Attitudes toward Fashion Influencers as a Mediator of Purchase Intention

Abstract

Fashion influencers are a new phenomenon and profession to which many young individuals may currently aspire; such is its impact in the digital and online world. Hence, the article serves an upcoming group of fashion-influencers-to-be, as well as firms that seek the help of such professionals. This study aimed to test the mediating role of the attitude toward influencers in the relation between, on the one hand, perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, and attractiveness, and, on the other hand, purchase intention. Path analysis was used to test a conceptual model in which attitude toward influencers mediates the relation between perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, attractiveness, and purchase intention. Among the seven components, the association between perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity, on the one hand, and purchase intention, on the other, was completely and significantly mediated through attitudes toward influencers. It was found that the attitude toward the influencer determines the purchase intent; this attitude is, in turn, conditioned by the competence, the resemblance, and the proximity that the consumer perceives in the influencer. Thus, to lead the consumer to buy a certain product, influencers must pay attention to perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity with the product (or service).

1. Introduction

This empirical study has, as its main objective, a focus on digital influencers, a new and important reality in the digital world and the age of digital marketing. How do they affect the sales of fashion products and brands? Do you have to be rich and famous and be a world beater to be an influencer (e.g., such as Cristiano Ronaldo and Roger Federer are, for their sponsorship brands)? Do you have to be beautiful (like Jelena Noura “Gigi” Hadid, or her sister Bella Hadid, or like Hailey Bieber, Justin Bieber’s wife)? Alternatively, do other traits stand out as being as important or even more important than this to obtain a following and influence people (e.g., more mundane but not necessarily easier traits to possess—such as competence, likeability, proximity, and the ability to seduce)? The large convenience sample of the survey may point to new and important directions regarding online and digital marketing in what is a pilot study and, as such, warrants a deeper subsequent study on the topic.

Indeed, digital influencers are nowadays a widespread phenomenon—ranging from cosmetics to lifestyle products and services—whereby skilled and marketing-savvy individuals are able to earn a living through their visibility online. Hence, understanding their real impact is crucial, mainly considering how they affect consumer behavior. Given this scenario, one aims to analyze how the attitude towards digital influencers affects purchase intention (the task of an influencer is to influence the close of a sale). More specifically, one intends to understand how the perceived characteristics of the influencers can predict purchase intention and find an explanatory model of purchase intention based on the impact of influencers on consumers. This will aid marketers in choosing digital influencers for specific fashion marketing campaigns (linked to sports, executive clothing, or beachwear, for example).

For this purpose, a survey was applied to 501 individuals from 24 April to 15 May 2021. A total of seven hypotheses were developed that discuss essential issues in marketing today—including trust, intention to buy, expertise, likeability, similarity to others, and attractiveness. The verification of the established research hypotheses was done by using the statistical technique of multiple regression based on the ordinary least squares (OLS) method, specifically based on a path analysis model.

This research has important theoretical and practical implications. On the one hand, it brings contributions to the marketing literature in terms of consumer behavior. On the other hand, it contributes to information science by highlighting the behaviors resulting from the interactions in online social networks, embodied mainly in the relationships between digital influencers and their followers.

In theoretical terms, the results suggest that the perceived characteristics of a given digital influencer that most influence purchase intention are those related to developing a relationship of trust to reduce the uncertainty felt by their followers. What is trust, and why is it important? Human beings enjoy and have a need to belong to groups [1]. Therefore, one tends to trust people from one’s in-group. To be able to reach anyone from outside a group will mean that gaining trust will be essential. Furthermore, in practical terms, the evidence suggests that the effectiveness of digital influencers in terms of purchase intention can be enhanced through improvements in their perceived credibility, perceived characteristics of expertise (preferably related to the product or service in hand), likeability, similarity (to the consumer), familiarity (the extent to which the consumer feels and relates to the influencer as an endeared and close friend), and attractiveness (normally seen as being a more superficial trait, attractiveness may also be seen as coming from deep within, something not readily visible to the beholder, something that we acknowledge that some people have, an inner warmth).

Besides this brief introduction, this research is structured into six other points. Next, the theoretical framework and research hypotheses are presented. Immediately following this, the materials and methods are discussed. Next, the results are presented, constituting the central point of the research. Immediately thereafter, the discussion of the results, the conclusions, and the references used are presented.

2. Theoretical Framework and Research Hypotheses

2.1. Fashion Influencers

Fashion influencers (FIs) are persons with many followers on social media; they produce and publish fashion content and have the power to influence followers’ opinions and purchase behavior [2]. Influencers carry out a meaning transfer process that stimulates interest and influences consumers’ attitudes toward products or brands and purchases [3]. E-word of mouth (E-WOM) is more effective when carried out by recognized personalities with a greater impact on online consumers’ purchase intention [4,5]. The extant literature recognizes that influencer marketing contributes to consumer purchase intention [2,6,7], and that fashion consumers are highly influenced by trends, which are conveyed by fashion influencers, namely through social media [8]; therefore, the fashion industry has strongly considered and used FIs [9].

2.2. Purchase Intention and Attitude toward the Influencer

Purchase intention (PI) (customer or buyer’s intent) measures consumers’ propensity to buy a product or service. PI is the “total of cognitive, affective, and behavioral toward adoption, purchase, and use of the product, services, ideas or certain behaviors” [10]. In 1969, Howard and Sheth developed a model of consumer decision—the Theory of Buying Behavior [11]. This theory contained psychological, marketing, and social issues, suggesting that these factors affect the consumers’ decision-making processes; the outcomes of this model are attention, comprehension, attitude, intention, and purchase. Finally, Javed et al. [12] presented the multi-step flow theory, suggesting that influencers have an extended network that passes the content to different levels.

Although the literature has shown the importance of attitudes toward purchase intention [13], attitude toward the influencer is not a well-defined construct [14]. Perceived influencer–product congruence positively affects followers’ perceptions of credibility and attitudes toward influencers [15].

2.3. Perceived Credibility

Perceived credibility is the degree to which the consumers trust the products as well as the degree to which they trust the content about them [16]. The Source Credibility Theory states that trustworthiness, expertise, similarity, and attractiveness are elements of source credibility [17,18]. Munnukka et al. [19], Lou and Yuan [20], and Balaban and Mustățea [21] found that attractiveness, trustworthiness, expertise, and similarity are perceived as important elements of the perceived credibility of social media influencers.

In situations of uncertainty, consumers tend to seek information from other consumers or similar others [22,23], so the way in which they perceive them becomes a central issue that will influence their purchase decision process. In this sense, the visual presentation of the extroversion of a given influencer has a direct positive influence on the perceived credibility of the same, which in turn will increase the chances of purchase intention [24], the main objective of commercial firms and campaigns.

Empirical evidence points to the positive influence of perceived credibility—also referred to as the authority of experts [25]—on the attitude toward the influencer, attitude toward the brand, and purchase intention [2,26,27]. In addition, the perceived credibility of any given content is one of the most important factors for subscription and purchase intention on YouTube [28]. Similarly, the credibility of influencers who are active on the social networks YouTube and Instagram is positively associated with purchase intention [3]. Therefore, one hypothesizes that:

Hypothesis 1 (H1).

Attitude toward influencers mediates the relationship between perceived credibility and purchase intention.

2.4. Trustworthiness

As a moral value, trustworthiness is the quality of attaining and maintaining a trusted, believable, measurable, and verifiable state. Thus, trustworthiness is the attribute of an individual that ensures credibility, dependability, and honesty. “Issues regarding the trustworthiness of social media are often biased and fake feedback generated from deliberate manipulation of online reviews” [29].

The trustworthiness of a given influencer plays a major role in changing the consumers’ perception of the levels of honesty, sincerity, and truthfulness [30], constituting a fundamental element in determining the purchase intention [31].

Empirical evidence points to the fact that the trustworthiness of social media influencers is an important predictor of purchase intention [21,30,32,33,34,35]. Abdulah et al. [36] found that trustworthiness, likeability, and familiarity influence Instagram users’ purchase intention toward fashion products. In addition, trustworthiness has an indirect effect on purchase intention when mediated by the quality of the digital influencers’ content [37]. In the specific case of celebrities, the trustworthiness variable was identified as one of the variables that positively affects purchase intention [38]. Moreover, Wiedmann and von Mettenheim [39] found that the most important requirement for purchase intention is the trustworthiness of influencers. Thus, one hypothesizes that:

Hypothesis 2 (H2).

Attitude toward influencers mediates the relationship between trustworthiness and purchase intention.

2.5. Perceived Expertise

The literature has pointed out that digital influencers have a relevant role in terms of persuasion, especially because they influence consumers’ purchase intentions. Among the determinants of purchase intentions is expertise [31], leading to much of what one sees when dental doctors invariably appear in toothpaste commercials [22,25], being also an influential element in the purchase intention when it comes to celebrities [38], very evident in the sale of perfume, which sells a lifestyle and a desire for status linked to the product, but also linked to other items as banal as shampoo (e.g., the influencer and footballer Cristiano Ronaldo) or, on the other hand, luxury items such as Louis Vuitton handbags and accessories (the case of the influencer and singer Bono Vox, from U2).

Empirical evidence points to the fact that the expertise of social media influencers is an important predictor of purchase intention [3,32,33]. In addition, trust (a much-studied variable in the social sciences) and perceived expertise have positive influences on attitude toward influencers, purchase intention, and attitude toward the brand [2]. Accordingly, one argues that:

Hypothesis 3 (H3).

Attitude toward influencers mediates the relationship between perceived expertise and purchase intention.

2.6. Likeability

Likeability can be understood as a psychological factor that influences how consumers react to a brand, product, or service, particularly as a result of how it is presented in terms of physical attractiveness, likeability, flattery, and association [40,41,42]. In this sense, creating content that appeals to likeability in branded communities on Facebook can lead to increased online engagement, which in turn can result in loyalty, word-of-mouth, and purchase intention [43]. One tends to buy from people whom one likes, and not the contrary [22]; hence, being able to establish empathy is an important and marketable trait present in prominent salespeople, whereby one witnesses an increasing blurring of the offline and online communities and what works in each. De Veirman, Cauberghe, and Hudders [44] showed that Instagram users with a high number of followers (20,000 or more) are considered more likable because they are considered more popular; this may suggest that low numbers of followers might negatively impact users’ likeability.

Regarding digital influencers, likeability has a positive predictive influence on both attitudes toward influencers and brands, word-of-mouth behavior, and purchase intention [14,45]. Similarly, the likeability of a specific brand has a positive influence on the purchase intention of its consumers [41]. Therefore, one argues that:

Hypothesis 4 (H4).

Attitude toward influencers mediates the relationship between likeability and purchase intention.

2.7. Similarity

Similarity can be understood as the degree to which people converge in terms of beliefs, values, or other aspects [46], influencing the word-of-mouth of the followers when considering the relationship between them and a particular influencer [14]. The much-aspired word-of-mouth on social media can be described as online gossip as consumers are both entertained and seek convergence and approval from their peers and influencers. In addition, similarity has a mediating effect between the endorsement relationship formed by an influencer and the intention to purchase the endorsed product [47].

According to Fu et al. [48], external (without deep interactions, e.g., name, age, gender, birthday, place of residence) and internal (deep interactions, e.g., values, interests, attitudes or opinions, and preferences) similarity affected users’ perceived usefulness, enjoyment, and trust transfer, which, in turn, exerted impacts on users’ social consumer behaviors. Empirical evidence points to the fact that similarity has an influence on the trustworthiness of a given influencer [34], which in turn is an important predictor of purchase intention [21,30,32,33,34,35,49]. Reputation is everything, one is tempted to conclude, as recent developments have shown, not the least of which occurred at the Oscars in 2022. In this sense, one hypothesizes that:

Hypothesis 5 (H5).

Attitude toward influencers mediates the relationship between similarity and purchase intention.

2.8. Familiarity

Familiarity can be understood as a set of experiences accumulated by a consumer about a given product or service over time [50]. Familiarity increases the level of comfort of the receiver toward the sender, making the sender seem more persuasive [51]. Empirical evidence suggests that consumers’ intention is strongly influenced by familiarity with the source and the convenience of the information [52]. If the follower is more familiar with the influencer, it has a positive relationship with purchase intention [51]. Moreover, influencers are likely to be more effective than celebrities when the audience is very familiar with the endorser [53]. When it comes to brands, the familiarity and quality of information have significant impacts on engagement with a brand on social media, which in turn can influence purchase intention [54]. On the other hand, attitudes toward a brand and purchase intention are influenced by both the likeability and familiarity of a given digital influencer [45]. One buys from similar others whom one likes and respects.

In this sense, familiarity is one of the determinants of endorsement by celebrities [55]. In addition, familiarity with a particular brand has a moderating effect on the recognition, attitudes, and purchase intention of co-appearing products [56]. Thus, one argues that:

Hypothesis 6 (H6).

Attitude toward influencers mediates the relationship between familiarity and purchase intention.

2.9. Attractiveness

McGuire’s attractiveness model states that the effectiveness of a message depends on the source’s “familiarity,” “likeability,” “similarity,” and “attractiveness” to the respondent [57]. The literature points out that attractiveness is one of the seven determinants of purchase intention, standing alongside other attributes such as trustworthiness, brand attitude, influencer’s credibility, para-social interactions, and expertise [31]. When materialism is high, attractiveness has a great influence on purchase intention [32]. In the specific case of celebrities, attractiveness is one of the determinants of purchase intention [38].

Regarding digital influencers, attractiveness was identified as one of the relevant attributes that influences elements such as brand image, brand satisfaction, brand trust, and purchase intention [39]. In addition, attractiveness has an impact on online engagement and purchase intention [58] and positively predicts attitudes toward the influencer, word-of-mouth, and purchase intentions [14]. Moreover, Liu et al. [59] found that attractiveness is indeed an important factor affecting consumers’ purchase intention. Therefore, one hypothesizes that:

Hypothesis 7 (H7).

Attitude toward influencers mediates the relationship between attractiveness and purchase intention.

3. Materials and Methods

3.1. Procedures

All procedures were carried out following the Declaration of Helsinki and its later amendments. The research protocol included informed consent (which contained the study’s objectives, ensuring the confidentiality and anonymity of the participants, and the procedures dedicated to the destruction of data after five years). Authorization from the author of the instrument to use it was requested. The instrument was then translated from English to Portuguese and then back-translated to English to ensure the quality of the translation. The protocol was disseminated online and we created a page on a social network for this purpose; therefore, the sample was random and of convenience.

3.2. Measures

The research protocol included a sociodemographic questionnaire with four components: gender (male/female), age, education (elementary school, high school, university), and professional status (active—student, employee, housewife, caregivers; and inactive—unemployed, retired, sick).

The research also included an instrument based on the literature, containing 30 items distributed by nine dimensions (

Table 1. Sociodemographic characteristics: respondents’ profile.

Variables		N	% Total	Cumulative %
Gender	Female	362	72.3	72.3
	Male	139	27.7	100.0
Age	M ± SD; Min–Max	21.1 ± 2.32; 15–25
Education level	Basic education	14	2.8	2.8
	Secondary education	289	57.7	60.5
	Higher education	198	39.5	100.0
Occupation	Inactive	55	11.0	11.0
	Active	446	89.0	100.0

): purchase intention (three items—two items by Chetioui et al. [2]; one item by Chun et al. [60]; one item by Lou and Kim [61]); perceived credibility (four items by Chetioui et al. [2]); trustworthiness (three items by Chetioui et al. [2]); perceived expertise (three items—two by Chetioui et al. [2] and one by Chun et al. [60]); likeability (four items by Chun et al. [60]); similarity (three items—the first one by Chun et al. [60] and the last two by Munnukka et al. [19]); familiarity (two items by Chun et al. [60]); attractiveness (four items by Munnukka et al. [19]) and attitude toward influencer (four items by Chetioui et al. [2]). Respondents were asked to evaluate each statement using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree).

3.3. Data Analysis

Data analysis included specific procedures of descriptive statistics (mean, standard deviation, minimum and maximum, skewness, and kurtosis) and inferential statistics (differences, correlations, regressions, and path analysis). Descriptive statistics were used to characterize the sample, items, and dimensions. Differences in purchase intention score by demographic variables were assessed using Student’s t-test (two levels) and F-test (three or more levels). Correlations were performed to analyze the associations between independent variables (perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, attractiveness, and attitude toward influencer) and the dependent variable (purchase intention). Multivariate regression models, one without demographic variables as covariates and another with covariates, were conducted to assess the association between perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, attractiveness, attitude toward influencers, and purchase intention. Path analysis was used to test the conceptual model in which attitude toward influencers mediates the relation between perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, attractiveness, and purchase intention. Criteria used in the path analysis were χ²/df < 2.000; CFI and TLI > 0.90); RMSEA < 0.05; SRMR < 0.06; PCLOSE > 0.05 (Kline, 2015). Significance was established at p < 0.05. Statistical analyses were performed with AMOS, version 27 [62].

4. Results

4.1. The Sample

The sample consists of 501 participants, of which 362 (72.3%) are women. The average age is 21.1 years (SD = 2.32; Min = 15, Max = 25). The majority of the sample (289/57.7%) completed high school; the remaining 198 (39.5%) completed higher education, and 14 (12.8%) elementary school. The vast majority of the sample (446/89.0%) is professionally active.

The sample connects to social media, on average, 6.6 times a day (SD =1.88; Min = 1; Max = 8); the most used social media platform is Instagram (95.4% of the sample), followed by YouTube (78.2%), TikTok (55.3%), Facebook (50.7%), Twitter (47.9%), Pinterest (39.5%), and others (9.2%). Moreover, the most used social media platform to follow influencers is Instagram (91.2% of the sample), quite far from the following social media, YouTube (40.9%), Tik Tok (24.6%), Facebook (14.0%), Pinterest (14.0%) Twitter (9.0%), and others (4.6%).

4.2. Frequencies

Table 2. Items and scales frequencies.

Items and Scales		M	SD	Sk (0.109)	Krt (0.218)	α
Perceived credibility Item 1	I do believe that fashion influencers I follow are convincing	3.60	0.83	−0.863	1.261
Perceived credibility Item 2	I do believe that fashion influencers I follow are credible	3.28	0.87	−0.269	−0.109
Perceived credibility Item 3	I do believe that fashion influencers advertising is a good reference for purchasing products	3.67	0.99	−0.783	0.241
Perceived credibility Item 4	I find purchasing product/service advertised by fashion Influencers I follow to be worthwhile	3.35	0.93	−0.479	0.099
Perceived credibility Total		3.48	0.75	−0.651	0.984	0.85
Trustworthiness Item 1	I do believe that I can depend on fashion influencers I follow to make purchasing decisions	3.26	1.03	−0.496	−0.439
Trustworthiness Item 2	I do believe that fashion influencers I follow are sincere	3.41	0.93	−0.540	0.222
Trustworthiness Item 3	I do believe that fashion influencers I follow use the same products they advertise	3.22	1.01	−0.242	−0.603
Trustworthiness Total		3.29	0.85	−0.332	−0.067	0.83
Perceived expertise Item 1	The fashion influencers I am following are experts in their field	3.13	0.99	−0.195	−0.287
Perceived expertise Item 2	The fashion influencers I am following have great knowledge	3.29	0.94	−0.377	−0.063
Perceived expertise Item 3	The fashion influencers I am following are experienced in their field	3.41	0.90	−0.564	0.197
Perceived expertise Total		3.27	0.83	−0.363	0.308	0.86
Likeability Item 1	The fashion influencers I follow are warm persons	3.57	0.80	−0.588	1.056
Likeability Item 2	The fashion influencers I follow are likeable persons	3.82	0.77	−0.890	2.156
Likeability Item 3	The fashion influencers I follow are sincere persons	3.47	0.80	−0.352	0.827
Likeability Item 4	The fashion influencers I follow are friendly persons	3.73	0.78	−0.733	1.663
Likeability Total		3.65	0.68	−0.837	2.521	0.89
Similarity Item 1	I am similar to the fashion influencers I follow on overall lifestyle	2.74	1.14	0.170	−0.885
Similarity Item 2	I have a lot in common with the influencers I follow	2.86	1.09	0.056	−0.736
Similarity Item 3	I am a lot alike the influencers I follow	2.75	1.07	0.132	−0.797
Similarity Total		2.79	1.03	0.126	−0.694	0.93
Familiarity Item 2	I have knowledge about the fashion influencers I follow.	3.41	0.97	−0.861	0.321
Familiarity Item 3	I easily recognize the fashion influencers I follow	3.63	0.98	−1.080	0.907
Familiarity Total		3.52	0.91	−1.042	1.025	0.84
Attractiveness Item 1	The fashion followers I follow are very attractive	3.55	0.83	−0.661	1.130
Attractiveness Item 2	The fashion followers I follow are very stylish	3.76	0.83	−1.054	2.049
Attractiveness Item 3	The fashion followers I follow are good looking	3.70	0.82	−0.891	1.633
Attractiveness Item 4	The fashion followers I follow are sexy	3.49	0.85	−0.516	0.788
Attractiveness Total		3.63	0.72	−0.906	2.631	0.89
Attitude toward influencer Item 1	I do believe that fashion influencers serve as fashion models for me	3.13	1.08	−0.351	−0.632
Attitude toward influencer Item 2	I do believe that fashion influencers present interesting content	3.65	0.88	−0.923	1.180
Attitude toward influencer Item 3	I do believe that fashion influencers provide new deals about different products and services	3.65	0.85	−1.010	1.487
Attitude toward influencer Item 4	I do consider fashion influencers as a reliable source of information and discovery	3.28	0.98	−0.445	−0.166
Attitude toward influencer Total		3.39	0.64	−0.580	1.760	0.84
Purchase intention Item 1	I most frequently have intentions to purchase products advertised by the fashion influencers I follow	3.05	1.10	−0.218	−0.821
Purchase intention Item 2	I generally recommend products and/or services advertised by the fashion influencers I follow	3.07	1.09	−0.242	−0.744
Purchase intention Item 3	I will buy the fashion item advertised by fashion influencers in the future	3.20	1.04	−0.402	−0.346
Purchase intention Total		3.10	0.98	−0.314	−0.454	0.90

M = mean; SD = standard deviation; Sk = skewness; Krt = kurtosis; α = Cronbach’s alpha.

shows descriptive statistics of items and scales. Skewness and kurtosis values show the normal distribution of variables. Concerning items, the likeability item 2 presents the highest mean, and similarity item 1 the lowest. Regarding scales, likeability total presents the highest mean value and similarity the lowest one. All Cronbach’s alpha values are above the recommended value.

4.3. Differences

Table 3. Differences by gender.

Variables	Gender	N	M	SD	t	df	p Value	d
Perceived credibility Total	Male	139	3.31	0.08	−2.679	200.844	0.008	0.75
	Female	362	3.54	0.04
Trustworthiness Total	Male	139	3.14	0.08	−2.529	499	0.012	0.85
	Female	362	3.35	0.04
Perceived expertise Total	Male	139	3.24	0.08	−0.573	499	0.567	0.83
	Female	362	3.29	0.04
Likeability Total	Male	139	3.54	0.07	−1.938	204.07	0.054	0.68
	Female	362	3.69	0.03
Similarity Total	Male	139	2.87	0.09	1.12	499	0.263	0.11
	Female	362	2.75	0.05
Familiarity Total	Male	139	3.26	0.09	−3.621	208.935	<0.001	0.89
	Female	362	3.62	0.04
Attractiveness Total	Male	139	3.55	0.07	−1.269	200.12	0.206	0.72
	Female	362	3.66	0.03
Attitude toward influencer Total	Male	139	3.29	0.07	−1.954	198.274	0.052	0.64
	Female	362	3.43	0.03

N = frequencies; M = mean; SD = standard deviation; t = t-test; df = degrees of freedom; p value = significance; d = Cohen’s d size effect.

presents the differences between genders concerning the total of the variables. Women have significantly higher values than men concerning perceived credibility, trustworthiness, and familiarity. There are no statistically significant differences between professionally active and inactive participants concerning the variables being studied; also, there are no statistically significant differences between different levels of education. Age correlates negatively and significantly with all the dimensions (perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, attractiveness, attitude toward influencers, and purchase intention); however, all the correlations are weak (r = −0.117 to r = −0.258). Moreover, the number of times that the sample connects to the social media correlates positively and significantly with all the dimensions between r = 0.154 (similarity) and r = 0.305 (familiarity).

4.4. Correlations

Pearson correlations show that all independent and dependent variables correlate positively, significantly, and substantially between them (

Table 4. Correlations between independent variables and dependent variables.

Variables	1	2	3	4	5	6	7	8	9
1 Perceived credibility	1
2 Trustworthiness	0.768 **	1
3 Perceived expertise	0.587 **	0.629 **	1
4 Likeability	0.657 **	0.690 **	0.627 **	1
5 Similarity	0.533 **	0.542 **	0.530 **	0.496 **	1
6 Familiarity	0.563 **	0.561 **	0.509 **	0.619 **	0.504 **	1
7 Attractiveness	0.503 **	0.450 **	0.461 **	0.587 **	0.386 **	0.546 **	1
8 Attitude toward the influencer	0.834 **	0.830 **	0.770 **	0.838 **	0.749 **	0.771 **	0.708 **	1
9 Purchase intention	0.663 **	0.653 **	0.596 **	0.557 **	0.594 **	0.543 **	0.438 **	0.733 **	1

** p < 0.001.

). The highest correlation occurs between attitude toward the influencer and likeability (r = 0.838), and the lowest occurs between attractiveness and similarity (r = 0.386).

4.5. Multiple Regression Analysis

Results in

Table 5. Multivariate regression analyses assessing the association of the independent variables and sociodemographic variables with purchase intention.

	Model 1		Model 2
	β	SE	β	SE
Perceived credibility	0.244 *	0.123	0.255 *	0.124
Trustworthiness	0.174 *	0.096	0.177 *	0.096
Perceived expertise	0.162 *	0.087	0.159 *	0.088
Likeability	−0.041	0.128	−0.042	0.129
Similarity	0.213 *	0.100	0.210 *	0.102
Familiarity	0.097	0.086	0.103	0.087
Attractiveness	0.018	0.115	0.018	0.116
Attitude toward influencers	0.048	0.629	0.039	0.636
Gender			−0.041	0.068
Age			−0.018	0.013
Education			−0.025	0.057
Occupation			0.009	0.094
F(df, p value)	81.954 (8, <0.001)		54.767 (12, <0.001)

* β = standardized regression coefficient; SE = standard error; * p < 0.05; F = F-value of ANOVA; df = degrees of freedom; p value = significance.

show that the multiple regression models fit the data well (Model I (not controlling for covariates): F = 81.95, p < 0.001, Model II (controlling for covariates): F = 54.77, p < 0.001), explaining, respectively, 56.4% and 56.3% of purchase intention’s variance. Standardized regression coefficients from Model II indicate that, among the eight independent variables, four (perceived credibility, trustworthiness, perceived expertise, and similarity) are significantly and positively associated with purchase intention.

4.6. Path Analysis

A path analysis was carried out to test the proposed mediation hypothesis, as depicted in Figure 1.

Figure 2 presents the result for the model with all linking paths. The data fit the model well (χ²(df) = 2.31; CFI = 0.96; TLI = 0.95; RMSEA = 0.05; SRMR = 0.04; PCLOSE = 0.34). Trustworthiness, perceived credibility, familiarity, attractiveness, and likeability are the dimensions that directly contribute the most to purchase intention. Trustworthiness and familiarity are the dimensions that directly contribute the most to attitude towards influencers. Among the seven components, the association between perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity (on the one hand) and purchase intention (on the other) is completely mediated through attitudes toward the influencer.

Table 6. Results of the mediation hypotheses (standardized regression weights).

Hypothesis	Paths	Statement of the Hypothesis	Indirect	Direct	Total	Mediation
H1	PC→AtI→PI	Attitude toward influencers mediates the relationship between perceived credibility and purchase intention.	0.119 *	0.062 (NS)	0.605 ***	Supported
H2	TW→AtI→PI	Attitude toward influencers mediates the relationship between trustworthiness and purchase intention.	0.415 ***	0.215 (NS)	0.100 *	Supported
H3	PE→AtI→PI	Attitude toward influencers mediates the relationship between perceived expertise and purchase intention.	0.318 ***	0.165 (NS)	0.219 **	Supported
H4	L→AtI→PI	Attitude toward influencers mediates the relationship between likeability and purchase intention.	0.248 **	0.128 (NS)	0.089 (NS)	Not supported
H5	S→AtI→PI	Attitude toward influencers mediates the relationship between similarity and purchase intention.	0.253 **	0.131 (NS)	0.198 *	Supported
H6	F→AtI→PI	Attitude toward influencers mediates the relationship between familiarity and purchase intention.	0.555 ***	0.288 **	0.127 *	Supported
H7	A→AtI→PI	Attitude toward influencers mediates the relationship between attractiveness and purchase intention.	0.331 ***	0.172 (NS)	0.027 (NS)	Not supported

*** p < 0.001; ** p < 0.01; * p < 0.05.

presents the results of the mediation hypothesis of the research model of this study. These results support five of the seven mediation hypotheses; hypotheses 4 and 7 were not supported.

5. Discussion

This study aimed to test the mediating role of the attitude toward influencers in the relation between, on one hand, perceived credibility, trustworthiness, perceived expertise, likeability, similarity, familiarity, and attractiveness, and, on the other hand, purchase intention. As such, correlations were established between the different constructs; all independent and dependent variables correlated positively and significantly between them. Overall, these results are in line with the literature on the subject [31], in addition to confirming other empirical evidence that points in the same direction [2,32,33,38].

Four independent dimensions (perceived credibility, trustworthiness, perceived expertise, and similarity) significantly contribute to explaining purchase intention. The perceived credibility of any given content is one of the most important factors for subscription and purchase intention [28]. The results are in line with empirical evidence suggesting a positive relationship between similarity and trustworthiness [34], which in turn is an important predictor of purchase intention [21,30,32,33,34,35]. According to the literature, the expertise of social media influencers is an important predictor of purchase intention [3,32,33]. Finally, there is a positive effect of similarity on users’ online purchase intention, according to Fu et al. [48].

Trustworthiness, perceived credibility, familiarity, attractiveness, and likeability are the dimensions that directly contribute the most to purchase intention. Trustworthiness and familiarity are the dimensions that directly contribute the most to the attitude towards influencers. Familiarity is a variable that positively influences, indirectly, purchase intention, being a result of a process of uncertainty reduction through the development of a trusting relationship [50]. Kim and Song [63] found indirect effects of the credibility of authenticity claims on purchase intention via competence and authenticity. The literature states that attractiveness is a determinant of purchase intention [31,57]. One tends to buy from people whom one likes [22]; the likeability of a specific brand has a positive influence on the purchase intention of its consumers [41].

Among the seven components, the association between perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity, on the one hand, and purchase intention, on the other, was completely mediated through attitudes toward the influencer. Our results suggest that perceived credibility and perceived expertise have an indirect influence on purchase intention when mediated by attitude toward influencers. More specifically, the greater the perceived expertise of a given influencer, the greater the positive response toward him/her by his/her followers. In turn, such positive responses have the positive power of mediating the relationship between perceived expertise and purchase intention. A link is hence seen between relevant knowledge and purchase intention, and “empty” marketing or simply appearing next to a brand name is not enough to persuade buyers—however well-known the influencer is, they will still have to understand the market and the product or service being advertised [2,25,26,27]. Consumers tend to seek information from other consumers or reference figures when they find themselves in a situation of uncertainty [23]. From this perspective, the indirect influence of perceived expertise on purchase intention, when mediated by the attitude toward influencers, may be related to a process of uncertainty reduction (note that some countries are more averse to uncertainty than others [64], so the results may vary from culture to culture; Portugal is one of the most uncertainty-averse countries in the world, immediately after Greece). In other words, our results suggest that the influencer is perceived as someone capable of reducing uncertainties concerning a product or service since he/she seems to possess useful knowledge capable of subsidizing the purchase decisions of the followers, thus increasing their purchase intentions. One sees here a similarity to business-to-business markets and industrial marketing, whereby industrial buyers rely heavily on the knowledge of experts before buying, which is understandable.

The perception that a particular digital influencer is somehow related to followers’ past experiences brings familiarity, similarity, and legitimacy to the influencer as someone worthy of credit, especially concerning the quality of the information provided. Such legitimacy and trust in the quality of the information provided will reduce the uncertainties perceived by a given follower during a purchase decision process, impacting positively on their purchase intention (many products appear to be similar in a supply-dominated world, so the role of influencers is seen to be increasingly important, to differentiate products amidst the “noise”). This result confirms other empirical evidence that points to the influence of familiarity on engagement with a brand on social media, which in turn can influence purchase intention [54], attitudes towards a brand, and purchase intention [45], as well as on endorsement by celebrities [55] (suggesting an inner desire among human beings for fame and fortune, despite the premature demise of many celebrities, much publicized in the media).

One would expect to find an indirect influence of attractiveness and likeability on purchase intention when mediated by attitude toward influencers. However, this research could not confirm these hypotheses. In fact, aspects related to perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity have a greater impact on purchase intention than attractiveness and likeability. This suggests that the more subjective aspects of this process (attractiveness and likeability) are less considered than the more objective ones (perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity).

Given the nature of this research, which has fashion influencers as its object of investigation, one argues that this result may be related to the influencers’ ability to present themselves as relevant in the medium in question. The perception that a certain influencer is relevant makes them a kind of central reference figure, which in turn can be used as a starting point to subsidize the decision-making process regarding a specific purchase or brand by a follower, in a world where time “is of the essence”.

6. Conclusions

Of the seven hypotheses established, five were empirically validated. From the analysis performed, it was possible to identify the characteristics of digital influencers, namely perceived credibility, trustworthiness, perceived expertise, similarity, and familiarity, that have an important indirect positive influence on purchase intention. This has important implications for want-to-be digital fashion influencers, as well as for the firms who may employ their services. On the one hand, these results are related to the process of uncertainty reduction (a role also played by brands, let one recall, which leads one to see that digital influencers are increasingly acting as brands themselves), through the development of a trusting relationship (much as one would entrust an important task to a close family member or friend), especially regarding perceived credibility and expertise (or knowledge), similarity (one likes to believe that one is similar often to very dissimilar people, so much credit to those who make one believe or perceive that they are similar), and familiarity (one might recall an event whereby one crossed paths with a very familiar face from TV only to be surprised that they did not say hello; this may be perceived as the ultimate triumph of an influencer—encouraging one to feel to be close friends with them, even though we have never actually met).

This study contributes to helping fashion industry marketers to better understand the influencer marketing impact on purchase intention and the factors that explain consumers’ attitudes toward influencers, consequently improving their choice of influencers that could generate purchase intentions among consumers (the battle for consumers’ attention has begun and will not let up any time soon—but will, actually, become more intense and so firms will need every advantage they can get). A limitation of the study is that the sample used herein was not entirely random and thus is not representative of the Portuguese population. In future studies, one suggests (a) the analysis of the use of social networks in the attitude towards digital influencers, as well as its consequent influence on the purchase intention; (b) the investigation of how followers’ characteristics, such as narcissism, desire for self-promotion, social comparison, and envy, influence attitudes toward influencers and their indirect impacts on the purchase intention mediated by the attitude toward influencers.