The effects of team identification on consumer purchase intention in sports influencer marketing: The mediation effect of ad content value moderated by sports influencer credibility

Abstract

This study investigates the effects of team identification on consumer purchase intention in sports influencer marketing in South Korea. Specifically, it examines the mediation effect of ad content value moderated by sports influencer credibility in the relationship between team identification and purchase intention. Team identification positively affected consumer purchase intention and ad content value. Also, ad content value mediated the relationship between team identification and purchase intention. Sports influencer credibility moderated the mediation effect of ad content value on purchase intention as well as team identification. This study provides implications for planning sports influencer marketing in the South Korean market.

Keywords:

• sports influencer marketing
• sports influencer credibility
• team identification
• ad content value
• purchase intention

PUBLIC INTEREST STATEMENT

Sports influencer marketing has received much attention from advertisers and consumers in South Korea. On social networking sites, sports influencers are able to connect closely with their fans and followers and subtly deliver branded content to them. Although various factors determine the effectiveness of sports influencer marketing, sports team identification may affect consumers’ purchase intention and evaluation of ad content value in influencer marketing. Thus, it is important to examine the complex relationship between sports team identification, ad content value, and purchase intention. Sports influencer credibility may also affect the relationships between team identification, purchase intention, and consumer perception of ad content value. By identifying the impact of team identification on ad content value and purchase intention, this study contributes to the understanding of sports influencer marketing in South Korea.

Influencer marketing is an increasingly popular, content-driven type of online marketing based on the idea that people tend to purchase products or services recommended by influencers (Influencer Marketing Hub, 2021). Due to its recent exponential growth, global ad spending on influencer marketing was projected to reach $10 billion by the end of 2020 (Mediakix, 2020). Sports influencer marketing has started attracting millennial consumers’ attention. For instance, Juventus soccer player Cristiano Ronaldo is the most followed individual on Instagram, and companies are more than glad to pay almost $1 million for his branded posts (GRIN, 2020; Lane, 2019). Although sports influencers are the perfect fit for sports, fitness, and athleisure brands (Raphael, 2017; Unbox Social, 2020), various brands in other fields have collaborated with sports influencers to take advantage of their fandoms and reputations (Barnett, 2020; Opendorse, 2021).

Social media influencers are individuals who have established social media followings and are regarded by their followers as role models and experts in their fields (Engaiodigital, 2020). Influencers range from nano influencers to mega influencers depending on the size and characteristics of their followings. For example, nano influencers typically have 1,000 to 10,000 followers and are suitable for small businesses with low budgets. Due to their close and authentic relationships with their fans, nano influencers have the highest engagement rate among influencers (HypeAuditor, 2019). On the other hand, well-known mega influencers with 500,000 to 2,000,000 followers help brands achieve maximum reach and increase brand presence. Although anyone from different backgrounds can be influencers, such as actors, models, authors, and entrepreneurs, and television and social media stars (Engaiodigital, 2020), this study will focus on mega sports influencers given the importance of achieving extensive reach and awareness using social networking sites (SNSs; Influencer Marketing Hub, 2021).

While developing a robust and effective social media strategy using sports influencers is crucial (Bellanger, 2018), no attempt has been made to investigate the impact of sports team identification. Team identification can be defined as the extent to which a fan feels a psychological connection to a team (Funk & James, 2001; Wann & Branscombe, 1993; Wann et al., 2001). Research has shown that attachment to a team is related to sports consumption behavior, such as attending events and purchasing team products (Fisher, 1998; Kwon & Armstrong, 2006; Madrigal, 2000). Team-related factors, such as team success, player attractiveness, and similarity to their fans, might have profound impacts on forming team identification, along with other environmental (e.g., socialization process) or psychological factors (e.g., the need to belong; Theodorakis et al., 2012).

The key objective of this research is to identify how team identification and the branded content posted by sports influencers systematically affect purchase intention in South Korea, which has the third highest rate of active SNS use globally (Statista, 2020). Furthermore, this study will investigate how sports influencer credibility moderates the complex relationships between team identification, ad content value, and purchase intention.

1. Literature review

1.1. Team identification

Team identification refers to the level of psychological attachment (Trail et al., 2000) or connection felt by a sports fan toward a favorite team and team’s performances (Kim & Kim, 2009; Wann, 2006; Wann et al., 2001). The concept of team identification can be explained by social identity theory (Tajfel & Turner, 1986), which describes psychological biases between in-group and out-group members. In-group members tend to have favorable attitudes toward their group members; due to this in-group favoritism, in-group members consider themselves superior to other groups (Hogg & Abrams, 1988). Similar to the notion of in-group and out-group members, the concept of team identification argues that sports fans are likely to engage in biased information processing toward their favorite team and the team’s performances (Wann, 2006).

Ample research has shown that fans with different levels of team identification produce significantly different affective, cognitive, and behavioral responses toward their favorite teams. Affectively, highly identified fans experiencing intense emotion express strong positive reactions to games that their team won and negative reactions to games that their team lost (Madrigal, 1995; Prayag et al., 2020). Their satisfaction with the team’s performance and goal relevance lead to optimism about the team’s chance of winning the next game (Madrigal, 2003). Cognitively, highly identified fans tend to have a self-serving bias, in which the individual attributes their success to internal factors but attributes their failures to external factors (Wakefield & Wann, 2006; Wann, 2006). Behaviorally, highly identified fans exhibit increased team-related consumption behaviors (Wann & Branscombe, 1993; Williamson et al., 2003) and purchasing team merchandise (Fisher & Wakefield, 1998; Kwon & Armstrong, 2002; Özgen & Argan, 2017).

Recent studies have supported the idea that team identification has a strong effect on behavioral responses. For instance, among fans with self-esteem spectator motivation, highly identified fans are more loyal and have significantly lower brand switching intentions than lowly identified fans (Lin, 2017). Cunningham and Eastin (2017) investigated the relationship between team identification, engagement, and self-efficacy with second screen use. Results showed that team identification and self-efficacy positively affected engagement with the second screen. Yim and Byon (2018) found that highly identified fans presented a more consistent pattern of consumption intentions than lowly identified fans regardless of the game outcome. In sum, highly identified fans presented a consistent level of consumption intentions irrespective of their game satisfaction or service satisfaction levels. Therefore, the following hypothesis is proposed:

H1: Team identification will have a direct positive effect on purchase intention.

1.2. Mediator variable: Ad content value

Successful influencer marketing starts with creating useful and valuable ad content for consumers. Based on the expected and outcome of advertising value (Ducoffe & Curlo, 2000), sports influencers are able to build authority, brand awareness, strengthen consumer engagement, and trust. Advertising value is defined as “a subjective evaluation of the relative worth or utility of advertising to consumers” (Ducoffe, 1995, p. 3). Ducoffe (1995, 1996) developed an advertising value model based on three antecedents of perceived value: informativeness, entertainment, and irritation. Informativeness is the ability of advertising to inform consumers of product or service alternatives. Entertainment is the ability of advertising to fulfill consumer needs. Finally, irritation describes the outcome of the techniques employed by advertisers to persuade consumers, who might find these techniques annoying, insulting, or manipulative. Generally, irritation reduces advertising effectiveness (Aaker & Bruzzone, 1985).

Extensive research has been conducted to understand ad content value in influencer marketing. For example, Khan and Saima (2020) found that information quality, entertainment value, and influencers’ trustworthiness directly affect influencers’ credibility and indirectly affect purchase intention. The informative value of influencer-generated content and the influencers’ trustworthiness, attractiveness, and similarity to their followers positively affect followers’ trust in the influencers’ branded posts, which leads to brand awareness and purchase intention (Lou & Yuan, 2019).

For both rational and emotional customers, advertising disclosure, attitudes toward ads, product utility, and the influencer’s credibility were found to be determinants in generating purchase intention in influencer marketing (Müller et al., 2018). Logan et al. (2012) found that informativeness and entertainment played key roles in affecting advertising value in SNSs. Although irritation did not directly affect advertising value, it had a direct negative effect on attitudes toward ads. Choi et al. (2019) revealed that interaction, informativeness, entertainment, professionalism, and freshness had direct positive effects on trust in information and indirect effects on purchase intention in one-person media. Byun and Kim (2020) reported that the advertising attributes of YouTube e-sports videos, such as information, individuality, and irritation, were closely related to advertising attitudes, product attitude, and purchase intention.

What then is the relationship between team identification and information processing of ad content? From the social identity and cognitive dissonance perspectives, extensive research has been conducted to understand the relationships between group categorization, biased information processing, and affective states (Knobloch-Westerwick et al., 2020; Schaller, 1991). The mere categorization of people into social groups automatically instigates a cognitive mechanism through which group-relevant information is processed in a biased manner (Schaller, 1991). In other words, categorization-based in-group favoritism might contribute to group stereotype formation. Due to confirmation bias, selective exposure is greater for messages (e.g., political news) that align with preexisting attitudes than for messages that contradict preexisting attitudes (Knobloch-Westerwick et al., 2020). People expecting less positive affect states are likely to engage in attitude-consistent exposure to enhance affect states or to avoid negative moods. Meanwhile, biased information processing could be adjusted by a need for cognition or cognitive reflection (Knobloch-Westerwick et al., 2020).

Similarly, highly identified fans tend to engage in biased information processing when they are exposed to brand cues (Ahn et al., 2013; Kwak et al., 2011). For instance, research on highly identified fans’ attitudes toward their favorite teams’ minor (e.g., Utah Jazz) or major (e.g., Golden State Warriors) logo changes showed that the fans had negative attitudes toward the redesigned logos but continued to purchase team merchandise regardless of the magnitude of the logo redesign (Ahn et al., 2013). Fans can adjust this attitude-behavior inconsistency by simply refuting counter-attitudinal messages or selectively exposing themselves to pro-attitudinal messages (Ahluwalia, 2000). Highly identified fans are more likely than lowly identified fans to undertake a biased memory search to continue to support their behavior. In addition, fans evaluate products licensed with a rival team’s logo as possessing less functional, emotional, and social value than products licensed with their favorite team’s logo (Kwak et al., 2011). Therefore, the following hypotheses are proposed:

H2: Team identification will have a direct positive effect on ad content value.

H3: Ad content value will have a direct positive effect on purchase intention.

1.3. Moderator variable: Sports influencer credibility

Influencer marketing is a hybrid of old and new persuasion tools that combines celebrity endorsement and content-driven marketing (Brooks, 2020; Childers et al., 2019). Celebrity endorsement is one of the most effective advertising strategies in that it helps increase brand awareness and recall, create positive feelings toward the brand, entertain consumers, and increase purchase intention (Choi & Rifon, 2012; Gupta, 2009). However, influencer marketing is different from celebrity endorsement as not all influencers are as popular or well-known to the public as celebrities. Mega influencers might enjoy greater popularity among the public and have the power to get their followers to share and amplify their brand posts (Childers et al., 2019). Although celebrity endorsement may be helpful in enhancing brand awareness and maximum reach, influencer marketing has a greater effect on encouraging consumer engagement by building emotional ties and authentic parasocial relationships with influencers (Solis, 2017). Nevertheless, benefits of both celebrity endorsement and influencer marketing lie in providing consumers with endorser credibility.

By integrating the source credibility model (Hovland et al., 1953) and source attractiveness model (McGuire, 1985), Ohanian (1990) proposed a celebrity endorser credibility model that consists of three elements: trustworthiness, expertise, and attractiveness. Expertise is defined as the extent to which a communicator is perceived to be a source of valid assertions about an object or issue, whereas trustworthiness is the degree to which consumers are confident in the communicator’s intent to communicate the assertions they consider most valid (Hovland et al., 1953). These two elements originated from the source credibility model. Attractiveness, which originally came from the source attractiveness model, refers to the attractiveness of the source (McGuire, 1985). Sports influencers who are perceived to be knowledgeable, reliable, and physically attractive are considered credible; subsequently, they generate positive attitudes toward the endorsed brand and increase purchase intention (Seiler & Kucza, 2017).

Meaning transfer theory and associative learning theory may provide a theoretical framework for understanding how influencer credibility works (McCracken, 1989; Till, 1998). McCraken (1989) emphasized the importance of understanding that celebrities symbolize a set of meanings to consumers. The same meanings are then moved to consumers by their purchase and use of the brand (Jain & Roy, 2016). Likewise, same process happens when an influencer promotes the brand in SNSs. Meanwhile, associative learning theory provides a mechanism for how an associative link between a brand and an influencer can be created. When an influencer promotes a particular brand in SNSs, an associative link is built between the brand and the influencer. Subsequently, new information about the brand will be integrated into the existing associative network (Meyers-Levy, 1989).

Recently, Taillon et al. (2020) found that social media influencers’ attractiveness and likability positively affected attitudes toward the influencers, word-of-mouth behavior, and purchase intention. Closeness positively moderated the effect of attractiveness on purchase intention as well as the effect of likability on attitudes toward the influencers. Compared with traditional celebrities’ brand posts, Instagram influencers’ brand posts make consumers perceive the source as more trustworthy, show more positive attitudes toward the endorsed brand, feel a strong social presence, and feel more envious of the source (Jin et al., 2019). Xiao et al. (2018) investigated factors affecting YouTube influencer marketing credibility within the framework of a heuristic-systematic model. They found that influencer credibility along with homophily, interactivity, and social advocacy positively affected information credibility, video attitude, and brand attitude.

Furthermore, Lim et al. (2017) examined the advertising effectiveness of social media influencers on purchase intention and the mediation effect of customer attitude. They found that source attractiveness, product match-up, and meaning transfer positively affected customer attitude and purchase intention. Chetioui et al. (2020) investigated how fashion influencers contributed to consumers’ purchase intention in Morocco. Using the theory of planned behavior, they demonstrated that perceived credibility, trust, perceived behavioral control, perceived subjective norms, perceived expertise, and perceived congruence positively affected attitudes toward fashion influencers. Based on these findings, the following hypothesis is proposed:

H4: Sports influencer credibility will moderate the effect of ad content value on purchase intention.

Research has been conducted to examine the interplay between team identification and the celebrity endorsement effect. For example, in the context of sports celebrity endorsements, compared with fans with low team identification, fans with high team identification perceived athletes’ credibility as significantly higher (Tzoumaka et al., 2016). Famous athletes serve as liaisons to their fans’ favorite teams, and highly identified fans actively support these athletes’ endorsements by purchasing their endorsed brands (Carlson & Donavan, 2008).

To effectively cope with negative publicity highly identified fans were more likely than lowly identified fans to use the rationalization and decoupling strategy to maintain their favorable attitudes toward the players and the team (Wang & Kim, 2019). This can be interpreted that highly identified fans were motivated to protect the athletes’ credibility by cognitively disassociating them from negative publicity. In the wake of a sports scandal, it is desirable for sponsors to maintain their relationship with the team because highly identified fans tend to support their own team but more harshly judge a rival team’s scandal (Chien et al., 2016). Taken together, sports influencer credibility is predicted to affect the effect of team identification (Knobloch-Westerwick et al., 2020), and the mediating effect of ad content value will depend on the level of sports influencer credibility. Therefore, the following hypotheses are proposed:

H5: Sports influencer credibility will moderate the effect of team identification on purchase intention.

H6: The mediation effect of ad content value in the relationship between team identification and purchase intention will be moderated by sports influencer credibility.

Figure 1 shows the research model of this study.

Figure 1. Research model

Note. Dotted line (H6) represents the mediation effect of ad content value moderated by sports influencer credibility.

2. Method

An online survey was administered through Google Forms to a convenience sample of university students majoring in advertising and public relations in South Korea who said that they had previous experience in using SNSs. Data were collected in May 2020. As an incentive to complete the online survey, extra course credit was provided to the student participants. In Korea, the social media user penetration rate in 2018 was highest for people aged 18–24 (82.1%), followed by those aged 25–34 (80.3%) and those aged 35–44 (64.5%) (Waldeck, 2020). The sample population of this study provides unique insights regarding the dynamics of team identification and sports influencer marketing in Korea. Due to South Koreans’ high autonomous-related self-construal, they are more likely to express their unique personality traits and characteristics in SNSs compared with other Asian countries (Lee et al., 2016).

2.1. Sample

A total of 350 respondents participated in this study. Of the 350 respondents of the online survey, 44.9% were male. The mean age was 22.5 years (SD = 2.03); 19.1% of the participants were 20 years old, 17.4% were 21 years old, and 16.6% were 22 years old. In terms of SNS platforms, 45.4% of the participants indicated that they used Instagram, followed by YouTube (29.1%), KakaoTalk (14.8%), Facebook (6.9%), and Twitter (4.0%). Finally, regarding the average hours of SNS use per day, 29.7% of the participants indicated that they used SNSs about two to three hours daily, 25.0% said they used SNSs one to two hours each day, 19.7% used SNSs three to four hours a day, 17.7% used SNSs more than four hours a day, and 8.0% used SNSs less than one hour per day. Because YouTube, Instagram, Facebook, and Twitter are the most popular influencer marketing platforms in South Korea (Info Cubic Japan, 2019; Mediakix, 2019), this study analyzed users of these SNS platforms (n = 299).

2.2. Measures

Measurement items from previous research were used to assess all the constructs. Prior to the data collection process, a translation–back translation method from English to Korean was used to ensure that the items were properly translated (Beaton et al., 2000), and pretests for the items were conducted on a small sample (Behr, 2017; Brislin, 1970). A questionnaire using a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree) was developed to measure team identification, ad content value, sports influencer credibility, and purchase intention.

Team identification is conceptually defined as the extent to which a fan feels psychologically connected to a team and its performances (Wann, 2006). Five items were developed to measure team identification (α = .91) after slightly modifying the organizational identification scales (F. A. Mael & Ashforth, 2001; F. Mael & Ashforth, 1992). Table 1 shows the measurement items, unstandardized factor loadings, standard errors, t values, standardized factor loadings, average variance extracted (AVE), and construct reliability (CR) for each construct.

Table 1. Overview of measurement items and reliability and validity measures

Construct	Subdimension	Item	B	SE	t	β	AVE	CR
Team identification	̶	When someone criticizes my favorite team, it feels like a personal insult.	1.00	̶	̶	.83	.94	.99
		I am very interested in what others think about my favorite team.	0.91	0.06	15.50	.77
		When someone praises my favorite team, it feels like a personal compliment.	1.12	0.60	20.25	.93
		When I talk about my favorite team, I usually say “we” rather than “they.”	0.97	0.06	16.96	.82
		My favorite team’s successes are my success.	0.86	0.06	13.99	.72
	Model fit indices: χ^2/df = 2.40, p < .05, IFI = .99, TLI = .99, CFI = .99, RMSEA = .07
Ad content value	Informativeness	SNS posts from the sports star influencers whom I follow supply relevant information on products.	1.00	̶	̶	.73	.80^a	.91^a
		SNS posts from the sports star influencers whom I follow provide the most up-to-date information on products.	1.16	0.12	9.65	.83
		SNS posts from the sports star influencers whom I follow supply product information when I need it.	1.13	0.13	8.71	.74
		SNS posts from the sports star influencers whom I follow provide useful product information.	1.04	0.12	8.90	.75
	Entertainment	SNS posts from the sports star influencers whom I follow are entertaining.	1.00	̶	̶	.76
		SNS posts from the sports star influencers whom I follow provide fun experiences.	1.12	0.10	10.98	.88
		SNS posts from the sports star influencers whom I follow make me feel better.	0.96	0.10	9.32	.74
		SNS posts from the sports star influencers whom I follow have many interesting contents.	0.80	0.10	8.11	.65
	Irritation	SNS posts from the sports star influencers whom I follow are irritating.	1.00	̶	̶	1.00
		SNS posts from the sports star influencers whom I follow disturb me.	0.84	0.04	19.93	.84
	Trust	SNS posts from the sports star influencers whom I follow are believable.	1.00	̶	̶	.83
		SNS posts from the sports star influencers whom I follow are reliable.	1.20	0.11	10.93	.96
	Model fit indices: χ^2/df = 2.50, p < .000, IFI = .94, TLI = .92, CFI = .94, RMSEA = .09
Sports influencer credibility	Trustworthiness	The sports star influencers whom I follow are honest.	1.00	̶	̶	.77	.86^b	.95^b
		The sports star influencers whom I follow are trustworthy.	0.99	0.07	13.48	.80
		The sports star influencers whom I follow are sincere.	0.84	0.07	11.87	.70
		The sports star influencers whom I follow are ethical.	0.99	0.07	13.83	.82
	Expertise	The sports star influencers whom I follow have good media reputations.	1.00	̶	̶	.73
		The sports star influencers whom I follow have outstanding performances.	0.92	0.11	8.13	.69
	Attractiveness	The sports star influencers whom I follow are attractive.	1.33	̶	̶	.68
		The sports star influencers whom I follow are handsome/beautiful.	1.25	0.12	11.59	.85
		The sports star influencers whom I follow are classy.	1.00	0.13	9.93	.66
		The sports star influencers whom I follow are friendly.	0.92	0.12	9.10	.60
	Model fit indices: χ^2/df = 2.24, p < .000, IFI = .95, TLI = .93, CFI = .94, RMSEA = .09
Purchase intention	̶	I would like to buy the brands that are endorsed by the sports star influencers whom I follow.	1.00	̶	̶	.79	.92	.98
		I feel like buying the brands impulsively when the sports star influencers whom I follow advertise them.	1.33	0.08	17.16	.89
		I am very likely to buy the brands that are advertised by sports star influencers whom I follow.	1.34	0.08	18.03	.93
		The brands that are endorsed by the sports star influencers whom I follow have the best quality.	1.09	0.08	13.70	.74
	Model fit indices: χ^2/df = 1.49, p < .05, IFI = .99, TLI = .98, CFI = .99, RMSEA = .04

Note. B = Unstandardized Factor Loading; SE = Standard Error; β = Standardized Factor Loading; AVE = Average Variance Extracted; CR = Construct Reliability.

^aThe CR and AVE values of ad content value were calculated based on the standardized factor loadings of each second-order factor: informativeness (.80), entertainment (.84), irritation (.13), and trust (.67). ^bThe CR and AVE values of sports influencer credibility were calculated based on the standardized factor loadings of each second-order factor: trustworthiness (.68), expertise (.81), and attractiveness (.78)

Ad content value refers to the subjective evaluation of the relative worth or utility of advertising to consumers (Ducoffe, 1995, 1996). Ad content value has four subdimensions: informativeness, entertainment, irritation, and trust (Ducoffe, 1996; Lou & Yuan, 2019; Soh et al., 2009). Informativeness and entertainment were measured with four items each (α = .91 and α = .87, respectively). Irritation was measured with two items (α = .90). In addition, two items were added to measure the trust aspect of ad content value (Lou & Yuan, 2019; Soh et al., 2009) (α = .82). Informativeness, entertainment, and trust reflect the positive aspect of cognitive and affective ad content value, whereas irritation captures consumers’ negative response to advertising (Logan et al., 2012; Lou & Yuan, 2019). Due to their opposite direction, items for irritation were reverse coded, and then the scores of all 12 items were averaged to create a composite index for ad content value.

Sports influencer credibility refers to a communicator’s positive characteristics that influence the receiver’s acceptance of a persuasive message (Ohanian, 1990). Sports influencer credibility has three subdimensions: trustworthiness, expertise, and attractiveness. A total of 10 items from past research (Lou & Yuan, 2019) were modified to measure sports influencer credibility. Four items were used to measure trustworthiness (α = .85). The expertise dimension was measured with two items (α = .70), and attractiveness was measured with four items (α = .80). The scores of all 10 items were averaged to generate the sports influencer credibility index. Finally, purchase intention can be defined as “an individual’s conscious plan to make an effort to purchase a brand” (Spears & Singh, 2004, p. 56). Purchase intention was measured with four items (α = .91), which were modified from the studies by Ahn et al. (2013) and Machleit and Wilson (1988).

To confirm the measurement reliability and the convergent and discriminant validity for each construct, first-order and second-order confirmatory factor analyses (CFA) were conducted. The CFA showed a reasonable model fit across four models. Even though the χ²/df values were significant, the model fit is highly sensitive to the sample size. Other fit indices were all within the ranges suggestive of a good fit. The standardized factor loadings of team identification, sports influencer credibility, and purchase intention were greater than .5, with a t value greater than 1.96 (p < .001), which suggests high convergent validity (Anderson & Gerbing, 1988). Regarding the irritation subdimension of ad content value, although the standardized factor loading of the second-order factor did not meet the cut-off value of .6 (Bagozzi & Yi, 2012), the irritation dimension was retained in the model due to the acceptable overall model fit indices. Discriminant validity was successfully achieved as the AVE for the latent construct was greater than the squared intercorrelation of any two variables (Fornell & Larcker, 1981). Table 2 shows the AVEs, correlations, and squared correlation coefficients of the constructs.

Table 2. AVE, correlations, and squared correlation coefficients

Construct	1	2	3	4	M	SD
1	.94 ^c	.10	.04	.08	4.59	1.46
2	.32**	.80	.29	.37	4.38	.82
3	.20**	.54**	.86	.18	5.14	.73
4	.28**	.61**	.42**	.92	3.58	1.36

Note. 1 = Team Identification, 2 = Ad Content Value, 3 = Sports Influencer Credibility, 4 = Purchase Intention. ^cThe diagonal elements (italic) represent the AVE values. Upper diagonal represents squared correlations of each construct.

*p < .05, **p < .01.

3. Results

To test Hypotheses 1 to 3, Process Macro (model 4) was employed (Hayes, 2013). This study employed 95% bias-corrected bootstrapped confidence intervals using 10,000 bootstrapped samples with a maximum likelihood model. Hypotheses 1 and 2 predicted the positive impact of team identification on purchase intention and ad content value, respectively. Hypothesis 3 posited that ad content value positively affects purchase intention. The results show that team identification had a significant direct positive effect on purchase intention (β = .10, p < .05, SE = .05, 95% CI [.01, .18]). Thus, Hypothesis 1 was supported. Team identification had a significant direct positive effect on ad content value (β = .32, p < .00, SE = .03, 95% CI [.12, .24]), and ad content value had a significant influence on purchase intention (β = .58, p < .00, SE = .08, 95% CI [.81, 1.12]). Thus, Hypotheses 2 and 3 were supported. Table 3 summarizes the results.

Table 3. Results of process macro (model 4) (n = 298)

Independent variable	Dependent variable	B	SE	t	β	LLCI	ULCI	H	Result
X	Y	.27	.05	5.09**	.28	1.85	2.85	H1	Accepted
R^2 = .08, F = 25.92 (p = .00), Total effect of X
X	M	.18	.03	5.70***	.32	.12	.24	H2	Accepted
R^2 = .10, F = 32.54 (p = .00)
X	Y	.10	.05	2.09*	.10	.01	.18
M		.97	.08	12.12***	.58	.81	1.12	H3	Accepted
R^2 = .39, F = 92.80 (p = .00), Direct effect of X (37%)
Independent variable	Dependent variable	Mediator	Indirect Effect	SE	Z	BC 95% CI
						LLCI	ULCI
X	Y	M	.18	.04	5.38**	.11	.26

Note. B = Unstandardized Path Coefficient; SE = Standard Error; β = Standardized Path Coefficient; BC = Bias Corrected; LLCI = Lower Limit Confidence Interval; ULCI = Upper Limit Confidence Interval; X = Team Identification; M = Ad Content Value; Y = Purchase Intention.

*p < .05, **p < .01, ***p < .001

Hypotheses 4 and 5 predicted the moderating effect of sports influencer credibility in the relationship between ad content value and purchase intention as well as in the relationship between team identification and purchase intention, respectively. To test Hypotheses 4 and 5, Process Macro (model 15) was used. Regarding Hypothesis 4, a significant interaction effect of sports influencer credibility with ad content value on purchase intention was found (B = .18, p < .05, SE = .09, 95% CI [.02, .35]). Ad content value had significant direct impacts on purchase intention when sports influencer credibility was one SD below (M = 4.40) and one SD above (M = 5.87) the mean (B = .71, p < .001, SE = .11, 95% CI [.49, .94] and B = .98, p < .001, SE = .11, 95% CI [.77, 1.19], respectively). These results show that sports influencer credibility additively affects the impact of ad content value on purchase intention. Thus, Hypothesis 4 was supported.

Regarding Hypothesis 5, sports influencer credibility had a significant interaction effect in the relationship between team identification and purchase intention. When sports influencer credibility was one SD below the mean, team identification had no significant direct impact on purchase intention (B = .08, ns, SE = .06, 95% CI [−.04, .20]). However, when sports influencer credibility was one SD above the mean, team identification had a significant direct impact on purchase intention (B = .12, p < .05, SE = .06, 95% CI [.01, .23]). Therefore, Hypothesis 5 was supported.

Hypothesis 6 predicted that the mediation effect of ad content value in the relationship between team identification and purchase intention would be moderated by sports influencer credibility. Test results confirmed the significant moderated mediation effect of ad content value (B = .03, SE = .02, 95% CI [.01, .09]). Specifically, although the path coefficients were all significant at both one SD below and above the mean, high sports influencer credibility (B = .18, SE = .05, 95% CI [.09, .28]) amplified the indirect effect of ad content value in the relationship between team identification and purchase intention more than low sports influencer credibility did (B = .13, SE = .03, 95% CI [.07, .19]). Thus, Hypothesis 6 was supported. Table 4 summarizes the results.

Table 4. Results of process macro (model 15) (n = 298)

Independent variable	Dependent variable	B	SE	t	LLCI	ULCI	H	Result
X	M	.18	.03	5.70***	.12	.24
R^2 = .10, F = 32.54 (p = .00)
X	Y	−.04	.26	−.14	−.54	.47
M		−.10	.46	−.21	−.99	.80
W		−.65	.37	−1.78	−1.38	.07
X x W		.03	.05	.54	−.07	.12
M x W		.18	.09	2.14*	.02	.35	H4	Accepted
Low W (4.40)		.71	.11	6.22***	.49	.94
High W (5.87)		.98	.11	9.13***	.77	1.19
R^2 = .41, F = 40.40 (p = .00)
Conditional direct effect		B	SE	t	BC 95% CI
X → Y					LLCI	ULCI
Low W (4.40)		.08	.06	1.35	−.04	.20	H5	Accepted
High W (5.87)		.12	.06	2.10*	.01	.23
Conditional indirect effect		B	SE		BC 95% CI
X → M → Y					LLCI	ULCI
Low W (4.40)		.13	.03		.07	.19
High W (5.87)		.18	.05		.09	.28
Index of moderated mediation		B	SE		BC 95% CI
					LLCI	ULCI
W		.03	.02		.01	.09	H6	Accepted

Note. B = Unstandardized Path Coefficient; SE = Standard Error; β = Standardized Path Coefficient; BC = Bias Corrected; LLCI = Lower Limit Confidence Interval; ULCI = Upper Limit Confidence Interval; X = Team Identification; M = Ad Content Value; Y = Purchase Intention; W = Sports Influencer Credibility.

*p < .05, **p < .01, ***p < .001.

A series of simple slope analyses were conducted to visualize the conditional effects in relation to Hypotheses 4 and 6. The association between ad content value and purchase intention was plotted when the level of sports influencer credibility was one SD below and above the mean. Figure 2 depicts the effect of ad content value moderated by sports influencer credibility. A simple slope analysis showed that the mediation effects of ad content value moderated by sports influencer credibility were significantly different at one SD below and above the mean. Figure 3 illustrates the mediation effect of ad content value moderated by sports influencer credibility. Figure 4 shows the standardized path coefficients of the research model.

Figure 2. Moderation effects of sports influencer credibility on ad content value

Figure 3. Conditional indirect effects of team identification

Figure 4. Unstandardized path coefficients for research model

Note. Dotted line (H6) represents the mediation effect of ad content value moderated by sports influencer credibility. Unstandardized path coefficients of H5 and H6 stand for one SD below and one SD above from the mean of sports influencer credibility.

4. Discussion

This study investigated the mediation effect of ad content value moderated by sports influencer credibility in the relationship between sports team identification and purchase intention in South Korea. Followers view influencers as trustworthy and having expertise, so they perceive influencers’ review or recommendations in the branded content as a form of social proof. Given the global popularity and fandom of sports stars in Korea, sports influencers have actively collaborated with many consumer brands to create favorable brand evaluations and purchase intention. With this in mind, this study examined how the mediation effect of ad content value in the relationship between team identification and purchase intention was moderated by sports influencer credibility. The major findings of the study are as follows:

First, team identification and ad content value had a positive impact on purchase intention. Highly identified consumers are more likely to purchase brands that are endorsed by sports influencers whom they follow in SNSs. This finding is consistent with those of past research that highly identified fans display increased team-related consumption behaviors (Wann & Branscombe, 1993; Williamson et al., 2003) and purchasing team merchandise (Fisher & Wakefield, 1998; Kwon & Armstrong, 2002; Lee & Ferreira, 2011; Lin, 2017; Yim & Byon, 2018). Highly identified fans tend to continue supporting their favorite teams even without considering the cause (Lee & Ferreira, 2011) and the outcome of the game (Yim & Byon, 2018). Thus, team identification could be a powerful predictor of behavioral responses in sports influencer marketing.

Second, team identification had a positive influence on ad content value. As long as sports influencers post useful and interesting branded content, highly identified consumers are likely to engage in biased information processing, which will result in developing consumer trust and increasing purchase intention (Choi et al., 2019; Khan & Saima, 2020; Logan et al., 2012; Lou & Yuan, 2019; Özgen & Argan, 2017). For highly identified consumers, ad content posted by sports influencers is perceived to be valuable due to categorization-based in-group favoritism (Knobloch-Westerwick et al., 2020; Schaller, 1991). To rationalize their cognitive process and avoid uncomfortable affective states, they tend to process group-relevant information (e.g., branded content posted by a famous sports star in their favorite team) biasedly by exposing themselves to messages that are consistent with preexisting attitudes (Knobloch-Westerwick et al., 2020; Schaller, 1991). Simultaneously, other research has indicated that a contrasting effect of biased information processing could occur among highly identified fans, who can make a more objective or critical evaluation of a product licensed with a rival team’s logo (Kwak et al., 2011). Thus, providing useful and entertaining content will attract more potential consumers.

Third, sports influencer credibility moderated the relationship between ad content value and purchase intention as well as the relationship between team identification and purchase intention. Higher sports influencer credibility incrementally affected ad content value and team identification. Regarding the conditional direct effect of team identification, purchase intention was significantly affected only when sports influencer credibility was high. The findings of this research resonate with those of past studies, which showed that not only the influencer’s usefulness or informativeness but also their trustworthiness, attractiveness, and similarity to their followers positively affected followers’ trust in the influencers’ branded posts (Lou & Yuan, 2019) and consumer purchase intention (Müller et al., 2018; Weismueller et al., 2020). By taking advantage of their reputation and credibility, sports influencers can promote the sales of their endorsed products (Chmait et al., 2020; Hazari, 2018; Tabor, 2020).

Finally, sports influencer credibility significantly moderated the mediation effect of ad content value in the relationship between team identification and purchase intention. The conditional indirect effect of team identification on purchase intention via ad content value was significantly greater when sports influencer credibility was high. As consumers become avid fans of particular sports teams, the success of their teams will have a huge impact on developing consumers’ self-confidence, self-esteem, social connection, and psychological well-being (Wann et al., 2015). Consumers will experience strong emotional bonds with their favorite teams and players. In this triadic relationship, highly identified consumers are more likely to perceive the usefulness of ad content posted by sports influencers whom they might follow on SNSs. Given that the indirect effect of team identification is greater than its direct effect, selecting authentic and genuine sports influencers who are willing to communicate with their followers should be a top priority (Lemmons, 2021). In particular, national sports influencers with high credibility may exert greater influence on purchase intention for low involvement products (Dikčius & Ilciukiene, 2021).

This study provides unique implications for brand managers and marketers attempting to utilize sports influencers in their marketing campaigns, particularly those entering the South Korean market. First, considering the catalytic effect of sports influencer credibility, it is important for marketers to increase sports team identification by developing effective sports marketing sponsorships. One of the biggest sports marketing trends is to utilize SNS channels in building fanbases and capitalizing on brand opportunities (Mulcahy, 2019). Fan communities on SNSs can hold brand-sponsored sessions or events to strategically deliver exclusive behind-the-scenes stories and livestreams of teams and players. More importantly, marketers should find appropriate sports stars who have a high fit with their brands and proactively communicate with consumers on these SNS channels. For example, various global and domestic brands, such as Coca-Cola, TAG Heuer, and Adidas, featured Korean soccer player Heung-Min Son in their marketing communication campaigns because of his popularity and global fanbase of 1.98 million social media followers (Kwak, 2019; Masters, 2019).

More broadly, it is important to increase women’s involvement in sports. If women become fans of a team or player, they will be more loyal to that team’s or player’s endorsed brands and will be more likely to purchase their products. The environment for women’s involvement in sports is developing more favorably than ever before. From an economic perspective, in the South Korean retail industry, female consumers in their 30s with a high disposable income have emerged as one of the fastest growing demographics for their trendiness and wide range of interests (Song, 2017). Due to the coronavirus pandemic, more female buyers are going online shopping to purchase essentials, and brands are adapting to meet their needs (Meyer, 2021). From a societal perspective, due to the importance of diversity and equality in sports, more and more women will be drawn into sports to fight gender bias and stereotyping. Also, the current trend of “athleisure” in South Korea reflects the growth of potential female buyers who desire active lifestyles (Hong, 2019).

4.1. Limitations and future research suggestions

Although this study offers useful implications for sports influencer marketing, it has several limitations. First, this study investigates the role of team identification in general. However, not all participants may be fans of a team or its athletes. A participant could be a fan of an individual player even without identifying with the team. While this study provided insights into participants’ general attitudes and behaviors, an experimental study would be more appropriate to reveal causal relationships among constructs. For example, by exposing the branded content posted by a real or fictitious player in a team, experiments can directly reveal subjects’ prior knowledge and associations.

Second, future research could factor in the need for cognition or cognitive reflection. Highly identified fans tend to exhibit unconditional attachment to their teams. Moreover, sports influencer credibility as a peripheral cue will affect the consumer decision-making process (Xiao et al., 2018). Biased information processing could be affected by the need for cognition or cognitive reflection because people with a high need for cognition tend to overcome intuitive but incorrect heuristics (Knobloch-Westerwick et al., 2020). These consumers are more likely to take the central route or systematic information processing.

Third, it would also be helpful to understand the broader scope of influencer marketing. Future research could focus on nano or micro influencers who have closer relationships with their followers than mega influencers with different SNS platforms. Individuals react differently depending on the type of influencers or characteristics of SNSs (e.g., the content type; Kim et al., 2019; Munar & Jacobsen, 2014). Also, team identification and interaction with sports influencers is likely to be affected by cultural differences (White & Lehman, 2005). By incorporating these factors, future research will provide more insightful suggestions for sports influencer marketing.

Additional information

Funding

The author received no direct funding for this research.

Notes on contributors

Jin Kyun Lee

Jin Kyun Lee received a PhD from the University of Texas at Austin (the Department of Advertising) in 2010. Currently, he is an associate professor of advertising at Hongik University in South Korea. His research mainly explores consumer psychology and behavior in a technology-mediated environment such as social networking sites. His other research examines strategic brand alliances and sports marketing. He teaches brand strategy, integrated marketing communications, and the principles of marketing. Prior to his current position, he was part of the journalism faculty at the University of Wisconsin Oshkosh. E-mail: [email protected]