness is considered the third dimension (in addition to trustworthiness and expertise) of source credibility [22,53], researchers have broadened the concept of attractiveness to likability, which may or may not include physical attractiveness [24]. Whereas facial appearance or physical attractiveness is more suitable when considering a human celebrity or model [58], it is likability that is more apt for spokescharacters as consumers find spokescharacters likable [6,11,39].

In past studies, source likability has been found to positively influence perceptions of trust [58,59]. Luo, McGoldrick, Beatty and Keeling [43] also found a link between onscreen characters’ facial appearance, likability, and trustworthiness in an e-commerce website context. Callcott and Phillips [11], through qualitative interviews with consumers, found that the cute and humorous appearance of spokescharacters make them likable, supporting the findings of other studies which have reported that cuteness [60] and humor [61] can enhance source trustworthiness. The positive influence of spokesperson likability on consumers’ brand attitudes and purchase intentions has also been well established in the endorsement literature [11,23,24,39]. Hence, it is reasonable to believe that:

H4: 

Spokescharacter likability will positively influence (a) spokescharacter trustworthiness, as well as the (b) recycling attitudes, and (c) recycling intentions of consumers in a recycling message.

2.4. Matchup Hypothesis

Misra and Beatty [62] (p.161) defined the concept of ‘matchup’ or spokesperson -product congruence as consistency between highly relevant characteristics of the spokesperson and highly relevant attributes of the endorsed product. According to the matchup hypothesis, communication persuasiveness is enhanced under the condition of high endorser-product fit or congruence [63,64]. Empirical research using human spokespersons has consistently found that a higher perceived endorser-product fit enhances endorsement effectiveness [65,66]. However, the matchup effect has not been tested adequately for spokescharacters (for exception, see [21]), particularly within the environmental communication context. The impact of spokesperson -product congruence has been explained from the perspective of social adaptation theory and schema theory (for a discussion see [67,68]).

We adopted the conceptualization of spokescharacter relevance from the study of Garretson and Niedrich [21], where they considered relevance as the degree to which the spokescharacter and advertised product match each other, that is, the degree to which a spokescharacter is perceived to be appropriate to represent the promoted product. The concept of spokescharacter relevance is similar to what researchers have used in the non-profit context to represent the matchup condition, for example., the perceived connection or fit between the spokesperson and the non-profit organization [69,70], as well as in a profit context, for example, as product- endorser congruence [63]. As Garretson and Niedrich [21] argued, it is the appearance of the spokescharacter that can lead to greater perceived relevance of the spokescharacter. The Michelin Man, whose body is made up of tires for example, is considered appropriate to promote the Michelin brand of tires. The qualities associated with animals (e.g., a cheetah is fast, a rabbit has endless energy) can also help marketers to achieve higher spokescharacter-product matchup (e.g., the Energizer Bunny for Energizer batteries to reflect the long-lasting nature of their batteries). For practitioners as well as consumers, the matchup between the spokescharacter and advertised product is an important consideration, with the former aiming to create highly persuasive communicators and the latter using it as a signal to assist in developing dispositions toward brands [21].

Research shows that under the condition of high matchup between a spokesperson and the product, the spokesperson’s credibility is high [19,71]. Further, in a non-commercial context, studies also show that higher perceived fit between the spokesperson and the advertised cause/non-profit organization leads to higher source credibility perceptions [69,70]. Similarly, the matchup condition has been shown to consistently and favorably impact consumers’ brand attitudes and behavioral intentions [19,63,65,67,72]. Therefore, the following hypotheses are offered:

H5: 

Spokescharacter relevance will positively influence (a) spokescharacter trustworthiness, as well as the (b) recycling attitudes and (c) recycling intentions of consumers in a recycling message.

The literature review resulted in a theory-based conceptual framework (See Figure 1) that was tested using Structural Equation Modelling.

3. Methodology

3.1. Stimuli Development

A list of twenty spokescharacters was developed by research team from the pool available from a brand spokescharacters survey [6]. These twenty spokescharacters belonged to three broad categories, i.e., animal, mythical, and product personified as described by Callcott and Lee [33]. Next, six versions of recycling messages were developed as follows. By adopting the procedure of Kronrod et al. [73], five recycling slogans were shortlisted from the website https://www.thinkslogans.com/ (accessed on 15 July 2020). Based on the feedback from four subject experts, one slogan was selected (i.e., ‘Recycle today for a better tomorrow’) as it was simple and easy to understand. The same experts were asked to select six different spokescharacters (two of each type) out of the list of twenty which they deemed suitable to promote recycling. Having multiple spokescharacters ensured sufficient variability in spokescharacter traits, which was needed to evaluate the different hypotheses in the model [69]. Based on their responses, the following spokescharacters were finally selected by the research team: Tony the Tiger and Coco the Monkey (animal type), Mr. Clean and Gnome (mythical type), Michelin Man and M&M Red (product personified type). A professional designer was hired to develop the artwork of the message. The six versions of recycling message used as stimuli in this study can be requested from the corresponding author.

3.2. Method, Measures, and Participants

To test the research hypotheses, data was collected by administering an online questionnaire designed in Qualtrics to U.S. residents who were at least 18 years old, using Amazon’s MTurk platform (with 95% and above approval ratings) and were compensated for their participation. Following best practices, two attention check questions were added to the online survey [74]. Each participant saw one of the six versions of recycling message for about 20 s and then answered questions related to spokescharacter’s attributes, recycling attitudes and intentions, and demographics. For the purpose of the analysis, data from the six groups of participants were later merged. After removing the responses for attention check failure and incomplete information, 314 usable responses remained and were used for data analysis.

Source credibility was measured by the traits of expertise, trustworthiness, and likability [24], whereas spokescharacter relevance measured the degree of matchup or fit between the spokescharacter and the recycling cause [21]. Measures were adopted from the scales used in existing studies, such as expertise and trustworthiness [58], relevance [21], likability comprising the traits of liking, cuteness, humor, appeal, and friendliness [15,39,43,75], spokescharacter familiarity [19], recycling attitudes [76], recycling intentions [15] and past recycling behavior [77,78]. Participants rated their responses on a seven-point Likert or Semantic Differential scale (see Appendix A).

As participants’ demographic variables and past recycling behavior can influence recycling intentions [56,57], they were employed as control variables in the model. Additionally, as we used well-known spokescharacters as stimuli in this study, the effect of spokescharacter familiarity (not the focus variable in this study) on recycling intentions was controlled for, given that spokesperson familiarity can influence behavioral intentions of consumers [24].

Participants demographic data are provided in

Table 1. Demographics of Participants.

Participants’ Characteristics (n = 314)		Frequency	Percent
Gender	Male	187	59.6%
	Female	127	40.4%
Age in years (mean = 40.42; median = 38; SD = 11.30)	24–30	69	22%
	31–45	151	48%
	46–60	78	25%
	Above 60	16	5%
Educational background	Less than high school degree	1	0.3%
	High school graduate	36	11.5%
	Some college but no degree	39	12.4%
	Associate degree in college (2-year)	42	13.4%
	Bachelor’s degree in college (4-year)	143	45.5%
	Master’s degree	46	14.6%
	Doctoral degree	3	1.0%
	Professional degree (JD, MD)	4	1.3%
Income in USD	Less than $20,000	28	8.9%
	$20,000–44,999	94	29.9%
	$45,000–139,999	160	51.0%
	$140,000–149,999	9	2.9%
	$150,000–199,999	11	3.5%
	$200,000+	7	2.2%
	Prefer not to answer	5	1.6%

. The mean, standard deviation, and Cronbach’s alpha (all above required 0.7; [79]), of different constructs are reported in

Table 2. Cronbach’s alpha values, mean and standard deviations of constructs (n = 314).

Measure	Mean	Standard Deviation	Cronbach’s α
Spokescharacter familiarity *	5.65	1.88	N.A.
Spokescharacter likability	5.18	1.07	0.90
Spokescharacter trustworthiness	5.77	1.06	0.95
Spokescharacter expertise	5.10	1.45	0.96
Spokescharacter relevance **	4.67	1.68	0.97
Recycling attitudes	6.15	0.88	0.93
Recycling intention	5.23	1.57	0.89
Past recycling behavior	5.00	1.60	0.90

* Measured by a single item, ** mean value above mid-point of scale, 4 (t = 7.143, p < 0.001).

.

4. Results

4.1. Measurement Model

Confirmatory factor analysis (CFA) was conducted to assess if the study’s data fits well with the underlying factor structure of constructs taken from previous studies [79]. The measurement model fit was assessed using standardized factor loadings, chi-square tests, and multiple fit indices. One item (i.e., friendly) loaded less on spokescharacter likability (0.475) and was deleted. Model fit indices (CMIN/df = 2.34, p = 0.00, TLI = 0.95, CFI = 0.95, RMSEA = 0.067, SRMR = 0.067) suggest that the data fits the model quite well. All factor loadings were significant (p < 0.001) and above 0.70, indicating convergent validity. Additionally, construct reliability (CR) for all constructs was above the recommended value of 0.7, further confirming convergent validity (see

Table 3. Convergent and discriminant validity analysis.

	CR	AVE	(1)	(2)	(3)	(4)	(5)	(6)
Likability (1)	0.888	0.666	0.816
Expertise (2)	0.960	0.827	0.500 ***	0.910
Relevance (3)	0.974	0.903	0.486 ***	0.717 ***	0.950
Trust (4)	0.947	0.783	0.568 ***	0.717 ***	0.635 ***	0.885
RATT (5)	0.943	0.704	0.164 **	0.283 ***	0.264 ***	0.436 ***	0.839
RI (6)	0.921	0.796	0.408 ***	0.378 ***	0.506 ***	0.492 ***	0.517 ***	0.892

** p < 0.01, *** p < 0.001.

). The correlation between constructs was lower than 0.85 and the square root of average variance extracted (AVE) for all constructs (reported in diagonalin Table 3) was greater than the pairwise correlations between constructs, confirming discriminant validity [79,80].

4.2. Structural Model

The conceptual model developed for this study was tested using SPSS AMOS 27 using maximum likelihood estimation. The structural model fit indices (CMIN/df = 2.33, p = 0.00, TLI = 0.92, CFI = 0.93, RMSEA = 0.06, SRMR = 0.098) showed the data fitted the model well. As per Hair, Black, Babin and Anderson [79], the “rule of thumb is that an SRMR over 0.1 suggests a problem with fit”. Hair, Black, Babin and Anderson [79] also state that one of the two fit indices RMSEA or SRMR should be sufficient to indicate a good fit. In our case, the RMSEA of the model is less than the recommended 0.07, indicating goodness of fit.

The amount of variance explained in recycling intentions was 48%. As reported in

Table 4. Structural parameter estimates (standardized coefficients).

Hypothesis	Path Direction	β	S.E.	C.R.	Result
H1	EXP → TRUST	0.46 ***	0.047	7.261	Supported
H2(a)	TRUST → RATT	0.50 ***	0.068	5.966	Supported
H2(b)	TRUST → RI	0.11	0.100	1.635	Not supported
H3	RATT→ RI	0.25 ***	0.092	4.914	Supported
H4(a)	LIK →TRUST	0.25 ***	0.047	4.915	Supported
H4(b)	LIK → RATT	−0.12	0.053	−1.741	Not supported
H4(c)	LIK → RI	0.15 ***	0.077	2.594	Supported
H5(a)	REL → TRUST	0.19 **	0.036	3.078	Supported
H5(b)	REL → RATT	0.004	0.035	0.050	Not supported
H5(c)	REL → RI	0.20 ***	0.051	3.385	Supported

EXP = Expertise; TRUST = Trustworthiness; RATT = Recycling attitudes; RI = Recycling intentions; REL = Relevance; LIK = Likability; S.E. =Standard Error; C.R. = Critical Ratio ** p < 0.01; *** p < 0.001.

, seven out of ten hypothesized relationships were supported. The effect of control variables on recycling intentions is reported in

Table 5. Effect of control variables on recycling intentions.

Control Variables	β
Spokescharacter familiarity	−0.10 *
Past recycling behavior	0.42 ***
Participants’ demographics
Income	0.04
Age	0.01
Education	0.04
Gender	−0.02

*** p < 0.001, * p < 0.05.

.

• Mediation analysis

A 10,000-sample bootstrapping procedure was performed to create 95% bias-corrected confidence intervals. Mediation analysis showed that spokescharacter trustworthiness fully mediated the effects of relevance, expertise, and likability on recycling attitudes. Further, recycling attitudes fully mediated the effects of trustworthiness on recycling intentions.

Table 6. Mediation analysis.

Indirect Path	Unstandardized Estimate	Lower Value	Upper Value	p-Value	Standardized Estimate
EXP → TRUST → RATT	0.140	0.087	0.204	0.001	0.231 ***
EXP → TRUST → RATT → RI	0.063	0.036	0.107	0.000	0.231 ***
EXP → TRUST → RI	0.056	−0.008	0.122	0.141	0.053
LIK → TRUST → RATT	0.093	0.054	0.157	0.000	0.126 ***
LIK → TRUST → RATT → RI	0.042	0.022	0.078	0.000	0.126 ***
LIK → RATT → RI	−0.042	−0.094	0.000	0.097	−0.032
REL → TRUST → RATT	0.046	0.018	0.088	0.004	0.093 **
REL → TRUST → RATT → RI	0.021	0.008	0.043	0.003	0.093 **
REL → RATT → RI	0.001	−-0.033	0.049	0.933	0.001
TRUST → RATT → RI	0.184	0.112	0.287	0.000	0.131 ***

EXP = Expertise; TRUST = Trustworthiness; RATT = Recycling attitudes; RI = Recycling intentions; REL = Relevance; LIK = Likability; ** p < 0.01; *** p < 0.001.

reports the mediation analysis with upper and lower level of unstandardized estimate and associated p-Value, and standardized estimate for various indirect paths.

5. Discussion

The spokescharacter traits of likability, expertise, and relevance predicted spokescharacter trustworthiness, which is consistent with extant spokesperson literature [49,58,69]. Moreover, we showed that spokescharacter relevance positively influences spokescharacter trustworthiness, recycling attitudes, and recycling intentions. This finding differs from what is reported in the spokescharacter literature [21]. In a commercial marketing context, Garretson and Niedrich [21] suggest that as some of the brand spokescharacters have become iconic owing to decades of existence, their relevance did not matter to consumers (e.g., Aflac Duck for Aflac Insurance). However, it is possible that in the environmental communication context, owing to the prevalence of ‘greenwashing’ and ‘green skepticism’ in society [81], the lack (and decline) of trust in formal environmental communicators such as politicians, scientists, and governments [8,82], and controversies surrounding the authenticity of environmental information [83], the relevance of spokescharacters was more salient for participants in this study. Thus, relevance was a significant predictor of spokescharacter trustworthiness and consumers’ recycling attitudes and intentions in messages focused on promoting recycling.

The findings of our study also support the general consensus in the literature that anthropomorphism have a positive effect in promoting sustainable behaviors. Whereas anthropomorphizing nature (e.g., showing earth with human traits) enhances the pro-environmental behaviors of consumers [29], our study shows that anthropomorphic communicators in environmental campaigns can bring aboutsimilar outcomes.

Our study also established that source relevance can have direct, positive influence on behavioral intentions. This contrasts with the finding reported by Choi and Rifon [63] that matchup effects on behavioral intentions are fully mediated through product attitude. A plausible explanation of this direct effect can be that consumers rationally evaluate source relevance (i.e., judicious assessment of the appropriateness of the source to endorse the environmental cause) and once convinced, they may be more likely to adopt sustainable behavior without undergoing the process of forming positive attitudes towards the behavior.

It was also found that spokescharacter trustworthiness fully mediated the effect of spokescharacter traits such as expertise, likability, and relevance on recycling attitudes, in line with existing spokescharacter research [21]. Additionally, spokescharacter likability exerted direct, positive influence on recycling intentions. The finding that spokescharacter likability exerted both direct and indirect positive influence (through trustworthiness and recycling attitudes) on recycling intentions, is consistent with the extant literature [22,24]. However, spokescharacter trustworthiness influenced recycling intentions only through recycling attitudes. A possible explanation of this finding can be that in the recycling literature, attitude towards recycling is shown to be an important predictor of recycling intentions [57]. As per The Theory of Planned Behavior, behavioral beliefs play an important role in the formation of attitude towards behavior, which then leads to behavioral intentions [55]. Additionally, source credibility research suggests that a credible source is perceived by the audience to possess requisite knowledge and experience and thus appear believable [18]. Thus, in our study, the exposure to trustworthy spokescharacters in the recycling message positively influenced recycling beliefs (as spokescharacters appeared believable in stating that recycling is beneficial), resulting in the formation of favorable recycling attitudes, and subsequently influencing recycling intentions.

Altogether, our study supports the findings of extant literature that using a credible and relevant (to the advertised product) source as an endorser can enhance communication persuasiveness [24], particularly in the environmental context [5]. By doing so, we demonstrate that even fictional communicators (e.g., spokescharacters) can be persuasive in environmental context, provided they are designed to be highly congruent with the advertised sustainable behavior and to engender elevated credibility perceptions among consumers.

In terms of control variables, past recycling behavior positively influenced recycling intentions, which is in agreement with extant recycling literature [57]. However, none of the demographic variables (age, education, gender, and income) had a significant influence on recycling intention, a finding which differs from what past research has established [84]. A probable explanation can be that spokescharacters appeal broadly to demographically diverse consumers [7,35] and so participants’ demographics did not play a role in predicting their recycling intentions. Additionally, spokescharacter familiarity (β = −0.10; p < 0.05) exerted negative influence on recycling intentions, possibly since higher familiarity with the spokescharacters used in this study made participants associate them more strongly with the products and brands they advertise (e.g., the Michelin Man for Michelin tires) and as a result, highly familiar spokescharacters were perceived to be less suitable to promote recycling by participants.

To conclude, our findings support the proposition of this study that spokescharacters can be considered effective communicators to promote sustainable behaviors like recycling among the public. This is notable given that the effects of past recycling behavior, spokescharacter familiarity, and participants’ demographics on recycling intentions were controlled for and participants were only briefly (one time, for 20 s) exposed to the recycling message.

The next section details the theoretical contributions this study made to the literature, followed by practical contributions, study limitations, and future research directions. The respecified model is presented in Figure 2.

5.1. Theoretical Contributions

This study applied the Source Credibility Model [18] and Matchup Hypothesis [19] to test the suitability of spokescharacters as communicators to promote sustainable behavior, taking the context of recycling. Additionally, the study examined spokescharacter traits that can enhance their persuasiveness. In the process, this study made four distinct contributions to the literature. First, this study adds to the literature that investigates how anthropomorphism can be used to promote sustainable behaviors. Whereas rich literature exists detailing the positive outcomes of anthropomorphizing nature itself and its non-human constituents on pro-environmental behavior (e.g., [29]), this study focused specifically on the persuasiveness of anthropomorphic characters as communicators, thus adding to the limited literature that explores the role of spokescharacters in promoting sustainable behaviors [7,29]. Even though our study used popular spokescharacters who consumers strongly associate with the brands they promote [6] as stimuli, participants still found them marginally relevant as recycling communicators. It is thus likely that consumers may perceive spokescharacters designed specifically to promote environmental causes to be even more appropriate environmental communicators. Research has also shown that the presence of spokescharacters makes communication humorous [11] and so, our study also answers a recent research call to explore the role of humorous communicators in environmental messages [28].

Second, we expand the environmental communication literature by empirically showing that spokescharacters are trusted communicators who can encourage the public to adopt sustainable behaviors such as recycling. Past research has established the effectiveness of using different trusted formal and informal communicators (e.g., scientists, family members, government officials, military leaders to name a few) in environmental communication [8,41,82]. However, limited studies have tested the efficacy of using spokescharacters in environmental communication. Motta, Ralston and Spindel [30] suggested that future research should explore additional trusted sources to engage the broader public with pro-environmental behaviors. Our study answers their call by showing that spokescharacters may be a plausible trusted source that can be used in environmental communication. Our study also adds to the limited literature that examined the role of source characteristics such as credibility and relevance in environmental marketing context [31] and relatively few studies have focused on the outcomes of source trustworthiness in environmental communication [31,41]. Our study fills this gap in literature by showing that spokescharacter trustworthiness positively influences recycling attitudes, which in turn enhances participants’ recycling intentions. Additionally, by showing that recycling attitudes fully mediates the effect of spokescharacter trustworthiness on recycling intentions, our study offers the theoretical mechanism through which spokescharacter trustworthiness enhances recycling intentions of consumers. Hence, our study establishes that by manipulating likability, expertise, and relevance traits, environmental marketers can enhance spokescharacter trustworthiness and as a result, their persuasiveness in environmental communication.

Third, in addition to confirming the indirect effect of matchup (relevance in our study) on consumers’ attitudes and intentions through source trustworthiness, our study additionally established that source relevance can have direct, positive influence on behavioral intentions of consumers, thus extending the limited previous research in non-profit contexts [69,70]. Finally, this study also adds to the limited literature that explores the independent role of source likability in enhancing message persuasion [22,23]. Recent studies have shown that cuteness [15] and humor [28] appeals in environmental communication can positively influence consumers’ behavioral intentions. Hence, as part of spokescharacter likability, these traits exerted positive effect on recycling intentions.

5.2. Practical Contributions

This study offers several practical contributions which are outlined next. First, the findings of our study provide evidence to environmental marketers that spokescharacters may be effective communicators in environmental campaigns for promoting sustainable behaviors like recycling. As a survey on spokescharacters [6] revealed, not all spokescharacters may be equally persuasive. Hence, environmental marketers need to know the traits that influence spokescharacters’ persuasiveness. Our study provides some insights for designing trusted and persuasive spokescharacters. High spokescharacter likability can be achieved by giving a cute and humorous appearance to the spokescharacter. Likability could also be enhanced by using relevant props or speech accents, developing a clever or witty personality for the spokescharacter, or creating a story around the spokescharacter by introducing the character’s family members, friends, and even pets [11,36,85].

The expertise of the spokescharacter can be augmented by giving a suitable background to the spokescharacter (i.e., knowledgeable, experienced, and skilled personality in the promoted cause such as recycling) through a story. Spokescharacters can be made relevant by giving them an appropriate appearance; for instance, an anthropomorphized recycling bin for promoting recycling, an aquatic animal for water and ocean conservation, a local, wild animal species for bushfire prevention, and an anthropomorphized bus for encouraging the use of public transportation. Finally, even though not directly tested in the study, spokescharacters created by city councils could be used to encourage children to adopt sustainable behaviors through holding events in schools, as the success of the Pride campaigns suggests [86].

5.3. Limitations and Future Research Directions

As with any research, this study has limitations. First, the data for this study was collected online through MTurk, raising the question about data quality. However, following marketing survey research best practices, we screened the data for outliers and eliminated responses that were incomplete or where respondents’ failed attention checks [74,87]. Second, a single, brief exposure (20 s) to a static recycling message under lab settings may not have produced the full effect of spokescharacters on participants’ attitudes and intentions as previous research shows that repeated exposure to spokescharacters in marketing campaigns make them persuasive communicators [21]. Future research could test whether multiple exposures to messages (i.e., increased frequency) enhances the effects identified in this research. Next, as we surveyed participants from the US, the findings of this study may not be generalizable in other cultural contexts or to other environmental issues apart from recycling. These limitations can be addressed by future research across other cultural contexts and by using other environmental issues (e.g., water conservation, adoption of solar energy) as context.

This study also offers meaningful future research directions to researchers. For instance, future research may explore additional antecedents of spokescharacter trustworthiness which may provide further insights to how perceived spokescharacter trust could be enhanced. One possibility is to test the effect of brand personality traits similar to the study by Folse, Netemeyer and Burton [37], who tested their effect on brand trust rather than on spokescharacter trust. Second, researchers can design and empirically test the effectiveness of new spokescharacters by manipulating different spokescharacter traits, such as cuteness, humor, type of appearance, and level of expertise and measure the influence on message persuasiveness. Researchers can also explore interaction effects of different spokescharacter traits on message persuasion (e.g., the effect of a highly cute but less humorous spokescharacter on persuasiveness). Additionally, as spokespersons’ gender can influence their perceived credibility [43,88], future studies can empirically test how spokescharacters’ gender moderates their effectiveness in an environmental marketing context. Qualitative studies with consumers may provide additional insights about designing new spokescharacters for environmental communication. Finally, future research can use field studies to confirm the efficacy of using spokescharacters to promote pro-environmental behaviors.

5.4. Conclusions

Environmental problems are largely attributable to unsustainable human behaviors. One way to encourage the adoption of sustainable behaviors such as recycling is through creating effective communication campaigns. This study explored how using anthropomorphized characters as endorsers affects the persuasiveness of environmental communication promoting recycling. The findings of this study suggest that using spokescharacters in a recycling message can enhance consumers’ recycling intentions. Moreover, we demonstrate that environmental communicators can manipulate the spokescharacter traits of likability, expertise, and relevance to enhance spokescharacter trustworthiness, which results in improved recycling attitudes and behavioral intentions of consumers. As limited literature has thus far examined the role of spokescharacters in encouraging sustainable behaviors, our study provides a foundation for future research in this important domain. As part of their communication strategy, environmental marketers may create and deploy spokescharacters in their communication campaigns to promote sustainable behaviors among the general public in the near future.