1 Introduction

In the digital era, marketers increasingly faced the problem that consumers regarded traditional forms of advertising (e.g., banner ads) as disturbing and untrustworthy information sources for their purchasing decisions (Wenzel, 2016; Chen & Xie, 2008). This ultimately led to the re-occurrence and rise of influencer marketing, which can be basically regarded as a type of word-of-mouth (WOM) on online platforms. More specifically, it is a special form of endorsement marketing which uses product or brand recommendations from ‘influencers’ to improve the effectiveness of consumer persuasion and to ultimately drive sales. However, influencer marketing typically has much broader goals such as increasing attention of potential customers, triggering positive WOM effects, and generating consumer engagement (e.g., Brown & Hayes 2008; Masuda et al., 2022). A (social) influencer is an opinion leader who has the power and the will to affect consumers’ decision-making (e.g., product choice) via social media channels such as YouTube, TikTok, and Instagram (Lin et al., 2018). Some influencers have built up communities with an enormous number of follows. For instance, Kim Kardashian has more than 300 Mio. followers (Kim Kardashian, n.d.). In the recent past, marketers tried to benefit from the influencers’ reach and consumers’ inclination to value third-party sources more than commercial information when planning a purchase. In 2015, German consumers, for instance, particularly trusted personal recommendations (78%), but were also very much inclined to believe other persons’ opinions online (e.g., from influencers) (62%) (Nielsen, 2015). Some years ago, almost half of the Germans could imagine buying a product advocated by influencers (PwC, 2019). In these old days of influencer marketing, followers typically valued the product endorsements of the online opinion leaders. This was an international phenomenon. For example, in the US 63% trusted influencers more than brands (Edelman, 2019). They often regarded these messages as less biased and more informative compared to many other sources. In this golden era companies had a strong drive to use the viral growth potential of influencers to promote their brand’s personality and to create a community of loyal customers.

However, in the meantime, influencer marketing had suffered from some considerable drawbacks as it is currently struggling with sincerity issues: Consumers increasingly doubt the authenticity of influencers, who are more and more arbitrarily cooperating with diverse companies and multiple brands at the same time. These brands often do not appear to correspond to an influencer’s interests or personality (Audrezet et al., 2020). Every second German consumer who follows influencers receives sponsored messages from his/her beloved opinion leader, which increasingly erodes consumers’ faith in influencers’ independence and objectivity (Wavemaker, 2019). These days, less than 37% of German consumers trust influencer advertising (Nielsen, 2022). Currently, only 13% of US consumers turn to influencers to inform their purchasing decisions (Oracle & Brent Leary, 2022). Considerable anecdotal and empirical evidence from international sources suggests that trust in influencers is declining (e.g., CXM, 2022; Entrepreneur, 2021; Forbes, 2020). Consumers have adjusted their media knowledge, which led to an increased general skepticism towards influencers and their motivations. Therefore, the sustainable success of influencer marketing is currently at its crossroads – leading to controversial discussions about the future of influencer marketing among practitioners and scholars alike (e.g., Irish Times, 2019).

It seems that the concepts of influencer credibility and authenticity are the future key constructs for the success of influencer marketing. This implies that for marketers, who value the diverse benefits of influencer marketing for their brands, such as increased purchase likelihood and better reputation among their potential and future customers (e.g., Bu et al., 2022; Farivar & Wang, 2022), it will become increasingly important to select the right influencers based on their sincerity (Tabor, 2020). However, while marketing research has a profound understanding of the credibility concept (e.g., Sokolova & Kefi 2020; Lindh & Lisichkova, 2017; Hu et al., 2019), an appropriate discussion of the nature, boundaries, and role of authenticity – in the context of influencer marketing – is still sparse. This is surprising as marketing literature has recognized the importance of authenticity across different market offerings, like historical sites (Goulding, 2000; Grayson and Martinec, 2004), (reality) television (Becker et al., 2019; Rose & Wood, 2005), and brands (Fritz et al., 2017; Beverland et al., 2008; Lee et al., 2009). Only a few marketing studies address the perceived authenticity of individuals (e.g., Moulard et al., 2015). Nevertheless, our knowledge about influencer authenticity and its relationship to key brand-related outcomes remains extremely limited.

In this research, we shed light on the importance of an influencer’s perceived authenticity. By doing so, we have conducted two experiments, which investigated four possible perceptual constructs acting as the concept’s consumer-sided antecedents: (i) influencer uniqueness (i.e., the perceived originality of an influencer in terms of his/her independence, creativity, and individuality), (ii) scarcity (i.e., the perceived popularity of an influencer in terms of his/her number of followers), (iii) longevity (i.e., the perceived persistence of the influencer’s activities), and (iv) consistency (i.e., the perceived perseverance of an influencer towards change). The findings of our studies suggest that an influencer’s authenticity is mainly dependent on both his/her uniqueness and consistency. This means that consumers particularly value the iconic status of influencers who distinguish themselves from others and ‘stand out’ from the usual. In addition, our empirical findings indicate that consumers have a slight tendency to especially find influencers as authentic, who are both persistent and consistent in their actions. This interplay possibly has, however, only a weak systematic influence. Furthermore, this research can show that influencer authenticity is a trigger construct that can ultimately affect consumers’ purchase intention towards a specific brand, reviewed and recommended by the influencer in his/her sponsored posts. We demonstrate that a serial mediation process involving several brand-related constructs (e.g., brand authenticity, brand attitude) is accountable for this reaction. Overall, the empirical findings support our anecdotal claim that marketers should pay increased attention to the choice of influencers by considering their purchase-relevant authenticity. And more specifically, their uniqueness (i.e., perceived individualism) as well as consistency (i.e., truth to their selves). This consideration pays off by positively affecting consumers’ willingness to buy the advertised brand.

2 Conceptual background and hypotheses

2.1 Influencers

Scholars often define influencers as individuals who interact with a specific target audience in which they regularly stimulate online engagement (e.g., discussions) and to whom they sell a product, service, or brand. Influencers include persons with different backgrounds: some of them are celebrities, while others are industry advocates, professionals, or non-professional peers (Childers et al., 2019). The term also includes individuals who make notable contributions on social media and gained considerable recognition from others. One could recognize these individuals also can be identified by the fact that they have created a dedicated audience and a large online following (Bu et al., 2022; De Veirman et al., 2017). Influencers are often regarded as ‘opinion leaders’ who have a strong public reputation. They communicate with their followers in various ways including posts, photos, videos, and other social activities (e.g., online meetings). These activities help them to develop their own online persona (Tafesse & Wien, 2018) or their personal brands (i.e., ‘human brands’; Kay et al., 2020) and to build trust by demonstrating their knowledge in different domains such as fashion, lifestyle, sports, and cooking. In this article, we simply refer to these individuals as ‘influencers’, but in literature, they are also often described as ‘social influencers’, ‘internet celebrities’, ‘digital opinion leaders’, or ‘market mavens’.

Recently, the background conditions explaining consumers’ positive reactions to sponsored posts or social media advertising by influencers received considerable attention among marketing scholars and practitioners alike. Extant literature often emphasizes that this kind of communication is perceived as more credible, seems more organic, and has a greater potential to reach the desired target audiences as compared to online advertising by companies (Lou & Yuan, 2019). However, various factors contribute to this mechanism. For instance, interactions between the influencer and consumers appear more credible than traditional advertising as the persuasive messages are seamlessly woven into the daily narratives posted by the influencer (Abidin, 2016). Research demonstrates that besides message characteristics (e.g., content length, uniqueness, originality, interactivity, informativeness, recommendation type; Ki et al., 2020; Woodroof et al., 2020), follower attributes as well as perceived influencer characteristics are essential. Concerning the former, it has been found that interest fit (Belanche et al., 2021; Martínez-López et al., 2020) as well as followers’ loneliness (Hwang & Zhang, 2018) and self-esteem play a critical role in their responses to influencer marketing. In addition, parasocial relationships between the follower and the influencer can lead to an imitation of the influencer’s habits by means of consumption (Ki & Kim, 2019; Sokolova & Kefi, 2020). Followers have the illusion of a face-to-face relationship with influencers, which makes them more susceptible to their opinions and recommendations (Colliander & Dahlén, 2011). When it comes to influencer characteristics, the perceived credibility (Breves et al., 2019; Reinikainen et al., 2020), trustworthiness (Schouten et al., 2020), as well as the physical attractiveness (Sokolova & Kefi, 2020) are often highlighted as important determinants of influencers’ impact on buying decisions. The discussion on essential influencer characteristics is currently stimulated by the introduction of congruency-based factors such as influencer-brand fit (Breves et al., 2019; Kim & Kim, 2022) and – more importantly – influencer authenticity.

2.2 Authenticity concept

While in marketing at least some agreement exists about the meaning of authenticity, no universally accepted definition is yet available. Various authors agree that the concept is the outcome of a verification process of the truth or a fact (e.g., Newman & Dhar, 2014; Eigenraam et al., 2021; Kumar & Kaushal, 2021). Beverland and Farrely (2010, p. 839), for instance, state „despite the multiplicity of terms and interpretations applied to authenticity, ultimately what is consistent across the literature is that authenticity encapsulates what is genuine, real, and/or true”. While this phrase postulates that being ‘authentic’ means that someone or something is ‘real’, these words can mean different things to different persons in different contexts. Authenticity is therefore often described as a socially constructed interpretation of an observation (Beverland & Farrelly, 2010; Grayson & Martinec, 2004). As such, individuals (e.g., consumers) should always be considered as co-creators of authenticity because they interact with the authentic object or person (Rose & Wood, 2005). They ultimately ‘define’ authenticity. While some consumers may perceive an object, person, experience, or brand as genuine and real, others may regard the same evaluation object as fake. Consistently, various scholars theorize that the meaning of authenticity is goal dependent. This means that personal objectives (i.e., making a well-informed purchasing decision while being exposed to potentially biased market communication and communicators with a particular self-interest) determine which characteristics contribute to a perceived authenticity.

In line with this argumentation, Grayson and Martinec (2004) identify two types of authenticity: ‘indexical authenticity’ and ‘iconic authenticity’. Concerning the former, authenticity can be used to describe something that is thought not to be a copy or an imitation. Here, an object is perceived to be authentic when it is believed to be ‘the original’. Similarly, an individual’s habits are authentic in case they are believed to reflect who the person really is. In contrast, a person would appear unauthentic if his/her actions are thought to solely meet social norms or to make money (e.g., by recommending a product or brand only because a company sponsors that post). Alternatively, iconic authenticity means that sometimes the word ‘authentic’ is used to evaluate whether an object’s physical appearance resembles something stereotypical. For example, silver pieces in a museum gift shop are authentic if they are thought to look like coins made by Spanish colonies in the sixteenth century. In line with this perspective, for the remainder of this article, we regard a person’s authenticity as a form of what Grayson and Martinec (2004) describe as ‘indexical authenticity’ as individuals primarily subjectively judge another person’s (e.g., an influencer’s) genuineness based on the extent to which the actions mirror the other party’s real self.

2.3 Influencer authenticity

2.3.1 Perceived authenticity of individuals

Both self-determination theory (Ryan & Daci, 2000) as well as attribution theory (Asch, 1946) help to explain the nature and role of the perceived authenticity of individuals (e.g., Kowalczyk & Pounders, 2016; Moulard et al., 2015). Here, various authors agree that in this context, authenticity can be described as the extent to which one remains true to oneself. People are regarded as authentic if they are genuine, original, and unique. They act in accordance with their inner believes, thoughts, feelings, and passion. The actions of an authentic person reflect his/her unbiased self (Ilicic & Webster, 2016; Moulard et al., 2016; Moulard et al., 2015). These actions can be understood as intrinsically motivated behaviors, which are triggered without any bias like commercial or opportunistic goals. According to self-determination theory, such behaviors are free of external influences, and they align with the personal concept of oneself. Intrinsic motivation originates from three native basic human needs: self-competence, self-determination, and social affiliation (Rohlfs, 2011). Authentic individuals strive for these needs and are not motivated by potential rewards nor punishment (Ryan & Deci, 2000).

As the true self and the inner motivation are not obvious to others, another person’s authenticity can never be evaluated with absolute certainty. Nonetheless, individuals do make assessments of another’s authenticity. The perception of another person’s (e.g., influencer’s) authenticity, despite whether the person is objectively authentic, is the construct of interest in this research. Individuals (e.g., consumers) typically try to infer another human’s motivation from the observable behaviors. This corresponds with the basic tenet of attribution theory (Asch, 1946), which describes individuals’ inner urge to understand their environment and other persons’ actions by making causal inferences. When making observations, people always try to discover connections and try to identify the most plausible reason for others’ behaviors (Schwaiger & Meyer, 2011). Here, the theory claims that individuals typically make inferences whether the behavior is intrinsically or extrinsically motivated. A person is judged as being authentic when an intrinsic motivation is perceived to steer the observable behavior. A behavior is regarded as being intrinsically motivated if it is unique, and it is consistent across different situations and different stimuli (Kowalczyk & Pounders, 2016). For instance, a person is evaluated as acting intrinsically motivated in case this person is much more friendly and sociable than other individuals (i.e., uniqueness), the person remains friendly despite being in a good or bad mood (i.e., consistency across different situations), and the person is friendly to everyone that he/she meets (i.e., consistency across different stimuli). Having this said, depending on the context, authenticity perceptions can vary concerning their most relevant components, which can include individuality (Wentzel, 2009), empathy (MacInnis et al., 2002), and naturalness (Zhu & Meyers-Levy, 2009). For example, Jones and her colleagues (2022) conceptualize a chatbot’s authenticity in terms of its appearance as a real individual with the capability to feel and experience empathy while communicating in a natural manner.

A considerable acceleration of the scientific discussion of the authenticity concept came from the intensified research in consumer research (Rose & Wood, 2005). Since then, academic studies gained momentum leading to further insights concerning the role and nature of this concept (Beverland & Farrelly, 2010; Napoli et al., 2014; Kumar & Kaushal, 2021). In the branding context, scholars agree that the meaning of authenticity revolves around the extent to which consumers perceive brands (i.e., both human brands such as celebrities or artists and product or service brands) as intrinsically motivated. Here, a brand is perceived to be authentic if it seems to be in business because it is enjoyable or provides hedonic value and not to increase profits or purely prestige. In the consumption context, authenticity is a consumer’s subjective judgment of genuineness attributed to a (market) object (Napoli & Kaushal, 2014; Davis et al., 2019). Consumers evaluate the authenticity of products based on their sincerity, innocence, or originality (Fine, 2003), or the judgment can be related to perceptions of the reality or truth of it (Grayson & Martinec, 2004). A product can be authentic when it is perceived as being true to its heritage, using traditional modes of production (Beverland, 2005). Switching to social entities, Ilicic and Webster (2016) as well as Kowalczk and Pounders (2016) define ‘celebrity authenticity’ as the consumers’ perception of the extent to which well-known human brands remain true to their true self. An authentic celebrity must act based on his/her personal values. Moulard et al. (2014) also conceptualized ‘artist authenticity’ in a similar manner. In the work of Cuesta-Valino and his colleagues (2022), a product manager’s authenticity is defined as the extent to which consumers evaluate this person as intrinsically motivated and to which degree, he/she is passionate and devoted to his/her products.

Hence, based on the above discussion, one can conclude that consumers regularly infer the degree of perceived authenticity of various market participants including objects (e.g., stores, brands) and subjects (e.g., managers, advertisers, testimonials, influencers). Based on the thing being evaluated, the factors used to evaluate can strongly vary. However, marketing scholars widely agree that authenticity is an impactful concept that can affect consumers’ feelings, opinions, evaluations, and has the potential to dictate actions (e.g., brand purchases) (e.g., Morhart et al., 2015; Spiggle et al., 2012; Davis et al., 2019).

2.3.2 Perceived authenticity of influencers

Based on the dominant understanding of authenticity in marketing, this research conceptualizes influencer authenticity as the extent to which consumers perceive a social media influencer as behaving in accordance with his/her true self. Authentic influencers are regarded as acting in correspondence with their personal characteristics, thoughts, feelings, and passions because of a strong intrinsic motivation (Ilicic & Webster, 2016; Moulard et al., 2016; Moulard et al., 2015). They produce content (e.g., posts) that is seen as being dominantly based on their dedication to a topic, product, or brand. In line with this perspective, Kapitan et al., (2021) define influencers’ authenticity as the consumer’s perception of acting in accordance with one’s values, preferences, and needs versus acting in such a way to please others or obtain rewards. Satisfying the audience (i.e., followers) or involved brands as well as any monetary incentives have minor importance to the authentic influencer. Their activities revolve around self-expression and independence. For authentic influencers, producing social media content is not a means to an end (e.g., to please followers or marketers), but more importantly the possibility to create a real added value for their audiences. In line with this reasoning, Audrezet et al. (2020) state that authentic influencers provide fact-based information about companies and their products. Furthermore, they argue that ‘transparent authenticity’ means that the influencer discloses information about the contractual terms of the partnership with a particular brand as well as posting unedited content. Authentic influencers are perceived to derive gratification from self-improvement, enjoyment, pleasure, and positive emotions. Consumers interpret influencers’ positive/negative statements about products or brands as recommendations based on their true and unbiased beliefs about market offerings. That is, only if they provide fair, intrinsically motivated reviews mirroring his/her own personal experiences. Adding to this understanding, Lee and Eastin (2021) describe social media influencers’ authenticity as the extent to which consumers perceive influencers to be kindhearted, engage in intrinsically motivated brand endorsements, reveal personal information about themselves, are naturally talented in their area of expertise, and are distinct from other influencers. Based on above discussion, one can say that overall, influencer authenticity is a construct that comprises a consumer’s evaluation of the extent to which a social media producer genuinely communicates purchasing-relevant information which resembles his/her true stance towards the presented and discussed products or brands.

Influencer authenticity must be conceptually separated from ‘influencer credibility’. The latter is a well-established multi-dimensional construct, which was first conceptualized – as a general consumer research construct – by Hovland and Weiss (1951). According to these scholars, individuals evaluate the credibility of a source based on (1) its perceived expertise (i.e., the extent to which the source is regarded as knowledgeable) and (2) its perceived trustworthiness (i.e., the extent to which the source is assumed to be honest and caring for the welfare of his/her audience). While in the context of social media endorsers some scholars focus only on one dimension (e.g., Chung & Cho, 2017), it is however widely agreed that at least these two dimensions should be regarded to fully capture an influencer’s perceived credibility (e.g., Sokolova & Kefi 2020; Lindh & Lisichkova, 2017; Hu et al., 2019). Sometimes, in the context of social influencers, other scholars regard the perceived attractiveness as another component of source credibility or delimited construct (e.g., Torres et al., 2019). Here, the concept regularly not only encompasses the physical appearance, but also other factors such as sympathy and familiarity. These are components, which are often not used to conceptualize authenticity or represent only parts of its nature. The argument that influencer authenticity and above-mentioned constructs are related, but conceptually disjunct constructs is supported by the fact that marketing literature widely agrees that these constructs have different antecedents (e.g., Moulard et al., 2015), while affecting similar brand-related concepts.

2.4 Conceptual model

Given our understanding of the nature of influencer authenticity, we conceptualized a stimulus-organism-response-model (see Fig. 1), which specifies the construct’s important consumer-sided perceptual antecedents and its impact on consumers’ purchase intention - a critical brand-related behavioral variable. Based on attribution theory (Asch, 1946) and the conceptualization of celebrity authenticity we define (1) uniqueness, (2) stability, (3) longevity, and (4) consistency as essential factors determining influencer authenticity (Moulard et al., 2016; Moulard et al., 2015). In contrast to recent works, which conceptualize influencer authenticity as strongly dependent on the style of communication an influencer endorses a brand with (Audrezet et al., 2020), we define influencer authenticity as a personality characteristic or general nature of an influencer. This nature is naturally mirrored by the perceivable characteristics of his/her social media account and the associated information on his/her person (identifiable in the profile description) and his/her popularity (described by the number of followers or likes) (Lindmoser et al., 2022). Hence, we do not focus on self-determination theory and on whether the influencer perceives himself/herself as authentic. In contrast, this research’s conceptualization involves whether an individual, specifically another consumer, perceives an influencer or the available information on the influencer (i.e., someone other than himself/herself) as intrinsically driven and therefore authentic. By doing so, we borrow from typical conceptualizations of previous works on artist authenticity, celebrity authenticity, and brand authenticity (e.g., Moulard et al., 2014, Moulard et al., 2016; Moulard et al., 2015) (see above). Furthermore, our research follows the basic definition of the ‘Entity-Referent Correspondence’ (ERC) framework of authenticity which defines the construct as “a consumer’s perception of the extent to which an entity corresponds to a referent” (Moulard et al., 2021, p. 99) and can be categorized according to the framework as research on ‘true-to-self authenticity’.

Fig. 1
figure 1

Conceptual model

Full size image

2.4.1 Perceived rarity as antecedent of influencer authenticity

According to the ERC framework, an authentic entity is strongly defined by cues of uniqueness and scarcity (Moulard et al., 2021). These variables are two subdomains of the concept of rarity which has been already investigated in the context of celebrity authenticity and brand authenticity (Audrezet et al., 2020; Moulard et al., 2016). Rarity can be described as the degree to which a person or brand is seen as uncommon. Society typically makes it more difficult to act against social conformity and withholds social approval (Jones & Davis, 1965; Kelley, 1973; Bellezza et al., 2014). A person or brand which ’goes against the grain’ by not conforming to social norms is perceived as intrinsically motivated and is therefore experienced as more authentic. Uniqueness, in this context, describes the perceived originality of a person. It is defined by the independence, creativity, and individuality of the person (Moulard et al., 2016). Anyone who is not considered to be ’typical’ because of, for example, his/her age, occupation, social status and/or political beliefs is evaluated as more authentic (Fine, 2003; Peterson, 2005). Influencers reveal constantly personal information through their profile description and messages on social media outlets. We argue, in line with attribution theory, that the more this information distinguishes the influencer from his/her peers, the more it is interpreted as a sign of uniqueness and signals an intrinsic motivated person. Therefore, we hypothesize:

H1. Influencer uniqueness has a positive effect on influencer authenticity.

Scarcity, the second subdomain of rarity, is defined as the extent to which consumers perceive that another entity (a person or a brand) is not widely available or accessible (Moulard et al., 2016). A notion supported by the branding literature where ubiquity and over-commercialization as signs of aggressive growth decrease authenticity (Gilmore & Pine, 2007). The limited accessibility gives consumers the impression that an entity takes pride in and is committed to its craft and has been investigated and reported as a sign of authenticity of brands, products (Beverland, 2005) and persons (Moulard et al., 2015). The effect of scarcity is determined by attribution theory. An entity which “does not search the spotlight” and limits its accessibility is interpreted as more intrinsically motivated. In the context of social media, a common indicator of accessibility and popularity is the number of followers an account or influencer has. In recent years, several studies investigated the effect of follower numbers on different aspects of consumer behavior in social media environments (Britt et al., 2020; Kay et al., 2020; Park et al., 2021). Results show that influencers with a low number of followers have more two-way-interactions with their followers (Britt et al., 2020; Campbell & Farrell, 2020), create more product knowledge through their communication (Kay et al., 2020) and are perceived to have fewer ulterior commercial intentions compared to influencers with higher follower numbers (Audrezet et al., 2020; Campbell & Farrell, 2020; Kay et al., 2020). A plausible theoretical explanation for these results can be drawn from the Persuasion Knowledge Model, which implies that consumers try to resist or ignore marketing and advertising content which attempts to be more persuasive (Friestad & Wright, 1994). Consumers thus perceive influencers with a high number of followers as more commercially driven and manipulative, and subsequently they resist these efforts and evaluate the influencer as more inauthentic. In this context, Britt et al. (2020) found no substantial differences in affective, social, and cognitive content communicated by influencers with high or low numbers of followers. Therefore, we argue that it is not the content communicated but the number of followers itself causing an effect of scarcity. Hence, we hypothesize:

H2. Influencer scarcity has a positive effect on influencer authenticity.

2.4.2 Perceived stability as antecedent of influencer authenticity

A second factor proposed to positively influence ‘true-to-self authenticity’ is the concept of stability (Moulard et al., 2020). According to attribution theory brands and persons perceived as unwavering are interpreted as authentic because the demonstrated stable behavior is attributed to an intrinsic motivation and not interpreted as caused by external pressures (Kelley, 1973). Stability comprises of two sub-domains: (c) longevity and (d) longitudinal consistency. Longevity represents the persistence and therefore the length of existence of an entity (Moulard et al., 2020; Moulard et al., 2016). In the context of social media, we argue that influencers who already communicate constantly over a longer period of time are perceived as more authentic than influencers who just started their influencer career. Accordingly, we hypothesize:

H3. Influencer longevity has a positive effect on influencer authenticity.

Longitudinal consistency describes the perseverance of an entity towards change over time (Moulard et al., 2020). Brands that exhibit a consistent image, even though specific aspects and attributes (e.g., ingredients, products) have changed over time are considered to be authentic (Moulard et al., 2016). Similarly, the degree of change of an influencer’s personality and characteristics over time influence how authentic the person will be perceived as his/her consistency (Moulard et al., 2015). According to self-determination theory a person considers himself/herself as more authentic if his/her self-concepts are consistent across time and contexts (Diehl et al., 2006). A logic also applied when the same person is evaluating others. Therefore, we propose that an influencer who does not change in personality and his/her associated image appears to be more intrinsically motivated – and is therefore perceived to be more authentic. Consequently, we hypothesize:

H4. Influencer consistency has a positive effect on influencer authenticity.

2.4.3 Effect of influencer authenticity on brand authenticity

Several studies have investigated applicability of the authenticity concept to the branding context (e.g., Dwivedi & McDonald, 2018; Moulard et al., 2015; Moulard et al., 2016). They agree that brand authenticity offers consumers several advantages such as the ability to offer a sense of continuity in ever-changing, complex environments characterized by uncertainty (Fritz et al., 2017), cultivating the credibility of brands (Schallehn et al., 2014; Becker et al., 2019), enabling a unique brand positioning (Lu et al., 2015), increasing customer satisfaction (Bruhn et al., 2012), or making consumers more receptive to brand messages (Audrezet et al., 2020). Brand authenticity describes the extent to which a specific brand appears to be genuine, real, and true. (Beverland & Farrelly, 2010). An authentic brand conforms to consumers’ mental frames how it ought to be and has proven that it has a fit with consumers’ expectations (Beverland et al., 2008; Cinelli & LeBoeuf, 2020; Eigenraam et al., 2021).

Influencer authenticity is strongly related to brand authenticity due to a naturally occurring image transfer as proposed by the meaning-transfer-model (McCracken, 1989). According to this model, in a so-called ‘rub-off effect’ (McCracken, 1989; Silvera & Austad, 2004) characteristics and meanings get transferred from a celebrity testimonial onto advertised products or services. In this context, the findings of Chung and Cho (2017) suggest that the trustworthiness of a celebrity has a positive influence on the credibility of the advertised brand. Silvera and Austad (2004) illustrate that ‘stylish’ celebrities make consumers think that the products endorsed by these celebrities are also ‘stylish’. Finally, Park et al. (2021) demonstrate that perceptions of influencer authenticity can be transferred to brands via endorsements from these influencers., Therefore, we hypothesize:

H5. Influencer authenticity has a positive effect on brand authenticity.

2.4.4 Effect of brand authenticity on brand attitude

Attitudes are a cornerstone of modern persuasion theories and are relatively enduring and general evaluations of an object, person, group, issue, or concept (Petty & Cacioppo, 1996). Ewing et al. (2012) demonstrate the impact of brand authenticity on brand attitude based on the Peirican approach (Grayson & Shulman, 2000; Grayson & Martinec, 2004; Beverland et al., 2008; Beverland & Farrelly, 2010). According to this approach drawn from semiotics, a product’s appearance and features define what it is supposed to be or represent but a consumer may or may not accept it. Perceived meaning of an object based on face value is subject to consumer judgement of its validity (Sommers & Kernan, 1967; Kleine & Kernan, 1991). Hence, perceived brand authenticity as a genuine sign of validity positively effects the general evaluation of a brand. The relationship between authenticity and attitude has been investigated in different contexts in previous studies. Brand authenticity positively influenced consumers attitudes towards green products in the USA (Ewing et al., 2012), private-label brands in France (Carsana & Jolibert, 2018), global brands in Pakistan and China (Safeer et al., 2021) and stores in Spain (Cuesta-Valino et al., 2022). Thus, we hypothesize:

H6. Brand authenticity has a positive effect on brand attitude.

2.4.5 Effect of brand attitude on purchasing intention

Finally, based on the classical interpretation of the theory of planned behavior (Ajzen, 1985) and the theory of reasoned action (Fishbein, 1979) we propose that intentions to perform behaviors can be predicted from attitudes toward the behavior. Therefore, we propose a positive effect of a positive brand attitude on behavioral outcomes like the intention to purchase the brand. Correspondingly we hypothesize:

H7. Brand attitude has a positive effect on purchasing intention.

3 Empirical study

3.1 Data collection

3.1.1 Research design, procedure, and participants

To investigate the proposed hypotheses, we exposed our participants to two online experimental conditions. In each experimental condition one subdimension of rarity and one subdimension of stability was manipulated. More precisely, experiment 1 can be described as a 2 (uniqueness high vs. uniqueness low) by 2 (longevity high vs. longevity low) design. The second experiment can be classified as a 2 (scarcity high vs. scarcity low) by 2 (consistency high vs. consistency low) design. Each condition in each experiment consisted of a scenario description and two pictorial stimuli in the form of an Instagram post and Instagram profile description.

A total of 163 subjects participated in both experiments that were conducted online. Students at a German speaking university (70.6% female; average age of 23.2 years; 60.7% of participants possessed a high school degree, 39.3% possessed a bachelor´s or master´s degree) were invited to participate in the experiment. A link to the experiment was provided in class or via social media. We controlled the environment of participants during the experiment. The experiments took place in a quiet environment and subjects were not allowed to interfere with other persons during data collection. All subjects participated via smart phone or laptop. The participants were randomly assigned to the conditions of the experiments. Gender, age and education showed the same distribution in all conditions of the experiment.

The online questionnaire begins with an introduction in which participants are instructed that the study is about evaluating an influencer on Instagram. Instagram was chosen as the platform of interest because at the time of the investigation it was the main social media platform for people aged 16 to 35 years in the country the university is situated in. Subsequently, the stimulus ad was presented, followed by the manipulation assessment measurements, dependent variables, and demographic variables.

In experiment 1 we exposed our participants to a picture of an Instagram posting of a chronologically old female influencer (uniqueness high condition) and a chronologically young female influencer (uniqueness low condition). We have chosen chronological age as a signal for uniqueness because influencer marketing is a profession predominantly practiced by a younger population group. In 2019, most influencers on Instagram were 18 to 34 years old (85%), only 12% of global influencers were aged 35 years and older and only 1% of influencers were more than 50 years old (Klear Research, 2019). Both influencers promoted a food delivery service called “Tasty Box”. The post was marked as “sponsored” to signify transparently that the influencer endorses a product in exchange for money. The pictorial cue was accompanied by a profile picture and scenario description indicating that the chronologically older food influencer is 62 years old and a social media pro (uniqueness high condition). The younger food influencer was described as a typical influencer and accompanied by a profile picture showing a young adult (uniqueness low condition). To manipulate the longevity associated with the influencer we provided a picture of the Instagram profile indicating the number of posts the influencer had communicated. A high number of 4,860 posts signified a situation of high longevity, a low number of 20 posts low longevity. Additionally, we provided the number of posts also in the scenario description and added also information on how long the influencer had been active on social media. In the low longevity condition, we indicated a duration of 11 months. In the high longevity condition, we indicated that the influencer is active for more than 6 years.

In experiment 2 we confronted our participants with a picture of an Instagram posting and an Instagram profile description of a male influencer. The influencer promoted a fictious camera called “Camy”. The post was marked as “sponsored” to signify transparently that the influencer endorses a product in exchange for money. Consistency was manipulated by providing information on the vita of the influencer. In the high consistency condition the male influencer was described as photography aficionado from his early days until today. In the low consistency condition the influencer was described as a person who recently developed a passion for photography but had many different interests and activities before that. Scarcity was manipulated by providing the number of followers the influencer had. Low scarcity was described with an account with 400 followers. High scarcity was achieved by a stated follower number of one million. Indicators for low and high scarcity where also included in the scenario descriptions.

To ensure a basic cognitive processing of our manipulations every participant was asked several questions after exposure to the stimuli indicating if he/she has correctly perceived and understood the specific scenario conditions. Participants who provided a wrong answer were given the possibility to review the stimuli material and revise their answers afterwards. Only participants indicating the correct answers were able to proceed with the questionnaire. All presented brand names and influencer profiles were invented for the purpose of the experiment and participants who indicated any a priori knowledge of the fictious brands or influencers were excluded from the experiments.

Manipulations in experiment 1 and 2 were tested with items derived from Moulard et al. (2016). In experiment 1 uniqueness was measured on four items, longevity was assessed using two items. In experiment 2 we used three items to rate consistency and three items to evaluate scarcity. All items were evaluated on a five-point scale (ranging from “strongly agree” to “strongly disagree”). We conducted an ANOVA with two factors for experiment 1 (uniqueness and longevity) and experiment 2 (consistency and scarcity). The results for experiment 1 indicate that stimuli material describing the chronological old influencer (uniqueness high) was perceived as more unique than material describing the younger influencer (old [M = 3.49] vs. young [M = 1.88]; F(1,159) = 145.38; p < 0.001). Furthermore stimuli material indicating a long activity on social media (longevity high) scored significantly higher on longevity than stimuli material indicating a short activity (long [M = 4.16] vs. short [M = 1.81]; F(1,159) = 374.21; p < 0.001). The results for experiment 2 indicate that the presentation of an unambiguous vita (consistency high) led to a significantly higher score in consistency than an ambiguous one (unambiguous [M = 3.92] vs. ambiguous [M = 2.50]; F(1,159) = 113.28; p < 0.001). Furthermore, the indication of a low number of followers (scarcity high) led to significant higher ratings in scarcity compared to a high number of followers (high number: [M = 2.29] vs. low number [M = 4.45]; F(1,159) = 320.105; p < 0.01). No other effects (main or interaction were significant). Hence, we assume that our manipulations were successful.

Dependent variables measured were influencer authenticity, brand authenticity, brand attitude, and purchase intention. All measures were derived from published literature (for an overview of multi-item measures and associated literature see Table 1). Participants reported influencer authenticity on six-item five-point scales and brand authenticity on eight-item five point scales. To assess brand attitude, we used four-item five point scales. All scales ranged from “strongly agree” to “strongly disagree”. Purchase intention was measured using a single item approach based on Mitchell and Olsen (1981). We asked respondents how probable they are going to buy the promoted product. Answers were indicated on a five-point scale ranging from “not at all probable” to “very probable”.

Table 1 Construct measurement (Experiment 1 and 2)
Full size table

3.1.2 Measurement

Two confirmatory factor analyses (CFA) were conducted for testing the psychometric properties of the main latent construct measures: one for the measures included in experiment 1 and another one evaluating the appropriateness of measurement within experiment 2. The measurement models fitted the data well (experiment 1: χ² = 253.462, df = 124, χ²/df = 2.04, RMSEA = 0.08, CFI = 0.93, SRMR = 0.06; experiment 2: χ² =234.888, df = 116, χ²/df = 2.03, RMSEA = 0.08, CFI = 0.95, SRMR = 0.06). Further, the construct validity as well as the reliability were ensured as indicated by (a) high Cronbach’s alpha coefficients (experiment 1: ranging from 0.87 to 0.91; experiment 2: ranging from 0.91 to 0.93), (b) satisfactory indicator reliabilities (experiment 1: ranging from 0.42 to 0.77; experiment 2: ranging from 0.45 to 0.89) and standardized item-to-construct loadings (experiment 1: ranging from 0.65 to 0.88; experiment 2: ranging from 0.67 to 0.94), and (c) composite reliabilities (experiment 1: ranging from 0.86 to 0.91; experiment 2: ranging from 0.90 to 0.93) and average variance extracted (AVE) values (experiment 1: ranging from 0.51 to 0.64; experiment 2: ranging from 0.61 to 0.69) exceeding the conventional threshold levels. Moreover, discriminant validity for all constructs was also established, which was demonstrated by AVE values exceeding corresponding squared correlations for all construct pairs (Fornell & Larcker, 1981). An overview of the measurement scales is given in Table 1, while Table 2 provides a summary of the relevant means, standard deviations, and inter-construct correlations.

Table 2 Descriptive statistics and correlation matrix (Experiment 1 and 2)
Full size table

Given all measures for the variables are self-reported and hence come from the same kind of source, common-method bias (CMB) can be a potential problem (Podsakoff et al., 2003). We attempted to mitigate this issue by introducing various procedures (MacKenzie & Podsakoff, 2012), by, for example, enhancing participants’ motivation to respond to our questions by giving careful instructions and minimizing the repetitiveness of the used scales. In addition to these data collection principles, we applied statistical procedures to assess CMB. By using CFA, we compared three measurement models: In the one-factor model, a common factor loaded on all items. The second model was a full measurement three-factor model, in which the items loaded on their respective factors (i.e., influencer authenticity, brand authenticity, brand attitude). The third model was a four-factor model, in which an unmeasured latent method factor loaded on the measurement items. To achieve model convergence, we specified all loadings of the method factor as being the same, which corresponds to the assumption that common variance affects all items equally (Rindfleisch et al., 2008; Homburg et al., 2015). For experiment 1, the one-factor model l (χ² = 486.020, df = 127, χ²/df = 3,83, RMSEA = 0.13, CFI = 0.81, SRMR = 0.10) fitted the data poorly. A significant difference was found between the two models (Δχ² = 223, Δdf = 3, p < 0.001). Comparing the four-factor model (χ² = 223.954, df = 123, χ²/df = 1.82, RMSEA = 0.07, CFI = 0.95, SRMR = 0.07) with our original three-factor model (see above) showed that the model fit did not substantially improve after inclusion of an unmeasured latent method factor. Furthermore, when comparing the standardized regression weights of both models, no substantial differences were identified. Comparable results emerged for experiment 2.

3.2 Data analysis

3.2.1 Determinants of influencer authenticity

To test the first set of perception-based hypotheses (H1 and H3), we used the data from experiment 1, in which we have manipulated influencer uniqueness (low vs. high) and longevity (low vs. high). A-priori analyses showed that the four experimental groups were homogeneous in respect to various control variables including gender (χ²(3) = 0.53, p = 0.91), education (χ²(3) = 15.42, p = 0.63), and age (F(3,159) = 0.63, p = 0.60). We applied a two-way ANOVA to examine the direct and interaction effects of the two perceptual constructs on influencer authenticity. To account for differences in group sizes, the sample was bootstrapped with 5,000 replications (Sadooghi-Alvandi & Jafari, 2013). Simple main effects analysis showed that influencer uniqueness had a significant effect on perceived authenticity (F(1,159) = 13.58, p < 0.001, η² = 0.80). As proposed, individuals who have perceived influencer uniqueness as high, also perceive the influencer’s authenticity as significant higher (M = 3.57, SD = 0.79) as compares to persons who witnessed low uniqueness (M = 3.13, SD = 0.71, p < 0.001). This supported H1. We did not find empirical support for H3, which assumed a relationship between perceived influencer longevity and influencer’s authenticity (F(1,159) = 0.69, p = 0.41). Furthermore, no interaction effect between uniqueness and longevity was found (F(1,159) = 0.17, p = 0.69). Table 3 summarizes the impact of influencer uniqueness and longevity on influencer authenticity.

Table 3 Impact of influencer uniqueness and longevity on
Full size table

By using the data from experiment 2, a similar procedure was used to investigate H2 and H4, which emphasize the effects of influencer scarcity and consistency on the perception of the influencer’s authenticity. The ANOVA showed a non-significant main effect of scarcity on authenticity (F(1,159) = 0.01, p = 0.94), which was in contrast to H2. However, we found a significant effect of influencer consistency on the same construct (F(1,159) = 22.73, p < 0.001, η² = 0.13). In support of H4, contrast analysis showed that influencer authenticity was significantly higher in the high-consistency group (M = 3.87, SD = 0.86) as compared to the low-consistency group (M = 2.28, SD = 0.71, p < 0.001). Furthermore, the analysis indicated a weakly significant interaction effect (F(1,159) = 2.87, p = 0.09, η² = 0.02). Individuals in the high-scarcity, high-consistency group tend to evaluate influencer authenticity as higher. The impact of influencer scarcity and consistency on influencer authenticity is summarized in Table 4.

Table 4 Impact of influencer scarcity and consistency on
Full size table

3.2.2 Brand-related effects of influencer authenticity

For testing H5-H7, we used model 6 of the SPSS PROCESS macro (Hayes, 2013). This macro applies an ordinary least squares regression-based path analytical approach to estimate the direct and indirect effects in mediation models. We used 5,000 bootstrap samples to estimate the 95% bias-corrected bootstrap confidence intervals (BCIs) for evaluations concerning indirect effects and confidence intervals. To cross validate our findings, we ran the model twice with data from experiment 1 (main sample) and 2. The latter was used as the holdout sample. In the serial multiple mediation analysis, influencer authenticity functioned as the independent variable, brand authenticity as the first mediator, brand attitude as the second mediator, and purchase intention was the dependent variable. For the main sample, the results demonstrated a significant indirect effect of influencer authenticity on purchasing intention (indirect effect = 0.14, boot SE = 0.05, 95% BCI [0.01; 0.22]). This indirect effect of heightened perceived influencer authenticity on increased purchasing intention was mediated by a positive effect of influencer authenticity on brand authenticity (b = 0.45, p < 0.001), which supported H5. A positive effect of brand authenticity on brand attitude (b = 0.71, p < 0.001) was observable, which was in support of H6. And finally, in support of H7, the test showed an effect of brand attitude of purchasing intention (b = 0.42; p < 0.01). The proposed mediation model only partially mediated the effect of influencer authenticity on purchasing intention as a direct effect still existed (b = 0.28, p < 0.01).

To validate our findings, the same analysis was used with data from the holdout sample. Here, consistent with our earlier findings, we also found a significant indirect effect of our independent variable (i.e., influencer authenticity) on individuals’ willingness to purchase the brand intention (indirect effect = 0.25, boot SE = 0.08, 95% BCI [0.11; 0.41]). Also, the proposed mediation sequence influencer authenticity → brand authenticity → brand attitude → purchase intention was empirically supported by the alternative data set. Figure 2 summarizes the findings of the two samples. A notable difference of the findings pertains the determinants of brand attitude: While this variable was only determined by brand authenticity in the main sample (R² = 0.53), in the cross validation sample the same variable was determined by brand authenticity (b = 0.36, p < 0.001) and influencer authenticity (b = 0.31, p < 0.001) (R² = 0.40). Hence, brand authenticity only partly mediated this effect in the second experiment. Besides this fact, the consistent results of the two samples provide considerable statistical support for the validity of the proposed hypotheses (H5-H7).

Fig. 2
figure 2

Serial mediation model: Effect of influencer authenticity on

purchase intention via brand authenticity and brand attitude

Full size image

4 Discussion

Marketers used to regard influencers (i.e., social media contributors who have gained notable recognition from others; De Veirman et al., 2017) as one of the companies’ most important advocates in describing their brands, sharing positive customer experiences, and recommend their products as well as services to current and potential customers (Masuda et al., 2022). This gave rise to the growing importance of ‘influencer marketing’ and companies’ attempts to win the most appropriate influencers, who can easily persuade their target group with social media posts, photos, and videos in which a company’s brand is emphasized. It is true that marketers can benefit from influencers’ reach and their ability to persuade others. This is because most influencers were typically perceived as trustworthy and having a unique expertise about the discussed or recommended brand. It is known that such perceptions make consumers to form pseudo-social (or ‘parasocial’) relationships with influencers – a process, which is quite similar to the fact that individuals develop feelings of intimacy towards media personalities after being regularly exposed to them (Alperstein, 1991). In the golden era of influencer marketing, online celebrities were a welcomed alternative to online ads. However, with companies’ increased practice to sponsor influencers and compensate for their efforts (i.e., granting them extrinsic benefits), consumers have learned that the information sourced from the often-beloved influencers may not be unbiased or intrinsically motivated. Consumers have adapted their ‘persuasion knowledge’ (Friestad & Wright, 1994). As a result, influencer marketing increasingly experiences drawbacks as it is currently struggling with sincerity issues (Audrezet et al., 2020). Today, consumers regularly question the influencer’s positive source characteristics – and most importantly, their ‘authenticity’ (i.e., a consumer’s evaluation of the extent to which a social media producer genuinely communicates purchasing-relevant information which resembles his/her true stance towards the presented and discussed products or brands). They only comply to his/her recommendations when it has been judged positively. When influencers authentically post content which aligns to their true selves, their persuasiveness dramatically increases (Kowalczyk & Pounders, 2016; Lou & Yuan, 2019; Woodroof et al., 2020).

Across two experiments, we obtained empirical support that ‘influencer authenticity’ is an important concept that mediates between several perceptual constructs, which determine its rise, and brand-related outcomes that ultimately converge into consumers’ increased likelihood to purchase products recommended by the influencer. More specifically, we show that consumers are more likely to perceive influencer authenticity for two main reasons: (a) in case when they appear to be ‘unique’ (Moulard et al., 2016), and (b) when they are perceived as being ‘consistent’ (Moulard et al., 2015).

Being unique means that the influencer does not appear to be a copy or an imitation of something or someone (Grayson & Martinec, 2004). This kind of perception has been identified as a critical factor also in other contexts such as celebrity authenticity, where it is used to be part of the rarity construct (Moulard et al., 2015). But also, in the context of influencer authenticity, other researchers regard the extent of uniqueness as relevant (e.g., Lee & Eastin, 2021). Here, uniqueness is typically understood as consumers’ perception that the influencer is distinct from others – both in terms of personality as well as content – and thus appears to be real. Extant literature emphasizes the importance of an influencer’s originality, which helps him/her to stand out. Duffy and Feist (2017), for instance, claim that consumers value influencers’ authentic self-expression because it is closely associated with their creative individuality. On the other hand, consumers treat content that seems to be ordinary, unoriginal, and thus as inauthentic disparagingly (McRae, 2017). In our research, uniqueness is a strong predictor of influencer authenticity. However, perceived scarcity (i.e., another dimension of influencer rarity) was not. This means that for consumers limits in the influencer’s accessibility (reflected by the number of fans) are not interpreted as being a result of the influencer’s intrinsic motivation (Park et al., 2021). Consumers regard the size of the influencer’s followership not as a sign of authenticity as, for example, social impact theory would suggest. This is in line with other research (e.g., Lee & Eastin, 2021), which suggests that similar constructs like an influencer’s visibility (i.e., being open and transparent to a multitude of others) is of minor importance. In this context also the act of following an influencer itself could be questioned as sign of a deeper relationship between follower and influencer. Like the results of John et al. (2017), ‘following’ a person could be just a symptom of being fond of that influencer but does not cause automatically a deeper personal relationship. However, consumers value consistency in personality and self-image. Authentic individuals can demonstrate that they do not alter their opinions, attitudes, and habits due to exogeneous factors, but they are dependable and predictable. We find strong support for the important role of integrity-factors in the recent literature (e.g., Farivar & Wang, 2022; Kapitan et al., 2021). For example, Lee and Eastin (2021) show that ‘sincerity’ (i.e., the extent to which a person is honest and truthful) is the most dominant factor explaining influencer authenticity. For them, sincerity is the key determinant of a parasocial relationship, which is the driving force behind influencer marketing. Also, Marwick and Boyd (2011) earlier claimed that followers carefully judge whether a celebrity is sincere as they pay attention to the extent to which the account portrays a true, unedited persona. Our research adds to these insights as the perceived reliability (i.e., perceived consistency) appears to be more valued than the perceived length of time the influencer is active (i.e., perceived longevity). The insignificant effect of perceived longevity is surprising but can be explained by the investigated media outlet. Social media outlets are, by their very nature, highly volatile and fast-changing environments. Therefore, the personal consistency of an influencer could be a stronger signal of stability than the time an influencer is actually active on a certain outlet.

Having this said, our research adds to the existing literature by drawing scholars’ attention to two critical factors (i.e., uniqueness and consistency) that determine influencer authenticity. Seemingly, these factors align to the current moral values of (younger) consumers, who apparently regard influencers as their desired likeness with which they identify (Choi et al., 2022). They focus on characteristics, which they cherish in their lives and that is to be consistent differently.

On the other hand, this research contributes by linking perceived influencer authenticity to brand-related outcomes. Specifically, we show that there is a positive spillover effect of influencer authenticity on a brand’s authenticity, which should be regarded as an important mediator. The critical role of brand authenticity has been highlighted by extant literature, which shows that the concept offers consumers several advantages – such as the ability to offer a sense of continuity in ever-changing, complex environments characterized by uncertainty (Fritz et al., 2017), cultivating the credibility of brands (Schallehn et al., 2014; Becker et al., 2019), enabling a unique brand positioning, increasing customer satisfaction (Bruhn et al., 2012), making consumers more receptive to brand messages (Audrezet et al., 2020), or even predicting consumers’ attitude towards brands (Napoli et al., 2014) and their willingness to purchase them (Fritz et al., 2017). To benefit from these potentials, brand managers must understand the nature of brand authenticity, its components, and boundaries. Consequently, literature mirrors various attempts to clarify the concept. For instance, in an early attempt, Beverland and Farrelly (2010) conceptualize authenticity being closely related to terms like ‘realness’ and ‘trueness’. Bruhn and his colleagues (2012) refer to this perspective by describing authenticity as a sense of trustworthiness. The same authors continue by discussing the multi-dimensionality of the concept and state that ‘originality’, ‘reliability’, ‘naturalness’, and ‘continuity’ should be regarded as the basic dimensions of brand authenticity. Several other approaches followed such as the work of Choi et al. (2015), who identify ‘consistency’, ‘heritage’, ‘sustainability’, and ‘origin’ as the main elements of brand authenticity. For some research, ‘credibility’ and ‘integrity’ also belong to the concept. For example, Morhart et al. (2015) state that an authentic brand should, amongst others, be able to keep its promises and to have moral pureness and responsibility. What these views have in common is that brand authenticity describes the extent to which a specific brand appears to be genuine, real, and true. (Beverland & Farrelly, 2010). An authentic brand conforms to consumers’ mental frames how it ought to be and has proven that it has a fit with consumers’ expectations (Beverland et al., 2008; Cinelli & LeBoeuf, 2020; Eigenraam et al., 2021). In this research, we follow the view of Napoli et al. (2014) and define brand authenticity as the consumer’s subjective evaluation of genuineness ascribed to a brand. An understanding that is quite dominant in current research (e.g., Das et al., 2022).

Furthermore, our two experiments shed light on the role of brand authenticity as a mediator between influencer authenticity and brand attitude. Typically, an image transfer takes place as it has been earlier asserted by the meaning-transfer-model (McCracken, 1989). Hence, influencer affect with their personality characteristics the perceived genuineness of the discussed or recommended brand. For instance, when consumers assume that the influencer making the recommendation is real and true, this positive mental frame disposes them to also regard the involved brand in a favorable light. Our results confirm such a positive transfer to the brand’s authenticity, but also a possible direct impact of influencer authenticity on consumers’ attitude towards the brand and purchase intention– disintermediating brand authenticity to some extent. These results can be explained with two main strategies of influencer-brand relationships. According to Audrezet et al. (2020) strategies of passion and transparency define an influencers authenticity on social media. Passionate influencers are strongly intrinsically motivated and do not seek a commercial interest when communicating brands. Therefore, in such an influencer-brand relationship the authenticity of an influencer may overshadow the brand and lead to direct effects on purchase intention. However, in the second scenario of a transparent influencer-brand relationship the influencer provides fact-based information about the product or service at the center of the brand and transparently reveals the commercial interest of such a communication measure. In this case the promoted brand itself plays a more dominant role and also brand authenticity as well as brand attitude are important gatekeeper constructs (Das et al., 2022; Eigenraam et al., 2021), which trigger purchase intentions. In our experiment, we did not manipulate influencer-brand relationships. Instead, we presented each stimulus with the same amount of transparency and passion, causing the aforementioned different roles of brand authenticity and brand attitude as mediators.

The current study provides several insightful practical implications for marketers. First, they can use influencer authenticity as a useful construct to segment (social) influencers. In principle, marketers could apply various approaches to classify influencers (e.g., popularity). For instance, serval studies suggest that an influencer’s credibility strongly affects consumers’ purchasing behavior (Chapple & Cownie, 2017; Djafarova & Rushworth, 2017). Consumers tend to evaluate a brand more positively when the endorser is perceived to be credible (Bergkvist & Zhou, 2016). Credibility consists of ‘trustworthiness’ (i.e., the perception of a person’s honesty and integrity) and ‘expertise’ (i.e., the perception of a person having the relevant knowledge, skills, or experiences) (Erdogan, 1999). An influencer’s credibility is likely to be damaged when his/her selling intent is disclosed by a company (Colliander & Erlandsson, 2015). Hence, the extent to which an influencer genuinely shares his true thoughts, personal views, and feelings is another important criterion, which is typically included in the concept of ‘influencer authenticity’ (Tsen & Cheng, 2021). Being authentic makes it easier for consumers to identify with the influencer as consumers think that they share similar interests and values (Kelman, 2006). Consumers not only identify with influencers because of perceived similarities but also because of a sincere desire to be like the endorser (Hoffner & Buchanan, 2005). When influencers present themselves as ordinary people who are approachable and have authentic personalities, they are very likely to trigger desired brand effects (Chapple & Cownie, 2017). Hence, marketers should consider influencer authenticity as the main segmentation approach to identify appropriate brand advocates in social media. More specifically, our research demonstrates that companies can benefit from a positive spillover effect from perceived influencer authenticity on consumers’ brand-related reactions. Clustering influencers according to their level of authenticity can help companies to identify the most appropriate influencers where such spillover effects are most likely. In absence of a direct authenticity measure, perceived influencer uniqueness and consistency, which are the two main drivers of authenticity, can be used as a proxy. Second, marketers should consider the possible (positive but also negative) spillover effects from influencer authenticity on their brand’s authenticity. Companies can benefit when the influencer seems to be consistent in his/her recommendations and opinions about a product as this also makes the brand more authentic. However, marketers should avoid mainstream influencers, who are stereotypes of their profession (i.e., controllable advocates of a brand), only share generic content, and get sponsored by various brands at the same time. De Veirman and her colleagues (2017) support this view as the authors have found that the perceived uniqueness of a brand (i.e., an element of brand authenticity) was reduced in case it was endorsed by a popular, average influencer.

The study has some limitations, which future researchers may attempt to overcome. A central limitation is its focus on the concept of one specific authenticity type. True-to-self authenticity, defined here as a consumer´s perception of the intrinsic motivation of an influencer, is one well established authenticity type within the Entity-Referent Correspondence Framework of Authenticity (Moulard et al., 2021). We believe that the two other types of the framework, true-to-ideal and true-to-fact authenticity, are ideal additions to conceptualize the full effect of the representation and communication of an influencer on consumer´s perceived influencer authenticity. True-to-ideal authenticity represents “the extent to which an entity´s attributes correspond with a socially determined standard” (Moulard et al., 2021, p. 99) and is strongly defined by categorization theory (Sujan et al., 1986) social constructivism theory (Leigh et al., 2006). Influencers are often role models for their followers and have to live up to a certain ‘ideal’ representation. Future studies should integrate this view and investigate how a discrepancy between social and/or cultural defined ideals and actual representation of these ideals effects perceived influencer authenticity. Additionally, true-to-fact authenticity defined as “a consumer’s perception of the extent to which information communicated about an entity corresponds with the actual state of affairs” (Moulard et al., 2021, p. 100) also signifies a valuable addition. Based on the notions of realism a fact is static and not a fluctuating ideal, it represents an underlying reality independent of the mind (Wartofsky, 1968). In influencer marketing the representation of an ideal may come into conflict with a fact-based approach creating a negative influence on consumers. For example, male teenagers are pushed by social media influencers into an unrealistic understanding of their body shapes and start to form muscle dysmorphia (Hawgood, 2022). Future research might consider these issues by detangling the effect of true-to-fact-based and true-to-ideal based authenticity cues on influencer authenticity and subsequently behavior.

Furthermore, another limitation of the study at hand is the focus on the influencers’ representation in form of information typically presented and available on social media outlets. We did not consider the design and content of the messages themselves or the interaction between influencer and sponsored product conceptualized by Audrezet et al. (2020). Future research endeavors should emphasize a better understanding of the role of the influencer as a person and the content communicated via the influencer and should shed light on the interaction between the two.

Finally, although we based our sample and stimuli material on typical industry standards, by selecting the most prominent age group of customers on the most prominent social media platform for influencers at the time of investigation, we cannot generalize our results for all customer groups and platforms. Therefore, we would like to encourage future research projects to overcome these limitations by investigating more demographical diverse samples on multiple platforms – ideally across different product contexts.