“Effect of celebrity endorsers’ attributes on purchase intentions in South Africa”

Celebrity endorsement is the most used strategy since consumers value celebrities in advertisements. Organizations invest in celebrity endorsements in the expectation that they will enhance consumer perception of a product or service, evoke favorable at- titudes toward a brand, influence purchase intentions, and eventually increase sales and profits through increased consumer preference for a product, brand, or service. This paper examines the effect of endorser characteristics on purchasing intentions in South Africa. The survey sampled South Africans residing in Gauteng Province. The Qualtrics questionnaire was completed by 145 respondents ranging in age from 18 to 55 and living in various areas of Gauteng Province. Given the research objective, the paper analyzed only responses from participants who stated that they had been exposed to celebrity-related adverts. The findings show that these respondents saw advertisements featuring celebrities. To evaluate the hypotheses, regression analysis was performed. The results reveal that consumer behavior, such as purchase intention, is influenced by attractiveness ( β = .271; p < 0.05) and attitudes ( β = .520; p < 0.05). However, in this study, expertise ( β = .089; p > 0.05) and trustworthiness ( β = –.095; p > 0.05) were not related to purchase intentions. Understanding celebrity endorser’s attributes can help organizations determine the most successful marketing strategies for attracting new customers and maintaining existing ones.


INTRODUCTION
Celebrity endorsers, well-known people who appear in advertisements (McCracken, 1989), are not new in product or service marketing (Farrell et al., 2000). They are a pervasive aspect that dates back to the 18 th century (Garthwaite, 2014). Therefore, it is not surprising that their growing pervasiveness has sparked interest and has emerged as a potential strategy for advertisers. Not only do they contribute to the advertising industry's remarkable growth (Patil et al., 2013), they also have become central because they assist businesses by reaching out to target markets. Recently, consumers have demonstrated an increased interest in products promoted by celebrities. As a result, celebrity endorsers are increasingly important in attracting attention and influencing consumer behavior (Ohanian, 1991;Spry et al., 2011). To the extent that celebrity endorsers can transfer their values to the brands they endorse, they are likely to influence overall customer perceptions. According to Bennett (2016), perceptions are concerned with how individuals choose, organize, and interpret information received from the environment.
Often celebrity endorsers alter consumer perceptions, but they also shape the consumer decision-making process. Consumer perceptions of celebrities are determined mainly by assessing attributes such as attractiveness, expertise, and trustworthiness (Udo & Stella, 2015) and how these characteristics are transferred to the products they promote (Silvera & Austad, 2004). Faced with this prevalence of celebrity endorsers, Bush et al. (2004) wonder whether the use of celebrity endorsers "makes a difference on the bottom line for advertisers" (p. 108).
South Africa, like other countries, has demonstrated the same level of interest and is riding high on celebrity endorsements, which has resulted in many new opportunities and immense potential. Alekar (2014) affirms that South Africans are generally receptive to and identify with celebrity endorsers. However, in the case of South Africa, despite the progress in using celebrity endorsers, literature on the topic remains underdeveloped compared to other countries and is confined to relatively few studies.  Nyamakanga et al., 2019) in the country dealing with this topic has either omitted to examine the influence of endorser attributes on purchase intentions or has ignored a broad range of factors found in the endorsement theories. As a result, a gap in knowledge of this phenomenon exists. Understanding these characteristics and their influence on purchase intentions is expected to result in a future increase in the usage of celebrity endorsers. Furthermore, by examining attributes of celebrity endorsers on purchase intentions, this study assists marketers in determining characteristics of celebrity endorsers that play a role in impacting purchase intentions. Therefore, this paper contributes to the body of knowledge in a nation like South Africa, where few studies on celebrity endorsers exist.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
According to Yoo et al. (2000), consumer perceptions may influence the decision to purchase a product or brand. This is also true when such decisions are made based on an individual. Such is the case with celebrity endorsers. Celebrity endorsement theories stipulate that individuals look up to celebrities and use them as buying guides. As a result, these celebrities affect how consumers perceive and behave towards them and the brands they endorse. As reported by Joseph (1982), Kahle and Homer (1985), and Lafferty and Goldsmith (1999), attitudes, expertise, trustworthiness, and attractiveness seem to be critical attributes affecting consumer perception of celebrity endorsers, suggesting that consumers might reject endorsed products or brands if they do not perceive them to have these characteristics. Even celebrity attributes that may appear unimportant, such as physical characteristics or beauty, may, in fact, influence consumer decisions and exert a significant influence on how they perceive endorsed products. As a result, the celebrity endorser's characteristics are frequently transferred to the endorsed product or brand. These perceptions will also result in more positive attitudes and purchases (Ling et al., 2010).
Therefore, marketers choose celebrity endorsers solely for their ability to influence consumer attitudes over time. Fishbein (1967) defines attitudes as a learned predisposition of human beings to act positively or negatively toward some object or idea. Thus, attitudes toward an object or idea are defined by judgments, emotional feelings, and action tendencies toward that object or idea. According to Gaber et al. (2019), consumers generally perceive an advertisement positively or negatively.
Consequently, perceptions provide insight into the effectiveness with which a combination of celebrity endorser characteristics is likely to produce purchasing intentions for endorsed products or brands. There are numerous reasons proposed as plausible explanations for the association between intentions and celebrity endorsers. The literature places a high emphasis on traits like attractiveness, expertise, and trustworthiness, to name a few.
Among the endorser attributes examined, attractiveness is the one factor consistently mentioned in celebrity endorsement theories. Celebrity endorsers' attractiveness makes one believe that the product or service endorsed will be accepted in the eyes of the consumers (Spears & Singh, 2004). For this reason, the physical attractiveness of a celebrity is often the most highlighted aspect of an advertisement. For this reason, people are more receptive to ads featuring celebrities due to their level of beauty. According to Ohanian (1990), attractiveness consists of traits (e.g. physical appearance, personality traits, and an attractive lifestyle). These make an endorser appealing to the audience. However, Till and Busler (1998) caution that attractiveness may be beneficial only for specific product types and is unlikely to work for all promoted products or brands. Examining this is outside the scope of the present study. Generally, attractive celebrities are seen as having a beneficial effect on the products or services they endorse. Accordingly, as observed by Bhatt et al. (2013) and Sertoglu et al. (2014), celebrity attractiveness influences how an advertisement is perceived and significantly influences the desire to purchase the endorsed product or brand.
In the case of South African celebrities, as with others, their attractiveness contributes to their recognition and recall. In countries like South Africa, where the use of celebrity endorsers is in its infancy, one would speculate that attractiveness also plays an essential role in understanding the influence of celebrity endorsers on purchase intentions. Given what extant literature provides related to the role of attractiveness in purchase intentions, the present study explores how the attractiveness of endorsers affects purchase intentions.
Studies examining the factors influencing the use of celebrity endorsers have revealed endorser expertise as an essential element ( In such instances, people are more likely to purchase the product (Ohanian, 1990). These arguments support the hypothesis that endorser expertise affects consumer perceptions of celebrity-endorsed products and purchase intentions (Sertoglu et al., 2014). As a result, South African consumers are likely to judge a product based on how knowledgeable a celebrity is.
The literature on celebrity endorsement draws on trustworthiness, suggesting that it has primarily influenced purchase intentions. The concept refers to the endorser's general believability (Silvera & Austad, 2004) expressed as perceived honesty, integrity, and credibility (Erdogan et al., 2001). As a result, one expects consumers to form purchase intentions while forming perceptions of endorser trustworthiness and thus eventually purchase the endorsed products or brands (Onua et al., 2019).
Applying this argument to South Africa, it would be fascinating to learn whether the trustworthiness of a celebrity endorser has any relationship with purchase intentions.
Considering the preceding literature, the significance of celebrity endorser traits in influencing purchase intentions emerges. Purchase intention is a form of decision that evaluates a consumer's reason for purchasing a particular brand (Shah et al., 2012) and is frequently associated with consumer behavior, beliefs, and attitudes (Sallam & Wahid, 2012). Thus, many factors (attitudes, attractiveness, expertise, and trustworthiness) stated in celebrity endorsement theories could change general purchase intentions. However, despite the rich development of scholarship on celebrity endorsers, including its effect on purchase intentions, this area remains under-researched in South Africa. Therefore, this study examines how consumers in South Africa perceive the impact of celebrity endorsers' attributes on purchase intentions.
Based on the literature review, the following hypotheses are tested: H1: Consumer attitudes are positively related to purchase intentions.
H2: Celebrity endorsers' attractiveness is positively related to purchase intentions.
H3: Trustworthiness is positively related to consumer purchase intentions.
H4: Celebrity expertise positively influences consumer purchase intentions.

Participants and data collection
The participants were recruited online. Those who demonstrated an interest in taking part in the study were provided with a link to an online survey. 192 participants responded to the survey questionnaire. Of those responses, 47 were omitted from the analysis due to their inability to meet the qualifying criteria of this study. As a result, with a final sample size of 145 valid responses, the study achieved a response rate of 75.5%. The sample size used is comparable to that of Ertugan and Mupindu (2019) examining the effect of perceived endorser credibility on consumer purchase intentions.

Measures and questionnaire design
A quantitative survey approach was employed to collect the necessary data. The questionnaire included items from the source attractiveness and source credibility models and the theory of planned behavior. The questionnaire used a sixpoint Likert scale, with 1 denoting strong disagreement and 6 denoting strong agreement. Section A of the questionnaire requested biographical information about the respondents, including age, gender, income, and level of education. Section B of the questionnaire requested financial information and asked participants to respond to celebrity endorsers' perceived trustworthiness, feelings about the recommended product, and purchase intentions. Some of the questions on financial product purchases are scenario-based rather than addressing the actual purchase.

Descriptive statistics
Measures of central tendency (mean), deviation (standard deviation), and shape (kurtosis and skewness) were calculated. Table 2 presents the results of this analysis. The results show that the data were normally distributed as none of the shape (kurtosis and skewness) measures fell outside the recommended range of -2 and +2 (George & Mallery, 2010).

Results of reliability
The paper established the reliability of the items used. Cronbach's alpha values exceeded 0.80. Cronbach's alpha values have been computed and are shown in Table 3.

Correlation analysis
Before performing the regression analysis, it is important to investigate the correlations between the variables in this study. The correlation coefficients for the variables assessed in this study are listed in Table 4.
As demonstrated in Table 4, the items of the constructs are all associated, with the strongest correlation (r = 0.700) between expertise and trustworthiness. In comparison to the other dimensions of credibility, attractiveness had the strongest relationship (r = 0.681) with attitudes (r = 0.681) and intentions (r = 0.615) (expertise and trustworthiness). Trustworthiness is positively associated with both attitudes (r = 0.535) and intentions (r = 0.371). Moreover, there was a positive correlation (r = 0.464) between attractiveness and trustworthiness. There was the lowest association between intentions and trustworthiness (r = 0.371) and between intentions and expertise (r = 0.336).

Hypothesis testing results
Regression analysis was performed to assess the hypotheses of this study, using attractiveness, expertise, trustworthiness, and attitudes as the independent variables and intentions as the dependent variable. Purchase intentions account for 50.6% of the variation in the model, which takes attitudes, attractiveness, expertise, and reliability into account (as shown in Table 5). According to the data in Table 6, consumer attitudes (β = .520; p = 0.000; p < 0.05) and attractiveness (β = .271; p = 0.00; p < 0.05) have a statistically significant positive relationship with customer purchase intentions. While the relationship between expertise and purchase intentions is positive (β = .089; p = 0.282; p > 0.05) it is not statistically significant. Trustworthiness (β = -.095; p > 0.05) is negatively related to intentions and not statistically significant (p = 0.282).

DISCUSSION
This study found a significant relationship between consumer attitudes and consumer purchasing intentions. This comes as no surprise, considering that previous studies have revealed a similar finding. For example, Ayeh (2015) found that consumer attitudes toward a brand or product strongly predict future purchase intentions. Another significant result was the existence of a link between celebrity attractiveness and purchasing intentions. Gong and Li (2017) observed that a celebrity endorser's attractiveness positively affects the advertisement of a product they endorse, showing that advertising with attractive celebrity endorsers is more likely to generate interest in the promoted product or service. This outcome supports Sertoglu et al. (2014): an endorser's perceived attractiveness positively affects the intention to purchase a product. Note: ** means the correlation is significant at the 0.01 level (2-tailed).
In terms of another variable examined in the study (expertise), there was no evidence of a statistically significant relationship between a celebrity endorser's expertise and consumer purchase intentions. These findings also corroborate Tzoumaka et al. (2016), who indicated that a celebrity endorser's expertise is unlikely to influence purchasing intentions.
Although not statistically significant, this study reveals a negative relationship between the trustworthiness of a celebrity endorsement and purchase intentions. The findings on trustworthiness are most likely due to respondents' belief that celebrities' fame has contributed to their trustworthiness (Silvera & Austad, 2004

CONCLUSION
The study investigated the effect of celebrity endorsers' attributes on purchase intentions in South Africa.
The findings indicate that those celebrities deemed as attractive will draw more favorable attitudes from consumers. As a result, as proven in prior studies, attractiveness can be utilized as an attitude indicator. In other words, if respondents find celebrities attractive, this assumes that they will have positive attitudes towards such products and will buy products promoted by such endorsers. As a result, this paper proposes that marketers interested in using celebrities to promote their products should regularly assess their social media platforms. That may provide some useful insight into which characteristics indicate attractiveness from the standpoint of followers.
This paper proposed a link between expertise and purchase intentions. However, such a connection was not supported. These findings refute generally held beliefs in the literature about the significance of expertise in celebrity endorsements. Thus, the conclusion reached is that, because their admirers have automatically accorded celebrities an expert position, this status may instantly transfer into individuals having a strong belief that celebrities are knowledgeable about the products they promote. This may be the case if celebrity followers assess celebrity endorsers' social media profiles, anticipating that the information posted on a celebrity's social media page will act as an additional source of trust.
Notably, no statistically significant relationship existed between trustworthiness and intentions. Possibly, respondents in this study do not see celebrity endorsers as the most effective means of providing consumers with credible information that could help in the purchase decision. However, identifying trustworthy celebrities is frequently viewed as critical when selecting an endorser. With these findings, marketers should ensure that their messaging is honest and does not contradict the marketed product.
In conclusion, this paper provides additional insight into the attributes that South African consumers may appreciate most in celebrity endorsers. Furthermore, the contradictory results of this study provide a context for further empirical investigation.