Typology of trust in micro-health and fitness influencers Open Access

Social media influencers and influencer marketing have become an integral part of brand communication. Similar to opinion leaders, social media influencers may impact the decision-making (Rogers and Cartano, 1962), as well as attitudes and behaviors (Godey et al., 2016), of their followers. Some companies even consider virtual influencers as partners (Byun and Ahn, 2023; Feng et al., 2024; Li and Ma, 2024). Companies frequently use influencer marketing through entrusting influencers with part of their marketing communication in efforts to improve customer attitudes and behaviors and subsequently company performance (Leung et al., 2022). Influencer endorsement increases customers’ brand awareness and brand appeal, as well as leads to increased purchases (Lau et al., 2023). As per brand transfer theory, influencers transfer part of their personality, reputation, credibility and values when advocating for a brand or product (McCracken, 1989; Tian et al., 2022). Such transfer can pose problems, though, if influencer-brand collaborations lead to concerns regarding influencers’ authenticity (Audrezet et al., 2018; Massi et al., 2024).

Whereas some scholars argue that the number of likes demarcate influencers (e.g. Kay et al., 2020), extant work suggests (Campbell and Farrell, 2020; Looi et al., 2023; Park et al., 2021) that influencers may be distinguished by the number of their followers. Whereas influencers with many followers are attributed to possess high expertise and cultural capital, influencers with few followers – so-called micro-influencers – encourage followers using accessibility and authenticity (e.g. Massi et al., 2024; Schouten et al., 2020).

Customers desire authenticity from brands. Communicated brand authenticity increases message receptivity, perceived quality and purchase intention (Audrezet et al., 2018). Authenticity can also positively affect influencer credibility (Massi et al., 2024). For example, compared to mega-influencers, micro-influencers (i.e. categorized as being between 10 and 100 thousand followers) create a greater sense of intimacy with, and their endorsements lead to higher perceived authenticity on, the endorsed brand (Park et al., 2021). In comparison to macro-influencers, micro-influencers have a higher number and a broader variety of advertised posts that follow subtle patterns that are often integrated into an influencer’s lifestyle and social activities (Hogsnes et al., 2024). Recent findings further suggest that, based on return on investment metrics, influencers with low followership outperform influencers with high followership (Beichert et al., 2024). Some scholars aver that an inverted U-shaped relationship exists between an influencer’s number of followers and their engagement with sponsored posts (Wies et al., 2023). The reasoning is that high followership implies a broader reach but a weaker influencer-follower relationship that reduces the likelihood of inducing follower behavior.

Research suggests that followers rely on micro-influencers because of their accessibility and authenticity (Park et al., 2021; Schouten et al., 2020). Followers perceive micro-influencer as more authentic and relatable than celebrities (Djafarova and Rushworth, 2017). This is because they have higher engagement rates through directly engaging more with their followers by answering questions and providing personalized advice (Kay et al., 2020). Micro-influencers can be regarded as peers or friends (especially compared to celebrities, Schouten et al., 2020). As such, influencers’ recommendations resemble word-of-mouth from trusted others, which is one of the most trusted forms of communication (Hennig-Thurau et al., 2004).

Previous research has considered trust as part of source credibility theory (see Table 1) and advertisement disclosure as an instrument to invoke trust (e.g. de Jans and Hudders, 2020; Karagür et al., 2022; Kim et al., 2024; Weismueller et al., 2020). Although that work has been valuable, researchers have neglected to study the various types of trust (McKnight et al., 1998; McKnight et al., 2002). Owing to this gap in the literature, the present study focuses on the importance of different trust types in (micro-)influencer marketing. It utilizes source credibility theory (Hovland and Weiss, 1951; Hovland et al., 1953) in influencer marketing (Pan et al., 2024) as a framework and replaces a single trust construct with a typology of trust (McKnight et al., 1998; McKnight et al., 2002). Our goal is to identify the role of four trust-based constructs – disposition to trust, institution-based trust, trusting beliefs and trusting intentions – in the realm of influencer marketing. The study thus contributes to the applicability and relevance of the typology of trust in the context of micro-influencers.

We chose micro-influencers in health and fitness as the study’s locus because both micro-influencers and their area of activity require trust. A prerequisite of trust is the presence of risk (Rousseau et al., 1998; Singh et al., 2020). Because micro-influencers have a smaller followership compared to influencers with more followers, social cues are less pronounced, thus creating new followers’ concerns about an influencer’s trustworthiness. Furthermore, health and fitness are areas dominant in credence attributes (i.e. product or service qualities that consumers cannot easily evaluate, even after purchase, and often rely on expert opinions or trust to assess, such as nutritional value, medical effectiveness, or environmental impact) that include a sense of risk regarding an individual’s (i.e. a follower’s) health. Micro-influencers in health and fitness typically possess specialized knowledge and are considered experts in their field; those characteristics can foster trust based on their expertise. Followers rely on them for advice on healthy fitness, nutrition and well-being (Powell and Pring, 2024; Zou et al., 2021). However, according to criteria of the World Health Organization (Winzer et al., 2022), 77% of food and beverage influencers’ cues would not be permitted in children’s advertising. For example, influencers affect healthy eating (e.g. De Jans et al., 2021) and perceptions of individuals’ body self-image (e.g. Tiggemann and Anderberg, 2020). Nonetheless, when seeking authentic and trustworthy advice, followers hold health and fitness influencers to high ethical standards because their recommendations often directly impact follower well-being. In the end, influencer statements can have positive and negative outcomes on followers’ health (Powell and Pring, 2024; Zou et al., 2021) – amplifying followers’ need for trust.

By integrating the typology of trust into source credibility theory, the present study enhances understanding of how customers build trust depending on nonsituational (i.e. personal) and situational characteristics. Furthermore, this integration allows comparing our trust dimensions with other (i.e. influencer) dimensions of source credibility theory (i.e. attractiveness and expertise). Thus, we endeavor to address issues regarding:

• how various dimensions of trust affect follower attitudes toward (micro-)influencers; and

• how these dimensions compare with (micro-)influencer attractiveness and expertise (i.e. dimensions in source credibility theory).

Using a manipulated online experiment, 414 micro-influencer evaluations showed that the effect of all four trust dimensions exceeded the effects of attractiveness and expertise. In influencer marketing, the results suggest that the trust formation process flows from disposition to trust through institution-based trust and trusting beliefs to trusting intentions. Furthermore, disposition to trust and institution-based trust are particularly relevant in the case of less experienced social media users or when there is a lesser degree of fit between micro-influencers and partnered brands. Finally, the integration of trust dimensions into the source credibility theory contributes to understanding of the complete trust formation process regarding micro-influencers and influencer marketing. Because the current study reveals the importance of nonsituational trust formation based on disposition to trust and institution-based trust, brands need to extend trust building beyond social media and influencer communication. This is because social media users also perceive that the fundamental brand image is important. Brands should, therefore, address trust building in their marketing and brand management activities.

Source credibility theory builds on the work of Hovland and Weiss (1951) and Hovland et al. (1953). The theoretical focus was initially on propaganda and media credibility before it was extended more generally to source credibility. Source credibility implies positive characteristics of a source that affect receivers’ acceptance of a message (Ohanian, 1990). A sender’s (i.e. a source’s) credibility is of utmost importance for a message to be persuasive (Hovland and Weiss, 1951; Ohanian, 1990).

Ohanian (1990) derived the three dimensions (i.e. attractiveness, expertise and trustworthiness) from source credibility scales to establish a measure for celebrity endorsement. Research on the comparatively new instrument of influencer marketing has broadly adopted that scale to assess credibility of influencers and determine followers’ attitudes toward them (e.g. Wiedmann and von Mettenheim, 2021; Xiao et al., 2018). Somewhat relatedly, empirical research has also found that influencer–brand congruence mediates the effect of influencer authenticity on influencer credibility (Massi et al., 2024).

Extant work on source credibility theory has extended the foregoing three dimensions. Previous studies, for example, have considered similarity (Munnukka et al., 2016; Xiao et al., 2018), congruence (Breves et al., 2019; Dhun and Dangi, 2023) and parasocial relationships (Reinikainen et al., 2020; Masuda et al., 2022) as extension of source credibility theory in influencer marketing. Lou and Yuan (2019) added the dimensions of entertainment and informativeness and then proposed a social media influencer value model. The effect of the three key dimensions in the source credibility theory on various dependent variables has generally been consistent and confirmed in previous research (see Table 1). Moreover, a recent meta-analysis positioned source credibility as a mediator of follower attitudes, engagement and purchase intention (Pan et al., 2024).

Although scholars have applied source credibility theory to influencer marketing contexts, no work has yet considered a typology of trust to aid understanding of the various effects of trust. As such, researchers have neglected the intricate trust formation process implied in a typology of trust. Further, they have overlooked influencer marketing and additional means to trust building in marketing. Presented in Table 1 is an overview of source credibility research regarding influencers.

Trust is one of the key success factors in customer relationships and electronic commerce, and, therefore, in influencer marketing (Kim and Kim, 2021). Trust generally relates to constructs such as confidence, honesty and reliability in relationships and transactions (Liu et al., 2011; Morgan and Hunt, 1994) and is directly related to perceived risk (Pavlou, 2003). Given these associations, scholars have extensively explored trust in numerous disciplines (e.g. electronic commerce, health care or organizations), viewing trust from unique disciplinary perspectives and creating various definitions of trust. Mayer et al. (1995, p. 712) defined trust as the “willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party.” A cross-disciplinary meta-analysis concluded that researchers have fundamentally agreed on the meaning of trust and thus defined trust as “a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another” (Rousseau et al., 1998, p. 395). This definition emphasized that trust is not a behavior or a choice; rather, it is an underlying psychological condition that can cause, or be a result of, such actions. Trust is based on individuals’ beliefs and confidence in the capability and willingness of another party in a given situation to fulfill trustor’s expectations, thereby adhering to the relationship norms and keeping promises (Ba and Pavlou, 2002; Schurr and Ozanne, 1985). Consequently, two conditions are prerequisites for trust to arise (Mayer et al., 1995; Rousseau et al., 1998; Schultz, 2007). One is that the trustor must have a confident expectation about an interdependent beneficial outcome. The second prerequisite is that the trustor relies on another party fulfilling trustor’s expectations in an underlying situation characterized by risk and uncertainty.

McKnight et al. (1998) conceptualized a typology of trust that included four trust constructs: disposition to trust, institution-based trust, trusting beliefs and trusting intentions. They discussed personality-based, institution-based and cognition-based foundations for initial trust formation and their corresponding model. Whereas cognition-based trust formation refers to cognitive cues, personality- and institution-based trust is predicated on externalities that lead to trusting beliefs and trusting intentions. The personality-based trust argues for trust development during childhood, as an infant seeks and receives help from his or her benevolent caregiver. This understanding results in a general disposition to trust others (McKnight and Chervany, 2001). With regard to influencers, disposition is trust refers to the general disposition of social media users to rely on influencer postings.

Institution-based trust is instead anchored in sociology and refers to trust in situational structures. For example, a person may trust because of guarantees, insurance or other safety mechanisms (McKnight et al., 1998). Moreover, Pan et al. (2024) have proposed a recent research agenda to explore the platform trust mechanism. Institution-based trust in the influencer context reflects on the acceptance of vulnerability in the broad online context (e.g. internet) or the narrow context of the social media platform. Trusting beliefs refer to specific beliefs about another party being benevolent, competent, honest or predictable in a situation (Mayer et al., 1995). Followers of micro-health and fitness influencers particularly rely on influencers general trustworthiness with regard to their messages and brand partnerships. Trusting intentions refer to the willingness of a person to depend on another party in a given situation to fulfill their expectations (McKnight et al., 1998; McKnight et al., 2002). Trusting intentions refer to the follower willingness to depend on a specific influencer message.

Displayed in Figure 1 is our interdisciplinary model of the trust typology. Whereas dispositional trust is a psychological state, institutional-based trust refers to a sociological category. Both concepts culminate in interpersonal trust (i.e. trusting intentions, which is a state of social psychology).

The present study integrates a typology of trust into source credibility theory to enhance understanding of the dimensions of trust in micro-influencer marketing. This integration shifts the focus from influencer trustworthiness to its relational counterpart, perceived trust. While trustworthiness reflects an aggregated degree attributed to a trustee, the typology of trust outlines the trust formation process across four underlying dimensions. Understanding this multidimensional process offers insights into the environmental and situational aspects of trust that stakeholders (e.g. brands, influencers and service providers) can leverage to enhance their roles in influencer marketing. Depicted in Figure 2 is the research model and related hypotheses.

Source credibility theory encompasses three key elements – attractiveness, expertise and trust (Ohanian, 1990). We briefly outline below the hypothesized effect of perceived influencer attractiveness and expertise on followers’ attitude toward influencers. Then, we discuss the integration of the typology of trust into source credibility theory and the effect of trust on followers’ attitude toward influencers.

Attractiveness. refers to the physical appeal as an important cue in an individual’s judgment of another person. Increasing the communicator’s attractiveness generally enhances favorable attitudes toward the person (Ohanian, 1990). Influencer attractiveness thus increases followers’ attitudes toward influencers’ messages/posts (Lou and Yuan, 2019; Munnukka et al., 2016). It also positively impacts followers’ purchase intention (Weismueller et al., 2020). Wiedmann and von Mettenheim (2021) further suggested that influencer attractiveness enhances brand image, trust and satisfaction. Although not all studies support the salutary impact of attractiveness (Dhun and Dangi, 2023), most extant work provides support for the positive relationship between influencer attractiveness and followers’ attitude. Hence, we propose the following hypothesis:

As part of Hovland et al.’s (1953) original proposition in source credibility theory, a source’s perceived expertise has a positive impact on attitude change (Ohanian, 1990). Research on influencers supports the positive effect of perceived expertise on influencers’ posts (Munnukka et al., 2016). Munnukka et al. (2016) further outlined the importance of perceived normalcy of an endorser, indicating its relevance in the case of micro-influencers. Expertise also influences electronic word-of-mouth (Dhun and Dangi, 2023), purchase intention (Masuda et al., 2022; Weismueller et al., 2020) and brand affect [e.g. brand awareness (Lou and Yuan, 2019); brand satisfaction (Wiedmann and von Mettenheim, 2021); and brand attitude (Dhun and Dangi, 2023)]. Overall, previous research has revealed that perceived expertise creates positive attitude changes in follower perceptions. Thus, we posit the following hypothesis:

In our development of the research model, we concentrate the discussion on integration of the trust typology into the context of source credibility theory vis-à-vis micro-influencer marketing. The present study adopted McKnight’s trust typology regarding the perception of micro-influencers. Source credibility theory already encompasses perceived trust, in addition to the attractiveness and expertise of an influencer (Dhun and Dangi, 2023; Ohanian, 1990; Wiedmann and von Mettenheim, 2021). The concrete trust formation process toward micro-influencers follows McKnight’s typology of trust.

The typology of trust is particularly apt in online contexts based on its outline in electronic commerce. In addition, it offers an approach to outline personal and situational trust formation – which is especially relevant when interactions are less direct (McKnight et al., 2002), as in a micro-influencer context. Whereas Kim and Kim (2021) proposed a positive effect of source credibility (as determined by expertise and authenticity) on trust, the present study replaces influencer trustworthiness (Ohanian, 1990; Table 1) with its relational counterpart, perceived trust (Lou and Yuan, 2019; Schultz, 2007). In this manner, the present study enhances understanding of how customers build their trust depending on nonsituational (i.e. personal) and situational dimensions and how these trust dimensions compare to the other (i.e. influencer) dimensions found in source credibility theory (i.e. attractiveness and expertise).

Whereas disposition to trust relates to personality trust (i.e. faith in humanity), institution-based trust pertains to security and its mechanisms (i.e. guarantees of the underlying structure). Personality-based trust exerts an overall effect on subsequent trust formation, such as toward the institution – social media and influencers in this study’s context. Users’ disposition to trust is the foundation for all subsequent trust processes (McKnight et al., 1998). Conceptually, another boundary trust process is trust placed in the institution or setting at large (McKnight et al., 2002). Hence, we propose a positive relationship between disposition to trust and institution-based trust:

Disposition to trust and institution-based trust subsequently affect the specific trust situation at hand (Schultz, 2007) – in the present research, forming trust toward a micro-influencer. Both dimensions help users engage in specific situations in which they perceive vulnerability and thus need to rely on trusting another party. Consequently, disposition to trust and institution-based trust exert positive effects on trusting beliefs and trusting intentions toward micro-influencers, as per the typology of trust (McKnight et al., 1998; McKnight et al., 2002):

Trusting beliefs and trusting intentions are thus formed regarding the specific situation, which entails interaction with a specific micro-influencer. In particular, fitness and health influencers may endorse brands and products based on the activities in which they engage and, therefore, influence purchase and consumption decisions (Hudders et al., 2021; Massi et al., 2024; Vrontis et al., 2021). Influencers may also promote values and lifestyles, such as healthy behavior and fitness. Such influencers may, for example, encourage their followers to exercise (Sokolova and Perez, 2021). During the COVID-19 pandemic, for instance, social media influencers shared guidelines and exemplified appropriate behavior in Finland to promote public health (Pöyry et al., 2022). Whether followers accept influencers’ public health messages depends on message information quality, influencer-user similarity and influencers’ credibility (Gupta et al., 2022). Influencers’ authenticity may also increase followers’ perceived credibility toward the influencer (Massi et al., 2024). Users’ trusting beliefs create a positive intention to trust an influencer in a specific situation (McKnight et al., 2002):

We now discuss the effects of trust on users’ attitude toward micro-influencers. Trust is a central element of source credibility theory (Ohanian, 1990). In fact, extant work supports the necessity of followers having trust in their influencers (see Table 1). Scholars also aver that influencers are more relied upon than celebrities (e.g. Djafarova and Rushworth, 2017; Schouten et al., 2020). Though the overall relevance of trust is widely accepted, research does not account for the different dimensions of trust toward influencers and in influencer marketing.

Munnukka et al. (2016) found a positive relationship between trust and followers’ attitude toward an influencer’s message. Trust also increases information credibility in YouTube influencers (Xiao et al., 2018) and in branded social media posts (Lou and Yuan, 2019). Research further supports the relationship between trust and purchase intention (Masuda et al., 2022; Weismueller et al., 2020) and indicates that the number of followers has a positive influence on this relationship. Contrary to the outlined findings on micro-influencers, Weismueller et al. (2020) proposed that followers perceived influencers as more trustworthy when they had a large number of followers. More broadly, perceived trustworthiness of influencers improved followers’ attitude toward the advertisement and the advertised product, as well as followers’ purchase intentions (Schouten et al., 2020). Yuan and Lou (2020), however, cautioned that perceived trustworthiness reduces followers’ interest in advertised products. Wiedmann and von Mettenheim (2021) further demonstrated that brand attributes (e.g. brand image, trust and satisfaction) mediated these effects. The effect of trust on brand attitude, however, was not replicated in the work of Dhun and Dangi (2023); instead, they found a positive effect of perceived trust on followers’ intention to spread an influencer’s message. The foregoing discussion suggests that previous research overall has found a positive relationship between followers’ trust in an influencer and followers’ attitudes toward the influencer. Hence, we propose the following hypothesis:

We used partial least squares structural equation modeling (PLS-SEM) to estimate the parameters in our measurement model (i.e. outer model) and structural model (i.e. inner model). The R package plspm (Sanchez et al., 2015) was used for data analysis. This variance-based SEM approach has limited distributional needs regarding the measured items and minimal computational requirements for the underlying algorithm (Hair et al., 2020). The use of PLS-SEM was appropriate because it supports the prediction and explanation of dependent variables in a theoretically grounded structural model (Hair et al., 2021; Sarstedt et al., 2014). In addition, PLS-SEM allows for complex models with respect to the number of variables and relationships, as well as affords modeling flexibility, has limited requirements for the distributional assumptions of the variables and sample size and provides convergence and stability of the results (Sarstedt et al., 2014). The present study especially benefited from PLS-SEM characteristics in its stability across similar dimensions of trust and captured non-normal distributions in potentially multi-modal constructs. In turn, PLS-SEM generally achieves high levels of statistical power for hypothesis testing (Hair et al., 2017).

As stimuli, we chose two health and fitness micro-influencers active on one of the most frequently used influencer social media sites, Instagram (Casaló et al., 2020). Instagram is one of the fastest-growing social media platforms and has considerable user engagement (Gregor and Olejniczak, 2023). Each micro-influencer focuses on content related to healthy eating (i.e. ingredients and recipes), active movement (i.e. training routines and exercise tips), as well as followers’ motivation and mindset for a healthy lifestyle. The influencers have 13.1 and 17.7 followers, respectively. Participants are shown the mobile representation of the micro-influencer profile that includes profile metrics, such as the number of followers. The stimuli also presented a recent post that includes a corresponding number of likes – in line with the 5.63% like-follower ratio identified by Kay et al. (2020). Instagram user engagement is characterized by a particularly large amount of time on the site, intention to follow trends and high sense of belonging. Users who follow brands on Instagram are further characterized by strong brand loyalty, as well as a high intention to join, become involved in brand communities and participate in brand activities (Phua et al., 2017). Relying on others – which followers generally do when they follow influencers – requires a level of trust (Rousseau et al., 1998), especially in areas of high credence attributes such as health and fitness (Powell and Pring, 2024; Zou et al., 2021).

All measurement scales were taken from extant work and modified to align with the context of health and fitness influencers. Participants responded to typology of trust items on a seven-point Likert-type scale, ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). They used seven-point semantic differential scales for all other constructs. The appendix lists all measurement items.

The trust dimensions were adapted from McKnight et al. (2002). Trust disposition and institutional-based trust were measured first, before presenting the stimuli. These scales referred to influencers in general. Trust disposition was assessed with four items; institutional-based trust, five items.

Then, each participant was presented with two different micro-influencers characterized as health and fitness influencers. To counter a potential order bias, the order of influencers was randomized. Participants read a snapshot of each influencer’s profile and one of their recent postings. The snapshot is the mobile representation of the micro-influencer profile including number of followers. This is combined with a recent post with a like-follower ratio of 5.63% (Kay et al., 2020) without any follower comments. In each post, the micro-influencers advertised a product that either fit their profile (i.e. congruent) or did not fit it (i.e. not congruent).

After introducing the influencer stimuli, participants responded to questions concerning the perceived attractiveness and expertise of each influencer. Five items tapped each of the two constructs, as per source credibility theory (Ohanian, 1990; Wiedmann and von Mettenheim, 2021). Attitude toward an influencer was measured using a five-item semantic differential (Spears and Singh, 2004; De Veirman et al., 2017). Trusting beliefs and trusting intentions were measured with five-item and three-item scales, respectively, from McKnight et al. (2002).

We conducted an online survey using a snowball method to collect data over two months in 2022, including one reminder. As seeds, four different persons (i.e. one senior and one junior professor and two students) distributed the survey invitation in their personal social media network. Participants were invited and asked to complete the survey. Upon completion, participants were then asked to distribute the survey to five persons in their social media network.

As control variables, we obtained participants’ demographic information (i.e. age and gender) and frequency of social media use. We collected participants’ consent for the study at the beginning of the survey. Participants consented to participate in the anonymous scientific survey willingly and withdraw at any time without negative consequences. Furthermore, survey items were randomly rotated in each questionnaire. We recoded measurement items to check for participants’ attention.

We received 209 completed questionnaires; one questionnaire was discarded because the participant did not consent to participate in the survey. All remaining questionnaires were further checked for completeness, response behavior (i.e. “straightlining”) and processing time (i.e. relative speed index [RSI] ≤ 2; Leiner, 2019). We identified nine questionnaires with RSI > 2 [2.01; 2.61], which we retained after a sensitivity analysis yielded similar results. After data cleaning, the final sample included 207 questionnaires, with 414 micro-influencer evaluations, thus exceeding the minimum sample size of n = 77 (effect size = 0.15; α error = 0.05; power 0.8).

Of the 207 participants, 61.4% (127) were female and 38.6% (80) male participants (Table 2). The sample covers a typical range of ages for social media. The mean age was 26.9 years old (SD = 6.06). The majority of the sample had a bachelor’s or master’s degree (75.8%), high school education (21.7%), completed training (1.9%) or a PhD degree (0.5%). Most of the sample used social media several times a day (64.6%) and 79.6% used Instagram regularly.

A variance-based structural equation analysis tested the study’s research model (see Figure 2). We examined the measurement models’ individual item reliability, composite reliability and discriminant validity. For validation purposes, we used bootstrap resampling with 5,000 cases and a significance level of 0.05. Beyond the structural designs against common method bias, we also calculated Harman’s single-factor test. The results revealed that the maximum variance explained by any single factor was 37.1%, thus remaining below the acceptable threshold of 0.50. In addition, full collinearity variance inflation factors (FVIFs) were calculated; this is a conservative approach for assessing common method bias (Kock and Lynn, 2012). Trusting beliefs’ FVIF = 3.55 was below the threshold of 5.0, and all other FVIFs were below the conservative threshold of 3.3.

The measurement models demonstrated reliable and valid measurements overall. When considering indicator reliability, two measurement items had to be excluded due to low loadings (IN1 = 0.460 and AT4 = 0.621). All other loadings of the measurement items were between 0.761 and 0.946 and exceeded the required value of 0.70. We can also assume construct reliability with values for Cronbach’s alpha (between 0.825 and 0.928) and Dillon-Goldstein’s rho (between 0.885 and 0.945) exceeded 0.70 for all latent constructs. In addition, the average variance extracted (AVE) for all constructs was at least 0.50 (between 0.657 and 0.826), so convergence validity can be assumed. Presented in Table 3 are the measurement results.

Discriminant validity was evaluated using Fornell–Larcker criteria, which entails evaluating the cross-loadings and the heterotrait–monotrait ratio 2 (HTMT2). All construct correlations were below the corresponding diagonal value that represents the square root of the average variance extracted. Moreover, the cross-loadings confirmed that every item had the highest loading with the corresponding construct. The HTMT2 ratios were all below the conservative threshold value of 0.85 – ranging from 0.055–0.816 (Table 4). Based on these criteria, the measurement approach shows discriminant validity.

The results of the structural model tended to confirm the overall research model. In particular, the findings supported source credibility theory, revealing a positive effect of perceived attractiveness (H1: β = 0.391, p < 0.001), expertise (H2: β = 0.381, p < 0.001) and trust (H9: β = 0.191, p < 0.001) on attitude toward micro-influencers. Thus, increased perceived attractiveness, expertise and trust lead to more positive attitudes formation toward micro-influencers. The results were less supportive of the typology of trust. Participants’ disposition to trust positively affected their institution-based trust (H3: β = 0.499, p < 0.001), but it did not affect participants’ trusting beliefs (H4: β = −0.090, p = 0.080) nor their trusting intentions (H6: β = 0.038, p = 0.311). Participants’ institution-based trust positively influenced their trusting beliefs (H5: β = 0.477, p < 0.001) but not their trusting intentions (H7: β = 0.010, p = 0.809). Finally, we obtained empirical support for the positive relationship between participants’ trusting beliefs and their trusting intentions (H8: β = 0.741, p < 0.001). The empirical results suggest that trust in micro-influencers forms gradually from disposition to trust over institution-based trust and trusting beliefs to trusting intentions. Although the typology of trust (McKnight et al., 2002) outlines such trust formation, the process within the context of influencer marketing appears less complex than the original model suggests.

In the context of influencers and influencer marketing, the results indicated a more staggered effect of trust formation. In contrast to trust formation in online shopping, trust building toward a micro-influencer may be more reliant on the personal influencer-follower relationship. When considering the total effect of all trust dimensions (0.029 + 0.069 + 0.142 + 0.191), the sum of the trust typology exceeded the effect of attractiveness and expertise on follower attitudes toward micro-influencers. The empirical results thus show the key role of trust in attitude formation. Summarized in Table 5 are the results of the standardized path estimates and results for the hypothesis tests.

Overall, disposition to trust explained 24.9% of the variance in institution-based trust. Thus, 24.9% of the differences in institution-based trust are solely determined by disposition to trust, whereas 75.1% are due to other factors. The model explained 19.3% of the variance in trusting beliefs in micro-influencers. The explained variance of trusting intentions was satisfactory (R² = 56.5%). Overall, the research model explained 60.8% of the variance of attitude toward the micro-influencer. The research model can thus explain 60.8% of the variability of attitude formation toward mirco-influencers and 39.2% of participants’ attitude toward micro-influencers have to be explained by factors outside the model.

When exploring boundary conditions, we first compared the research model for both micro-influencers used as stimuli in the study. The group comparison revealed one significant difference in the relationship between social media experience. The comparatively older health and fitness influencer had a higher effect from perceived experience on participants’ attitude toward the micro-influencer (β_y = 0.295, β_o = 0.486, p = 0.025). Because the significance pertained to the difference in level but are both positive, the finding still afforded generalizability across influencers – thus inferring stereotypical attributions of experience to older persons.

We also examined differences by other demographic variables. Although participants’ gender showed no statistically significant differences, there was a significant difference when comparing younger (<30) and older (≥30) customers. In terms of personal perceptions, the effect of institution-based trust was higher on trusting beliefs for older than for younger participants (β₃₀₋ = 0.414, β₃₀₊ = 0.609, p = 0.033). Two differences emerged concerning media usage. First, the effect of disposition to trust on institution-based trust was higher for regular than for heavy social media users (β_regular = 0.635, β_heavy = 0.432, p = 0.007). This result confirms intuitive assumptions. Second, trusting intentions had a greater impact on attitude toward micro-influencers for heavy social media users than for regular users (β_regular = 0.093, β_heavy = 0.229, p = 0.044). Consequently, heavy social media users were willing to place more trust in micro-influencers.

Finally, we examined whether the placement of congruent versus noncongruent products created differences in perceived attitudes toward micro-influencers. The results yielded a single difference. Disposition to trust had a higher effect on institution-based trust for noncongruent than for congruent product placements (β_noncongruent = 0.577, β_congruent = 0.430, p = 0.033). The results are intuitive that noncongruent product placements rely more on disposition to trust than congruent products.

Source credibility theory has seen abundant application vis-à-vis influencers and influencer marketing (Table 1). The present study’s results confirmed the overall applicability of the source credibility theory (Ohanian, 1990). In the context of micro-health and fitness influencers, participants’ attitude toward the influencers was predominantly formed by trust based on the total effects from the typology of trust (0.431) compared to attractiveness (0.391) or expertise (0.381). Trust formation, therefore, played a significant role in participants’ attitudes toward micro-influencers, thus confirming the general importance of trust and previous research results. However, now the corresponding roles can be attributed to the relevant trust dimensions of McKnight’s typology (McKnight et al., 2002).

Trusting intentions replaced perceived trustworthiness creditably, which has been conceptualized as social media users’ perceptions of trusting beliefs or trusting intentions in previous studies (e.g. Feng et al., 2024; McKnight et al., 2002). The addition of the trust typology provided enhanced understanding of the trust formation process. For example, because users with less experience in social media had to rely on disposition to trust, the direct relevance of disposition to trust and institutional-based trust was more pronounced in the context of (micro-)influencers and influencer marketing. Similarly, the results revealed that older participants found institution-based trust highly relevant. As such, these participants relied on the regulations and boundaries that their environment created, which referred here to the internet and social media processes generally adopted by their service providers.

The typology of trust revealed the indirect effect of disposition to trust through institution-based trust and trusting beliefs on trusting intentions. Accordingly, the context of micro-health and fitness influencers suggested a trust formation process from disposition to trust, to institution-based, to trusting beliefs and last to trusting intentions. Trust formation in influencer marketing may thus be more reliant on staggering the trust dimensions due to the personal influencer–follower relationship. This notion implies that – similar to the general trust concept – the typology of trust seems to depend strongly on the four dimensions that constitute the typology of trust. In contexts of more situational trust building (e.g. toward micro-influencers), the trust formation process may well be more directly impacted by the specific trust dimensions, such as trusting beliefs and trusting intentions. The less situation-specific dimensions (i.e. disposition to trust and institution-based trust), however, have a lesser impact on the trust formation process. Consequently, this study’s findings call for a typology of trust moderated by the situational specificity of the trust formation process.

Whereas previous research has relied on overarching understanding of perceived trustworthiness, the current study’s results show the direct link between users’ predisposition and environmental cues that create trusting beliefs (most importantly, becoming vulnerable in the situation; e.g. McKnight et al., 2002; Rousseau et al., 1998) and trusting intentions in a specific situation. As previously noted, situation-specific dimensions (i.e. trusting beliefs and trusting intentions) are especially relevant when more people have situation-specific knowledge and experience. When individuals are less aware of situation-specific cues, they rely on their non-situation-specific processes (i.e. disposition to trust and institution-based trust).

Our results further indicate that noncongruent product presentations are in greater need of non-situation-specific trust (i.e. disposition to trust). Noncongruent micro-influencer collaborations refer to cases in which brand–influencer congruence is perceived to be low. Under the assumption that followers prefer similarity with micro-influencers and considering previous interactions with an influencer, followers can generally refer to situation-specific trust (i.e. trusting beliefs and trusting intentions). However, if micro-influencers partner with less congruent brands, followers reflect on the disposition to trust and institution-based trust to assess the less congruent situation.

The present study provides valuable theoretical contributions to the trust typology toward micro-influencers and in influencer marketing. First, we confirm the established applicability of source credibility theory and extend source credibility theory by including different trust dimensions – non-situation-specific disposition to trust and institution-based trust, as well as trusting beliefs and trusting intentions – into source credibility theory.

Second, examination of the boundary conditions further reveals the importance of the various trust dimensions. Social media users and followers form trust differently, based, for example, on their social media experience and the degree of brand–influencer congruence. Depending on the boundary conditions, situation-specific or nonspecific trust dimensions are more relevant, though all dimensions still contribute to the trust-building process.

Third, the findings suggest a direct trust formation process from disposition to trust, to institution-based trust, to trusting beliefs, to trusting intentions. McKnight et al. (2002) suggested various paths in their typology of trust for which the present study found no support. Instead, our findings propose a less complex and more straightforward way of trust building with influencers and in influencer marketing.

Finally, beyond user characteristics, micro-influencer characteristics contribute to the predisposition of followers. The comparatively older micro-influencer was attributed to have a higher degree of perceived experience, and that experience had a higher impact on followers’ attitude toward the influencer.

The present study reveals the importance of trust in (micro-)influencer marketing. Consequently, brands need to build trustworthiness in their partnered communication. By paying keen attention to the trust formation process and to institutional factors, personal credibility and experience with influencers, brands can customize their strategic approaches and influencer partnerships to develop deeper and more effective relationships with their target audience. This is particularly relevant in risky situations and with credence goods and services (e.g. fitness and health here). Thus, brands should co-create with their partnered influencers to design the sponsored communication in a way that aligns their own presentation adequately with the sponsored brand.

Moreover, because the study revealed the importance of nonsituational trust formation based on disposition to trust and institution-based trust, brands need to extend trust building beyond social media and influencer communication. Beyond influencer marketing, brands should pay attention to their branding strategies. This is because social media users also perceive the fundamental brand image as important. Managers should strive to cultivate a high level of trustworthiness in their brands to align with consumers’ disposition to trust. Exemplified means are transparency, consistent quality and service, authentic endorsements and responsible practices. Customers with less social media experience are more likely to rely on their disposition to trust. These customers can be reassured through consistent messaging across multiple marketing channels. Brands should, therefore, address trust building in their marketing and brand management activities across multiple touchpoints. Thus, brands can further support their influencer marketing through creating trustworthy signals in their other communication and distribution channels. One key approach seems to be a consistent branding strategy.

Next to disposition to trust, institution-based trust refers to the acceptance of vulnerability in the situational setting. Consequently, brands need to rely on platform trust mechanisms (Pan et al., 2024) that in influencer marketing are particularly related to the Internet and the social networking site. Beyond closely monitoring actions of the social service providers, brands can actively engage in discussions that democratize social media platforms and actions that align with the brand identity.

Since trusting beliefs are key in driving trusting intentions, brands can build stronger trustworthy connections with their target audience through partnering with micro-influencers who emanate strong trusting beliefs. One way for brands to leverage consumers’ trusting beliefs is by carefully selecting influencers with a proven track record of trustworthiness in previous collaborations. Micro-influencers are particularly relevant partners, as they often have high engagement rates with their followers (Kay et al., 2020). Consequently, if followers believe that their partnered (micro-)influencers are honest and reliable, their willingness to engage with the brand is higher. For example, health and fitness influencers are held to high ethical standards, as their recommendations often directly affect follower well-being. Health and fitness influencers collaborating with brands should be perceived as authentic, ethical and highly responsible.

Brands can increase trust building further by increasing institution-based trust. Beyond direct followership, partnered (micro-)influencers should also have a good reputation on their social media platform. This criterion also enhances trust building in target segments, as older customers rely more on the trustworthiness of the environment. In addition, brands should provide guidelines and ensure that influencers communicate the partnerships openly so that transparency is guaranteed for brands’ influencer marketing activities. Thus, brands should embrace the path of sincere authenticity, as suggested in the authenticity management framework (Audrezet et al., 2018).

The results further highlight the importance of having a high degree of congruent brand presentation. Consequently, brands should pay acute attention to brand-influencer congruence and select micro-influencers according to their fit with the sponsored brand. When partnering with micro-influencers who are congruent with a brand, trusting beliefs and trusting intentions are more pronounced. Brands should focus on reinforcing the influencer’s expertise and consistency in product alignment. If the brands do not readily align with an influencer’s usual content, brands can emphasize disposition to trust and institution-based trust. In these cases, the trust formation process builds more on the influencer’s characteristics and the platform. Brands should proceed to explain the rationale for the brand-influencer partnerships and be transparent about it. Brands should start into a brand-influencer collaboration by clearly defining their identity, including key elements such as mission, values, tone and aesthetics. Next, managers can assess an influencer’s tone, style, values, authenticity and reputation based on the influencer’s content. Then, they can align goals and metrics – a small campaign may test the brand–influencer collaboration. Furthermore, the fit between the influencer and the brand has to be monitored over time to ensure that a long-lasting trusting relationship can transfer to the micro-influencer’s followership.

The trust-building process naturally begins with new influencer partnerships. In these cases, the target audience will rely on their disposition to trust and their trusting beliefs toward the micro-influencer. Accordingly, customers may have difficulty trusting the new brand–influencer partnership. Followers of micro-health and fitness influencers are particularly sensitive to trust, as followers often look for advice on improving personal health and fitness. Trust in this context is critical, as it impacts not only followers’ perceptions of a micro-influencer but also their willingness to take health-related actions based on an influencer’s health and fitness recommendations. Brands can address this issue through transparent communication and, as mentioned above, reinforce micro-influencers’ general trustworthiness through consistent and aligned messaging from the brand.

After the initial partnership, the strategic approach toward long-lasting trust relationships necessitates building long-term relationships with the partnered (micro-)influencers. Influencer marketing thus represents a strategic approach beyond single social media campaigns that need continuous support and cultivation to obtain the most out of these partnerships. Hence, influencer-generated content should be accompanied not only by brand management outside of the social media network but also on the social network to ensure an authentic and thus consistent and trustworthy communication strategy.

The present research study has certain limitations that are suggestive of avenues for further research. First, the present study focused on health and fitness micro-influencers who are said to require a high level of trust. Further research could be extended to explore other types of influencers (e.g. beauty, lifestyle and gaming influencers). Another relevant influencer type is the so-called virtual or AI influencer (Byun and Ahn, 2023; Feng et al., 2024; Li and Ma, 2024), which also raises considerable trust questions. Virtual influencers are managed by companies and agencies; thus, whether their communication is honest or authentic should be a key concern of social media users and potential endorsers. Therefore, scholars should investigate the role of trust in virtual influencers and the potential fear of perceived arbitrariness in brand endorsements.

Second, the present undertaking selected micro-influencers who require high levels of trust and are in particular need of trust considering initial trust formation. However, beyond micro-influencers, subsequent empirical research could account for influencers’ reach to address trust formation in depth. Also, future studies might examine macro-influencers and celebrities to inspect the trust formation process in influencers and influencer marketing judiciously. Our focus on a specific type of micro-influencer was based on the dominant credence attributes in health and fitness. However, researchers can address the role of trust formation in other influencer contexts. It might also closely consider content-specific influencer–brand cooperations and endorsements to augment understanding of trust formation in situation-specific contexts.

Third, our experimental design and data collection process may have introduced biases. Our cross-sectional analysis represented one particular observation of self-reported data at a given point in time across two micro-influencers. Longitudinal analyses may provide in-depth insights into the formation of trust and the ongoing relationship between followers and influencers. This is of particular importance, as trust is built over time (e.g. Rousseau et al., 1998; Djafarova and Rushworth, 2017). A longitudinal study may, for example, offer more depth on how trust builds or diminishes over time, particularly in relation to repeated brand collaborations. Moreover, micro-influencers may draw their authenticity (Massi et al., 2024) and sense of intimacy from their interactions with their followers over time (Park et al., 2021). Thus, trust building at its core is a longitudinal process. Similarly, further work may address the duration of the relationship between influencers and partnered brands. Based on this study’s results, trust formation may well change, so followers should be familiarized with the partnered brands.

Fourth, our sampling relied on referrals to obtain additional participants. This could have introduced unwanted bias because referrals likely share similar characteristics. Addressing Instagram users, the used seeds for referencing captured a population characterized as younger and more female. Though these characteristics are generally in line with the Instagram population, further research should opt for randomized sampling techniques.

Fifth, some researchers have begun to incorporate observational data in studies of influencer marketing (Beichert et al., 2024; Wies et al., 2023). Subsequent efforts could thus include mixed-method studies. Beyond field data, neurophysiological measurements may further help explain consumer processes in influencer marketing.

Sixth, our results on the typology of trust suggest a more intricate trust formation process in influencer marketing beyond a singular trust measure. Therefore, the present findings call for empirical research on how brands can address this complex process. For example, potential mediators or moderators (e.g. social media usage intensity and prior exposure to influencers) may provide more nuanced insights into the effects of trust on influencer credibility. In addition, what other instruments and means are best combined with influencer marketing to ensure the success of a brand’s marketing activities?

Disclosure statement: The author declares no conflict of interest. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Table A1