The role of influencer – follower congruence in the relationship between influencer marketing and purchase behaviour

.


Introduction
Today, social media platforms are often used by consumers before they make any purchase decision.Particularly, consumers rely on influencers who have become one of the most powerful driving forces for purchase decisions and are said to create an intense impact on brands (Lu and Seah, 2018;Saul, 2016).By using influencer marketing, brands aim at getting brand mentions from influencers, which leads to brand awareness and increased sales (Lou and Yuan, 2019).Previous research indicates that influencer marketing is likely to get more attention from consumers compared to brand-generated ads (Pereira et al., 2014), which could be considered a consequence of branded influencer content being more authentic, organic, and having direct contact with potential buyers.Influencers create content that is likely to draw the attention of customers (Cheung et al., 2022).These are some of the reasons why the value of influencer marketing has grown more than 800% over the last 5 years (Statista, 2021), resulting in companies dramatically increasing their marketing spending on influencers.
Finding powerful influencers is relevant for social media marketing success that can transform a potential customer into a real customer through authentic content that an audience can relate to.However, the credibility of social media influencers has decreased (Mscience, 2019), which makes it more relevant for companies and brands to find a credible influencer to promote them.Another big challenge for companies is to choose influencers that share the same values with their customers and can influence their purchase decisions.In other words, it is difficult for companies to select influencers that their target audience feels congruent with.
There are three types of conventionally recognized congruence that determine the effectiveness of endorsements and they include brandendorser congruence, brand-consumer congruence, and consumerendorser congruence.All three types are important factors to influence purchase intention and determine the effectiveness of advertising (Pradhan et al., 2016).Several studies have shown that the congruence between the identity of the brand and the identity of the influencer increases willingness to buy, and advertisement performance efficiency (Farhat and Khan, 2011;Bejaoui, Dekhil, and Djemel, 2012).The relationship between brand and consumer congruence is also well-supported and studied.Consumers are more likely to have a positive attitude towards brands that are congruent with their own self-concept in the context of social or cultural values (Sirgy, 1982).Strong congruence between brands and consumers has a strong impact on brand devotion, brand satisfaction (Achouri and Bouslama, 2010), preference of the brand (Branaghan and Hildebrand, 2011), trust (Freling and Forbes, 2005), and purchase behaviour (Wu, Tsai, and Lo, 2011).
While consumers use brands to display their self-concept to themselves and others (Aaker, 1999;Belk, 1988), it is likely that they may identify with certain influencers for the same reasons.As reported by Kamins and Gupta (1994), endorsers who have similarities with consumers are perceived to be more persuasive and viable.Previous research indicates the need for further exploration in the field of influencer marketing due to the enhanced ability of celebrities to engage with fans, leading to the development of unprecedentedly close relationships with their followers.However, there is limited knowledge regarding this aspect (Aw and Labrecque, 2020).This research therefore raises the following question: what is the role of influencer-follower congruence in the relationship between influencer marketing and purchase behaviour?
This study aims to make three key contributions.First, this study contributes to the current body of literature that addresses congruence between influencers and followers.Although congruence, or a match-up between the self-image of the consumer and the endorser's image, may affect consumers' responses (Choi and Rifon, 2012), making this type of congruence particularly important in the influencer marketing context, knowledge about this match-up is still scarce.Most of the previous studies focused on product-celebrity congruence (e.g., Yoo and Jin, 2015).The studies of Zogaj et al. (2021) and Shan et al. (2020) are some of the few works in academia devoting attention to this issue.Second, this research uses an innovative survey distribution approach, collecting data via a well-known Instagram influencer's account.It allows addressing the relationship between a specific influencer and her followers, and it is the influencer herself (that is also a co-author of the study), through her Instagram profile with more than 207,000 followers at the time of data collection period, who asks her followers directly about their social relationship explicitly disclosing to the participants that the author of the study was also the influencer herself.This provides a unique opportunity to examine the influencer-follower relationship in a real-world context providing valuable insights into the dynamics and impact of influencer marketing at a micro level.To the best of our knowledge, it is the first time that a specific relationship between an influencer and her followers is analysed as the scarce studies that do exist are based not on specific relationships but on generic ones.Third, this study provides managerial insights into how influencer marketing should be approached by companies from the congruence perspective.

Social media and influencer marketing
The emergence of social media networks has presented companies and brands with opportunities to utilize these platforms to accomplish their marketing and business objectives.Today, social media marketing is understood as an interdisciplinary and cross-functional process that leverages social media, often in conjunction with other communication channels, to achieve organizational goals by delivering value to stakeholders (Felix et al., 2017).It can also be defined simply as "the practice of utilizing social media channels to promote business products or services" (American Marketing Association).
Influencer marketing is regarded as a form of social media marketing that assists companies and brands in increasing awareness and bolstering their image (Reinikainen et al., 2020;Sokolova and Kefi, 2020).It is defined as a strategy in which "a firm selects and incentivizes online influencers to engage their followers on social media, leveraging these influencers' unique resources to promote the firm's offerings, with the ultimate goal of enhancing firm performance" (Leung et al., 2022).
Social media influencers have emerged as a new type of endorsers (Sands et al., 2022), functioning as self-generated content creators who receive monetary or in-kind compensation in return (Campbell and Grimm, 2019).They are described as 'a new type of independent third-party endorser who shape audience attitudes through blogs, tweets, and the use of other social media' (Freberg et al., 2011) and as 'people who built a large network of followers and are regarded as trusted tastemakers in one or several niches' (De Veirman, Cauberghe, and Hudders, 2017).An updated definition has been proposed by Lou and Yuan (2019) who defined an influencer as 'first and foremost a content generator: one who has a status of expertise in a specific area, who has cultivated a sizable number of captive followers-who are of marketing value to brands-by regularly producing valuable content via social media'.
Influencer marketing is seen as a new type of marketing strategy where the influence of opinion leaders drives consumers' purchase decisions and brand awareness.Over the last years, this strategy has been actively adopted by various organisations (De Veirman, Cauberghe, and Hudders, 2017;Lou and Yuan, 2019;Scott, 2015).It is often defined as a form of marketing in which companies invest in selected influencers to promote their brands through the influencer's created social media content to their target audience (Giles and Edwards, 2018).
While social media influencers are often described as celebrities and public figures, they are regular social media users who became known online by creating interesting and relevant social media content (Lou and Yuan, 2019).Social media influencers typically have expertise in selected areas and gain followers because of the specific content they create, whereas traditional celebrities could also be considered as social media influencers if they regularly create viral content on social media (Lou and Yuan, 2019).Also, celebrities achieve fame through traditional media and social media influencers are more seen as regular who became known on social media thanks to their relatable, close to followers and original content (Campbell and Farrell, 2020).Influencers are classified into celebrity influencers, megainfluencers, macroinfluencers, micro influencers, and nano influencers (Campbell and Farrell, 2020), while other authors also classify them into celebrities, macro-influencers, professionals, micro-influencers, and nano-influencers (Schaefer, 2019).Thus, influencers differ in the size and the attributes of the audience, which is important for brands, when they choose them.

The relevance of credibility
In the context of endorsement, the source credibility model (Hovland and Weiss, 1951) indicates that information transmitted by a trustworthy source of communication can affect the attitude and beliefs of the customer (Erdogan, 1999).The characteristics of the source become important when there is a need to persuade the audience (Buda and Zhang, 2000).An influencer who is perceived as more credible can affect the purchase behaviour and willingness to buy of the target groups more, as they are seen to possess higher levels of competence and trustworthiness (Silvera and Austad, 2004).In prior studies, credibility has been explored as a one-dimensional or multi-dimensional reality.In this study, we use a multi-dimensional approach with three components of credibility: expertise (competence), trustworthiness, and attractiveness (Lord and Putrevu, 2009;Ohanian, 1990).
Experience.In the context of this research, we refer to competence as the experience that the influencer possesses.Competence is the degree to which a communicator is treated as a source of legitimate statements (Hovland et al., 1953;Hovland and Weiss, 1951) and being able to provide the right information (Sussman and Siegal, 2003).As explained in previous studies, competence arises from awareness, practice, or training in the area concerned (Erdogan, 1999).According to prior research, consumers take advice and recommendations online before buying (Elwalda and Lu, 2016).Influencers usually have expertise in selected areas (Lou and Yuan, 2019) and their content is usually in line with those areas.Due to this, it is expected that influencers are considered a reliable source of information when it comes to shaping the attitudes and behaviour of their followers.While competence usually arises from awareness, practice, or training in the area of expertise (Erdogan, 1999), people are likely to rely on the opinion of the competent source which often leads to a positive attitude towards the D. Venciute et al. promoted product or brand and purchase intention (Ohanian, 1990).While previous studies found that, before buying, consumers consider advice and recommendations they found online (Elwalda and Lu, 2016), it is more likely that they will trust advice from influencers they perceive as competent.Thus, the following hypothesis is raised: H1.An influencer's experience has a positive direct impact on followers' purchase behaviour.
Trustworthiness.Trustworthiness describes how the endorser is truthful, trustworthy, and accurate (Ohanian, 1990).According to Lee and Koo (2015), trust is one of the most important aspects of relationships between influencers and their audience, which leads to an assumption that influencers will be able to influence the consumer only when they trust them.
Trustworthiness is seen as "the honesty, integrity and believability the endorser possesses" ( Van der Waldt et al., 2009, p. 104) and is one of the key factors to evaluate the credibility of the source.A source possessing trustworthiness helps generate a positive attitude towards a brand (Garretson and Niedrich, 2004).In turn, this may stimulate consumers to make a purchase decision due to their trust in what they saw or heard (Hubert et al., 2018).In the context of celebrity endorsement, product reviews by celebrities on social media were found to generate greater confidence (Djafarova and Rushworth, 2017).While influencer marketing is based on electronic word-of-mouth (eWOM) (Martínez-López et al., 2020), influencer-generated content may seem more trustworthy given the consumers' expectations for the influencers to act based on shared values (Porter and Donthu, 2008).Moreover, previous research found an affirmative impact of source trustworthiness on purchase behaviour (Mishra et al., 2015).Thus, it is proposed that: H2.The trustworthiness of an influencer has a positive direct impact on followers' purchase behaviour.
Attractiveness.In addition, when applying the source attractiveness model, previous research shows that endorsers' attractiveness is positively related to consumers' purchases intentions (Lee and Koo, 2015) and more positive attitudes towards a brand (Eisend and Langner, 2010).While influencers have a huge influence on attitude and willingness to buy when they are easy-going on social media and have popularity, they are also often perceived as attractive.
Previous studies found a positive relationship between endorser attractiveness and consumer attitude (Erdogan, 1999), which indicates that communicators with appealing attributes can have a favourable influence on customers' purchase behaviour (Till and Busler, 2000).Even though generally we are more likely to link the attractiveness of a source to its physical appearance and how it positively affects information processing (Roozen and Claeys, 2010), it is also related to the endorser's personality and intellectual skills (Erdogan, 1999).Previous research has discussed that even an unappealing message with weaker arguments can be overshadowed by the attractiveness of an endorser (Kahle and Homer, 1985).Thus, it is expected that the attractiveness of a source will have a positive relationship with purchase behaviour and the following hypothesis is raised: H3.The attractiveness of an influencer has a positive direct impact on followers' purchase behaviour.

Content usefulness
While the credibility of a source might be one of the most relevant factors when it comes to influencers, a common understanding nowadays is that to gain the affection of followers, influencers should create content that is valuable for the audience.Content usefulness can be described as the ability of content to help users make decisions or make progress toward their goals.Moreover, identifying the content that an influencer creates can even be seen as a feature of trustworthiness in a way that it helps customers who are seeking information and can rely on someone they trust.Content on social media is a tool that may influence customer beliefs and purchase behaviour (Müller and Christandl, 2019).The importance of content is even emphasized in the definitions that describe influencers as content generators who regularly produce valuable content via social media (Lou and Yuan, 2019).
Based on Müller and Christandl (2019), content usefulness is very important in social media platforms as it influences customers' beliefs and purchase behaviour.Therefore, content usefulness as an influencer marketing factor that will be included in the study.
Given that branded content produced by influencers is usually informative, entertaining, or inspirational, we expect that if followers perceive it as useful, it will directly influence their purchase intention and decision.Therefore, it is proposed that: H4.The usefulness of an influencer's created social media content has a positive direct impact on followers' purchase behaviour.

Influencer-follower congruence
The concept of congruence is often described in different ways such as fit, similarity, match-up, or typicality and only a small number of researchers used this term in the context of marketing (e.g., Malär et al., 2011).
In the field of celebrity endorsement, congruence between an endorser and a brand was found to influence purchase behaviour (Shimp, 2008).Generally, the selected endorser is said to largely impact the effectiveness of celebrity endorsement (Kim et al., 2014).Previous research has also indicated perceived congruence between the celebrity and the brand as a convenience match-up factor (Kamins and Gupta, 1994;Till and Busler, 2000).These studies show that the choice of endorser matters for companies and brands as it may directly impact the effectiveness and success of the ad or campaign.Given that social media influencers are often perceived as more trustworthy, genuine, and honest than other endorsers (Djafarova and Rushworth, 2017), it is believed that the match between the brand and a social media influencer is a required condition for good advertising (Sipka, 2017).
Despite the factors that may directly influence the purchase behaviour of followers in the context of influencer marketing, it is important to understand that traits, such as similarities and value closeness, in influencer-follower congruence may influence purchase behaviour (Albert et al., 2017).Previous research has suggested that when consumers identify with the endorser, there is a positive impact on purchase behaviour (Daneshvary and Schwer, 2000).While consumers are more likely to be drawn to brands that, in social and cultural contexts, are congruent with their own self-concept (Sirgy, 1982), it is more likely that the effect of influencers that consumers can identify with will be higher when it comes to purchasing behaviour compared to those they do not feel congruent with.Previous research has indicated that consumers prefer brands whose personality is congruent with themselves (Malär et al., 2011) or corresponds with their cultural values (Holt, 2004;Torelli et al., 2012).Also, in the social media context, self-congruence between consumers and influencers affects consumer behaviour (Shan et al., 2020).Zogaj et al. (2021) include two main dimensions in the self-congruence concept: actual self-congruence and ideal self-congruence (Malär et al., 2011;Zhu et al., 2019;Zogaj et al., 2021).Whereas the "the actual self describes a person's own self-image, that is, how one perceives him/herself in the "here and now", in contrast, the ideal self describes the person whom one would like to be" (Zogaj et al., 2021, p. 417).The main finding of the authors shows that actual and ideal self-congruence affect perceived trustworthiness and perceived competence differently and that actual self-congruence triggers perceived trustworthiness more strongly than ideal self-congruence.Shan et al. (2020) suggest that a 'high degree of congruence between the image of a social media influencer and the consumer's ideal self-image leads to effective endorsement outcomes' (p.605).The D. Venciute et al. authors also develop the concept of parasocial identification and claim that despite the followers not knowing the influencers, they feel like they intimately know them (parasocial relation) and a great degree of self-influencer congruence leads to a greater parasocial identification with the social media influencer, which in turn mediates the relationship between self-influencer congruence and endorsement effectiveness.Previous studies have indicated that a meaningful fit between the influencer and the brand is necessary to achieve positive brand outcomes (Hudders et al., 2021).Additionally, influencers leverage their established relationship and the trust they have built with their followers (Kim and Kim, 2021).Generally, influencers who are perceived as similar to their followers are more likely to be seen as credible and, consequently, increase the intention to purchase the advertised product (Sokolova and Kefi, 2020).Based on these findings, the following hypotheses propose that the actual congruence between an influencer and a follower, referred to as influencer-follower congruence, will moderate the relationships between perceived content usefulness, trustworthiness, attractiveness, experience, and purchase behavior.We anticipate that influencer-follower congruence plays a pivotal role in fostering or establishing a relationship between influencer marketing and purchase behaviour.
H5a. Influencer-follower congruence positively moderates the relationship between perceived usefulness of the influencer's social media content and purchase behaviour, such that for followers who are congruent with an influencer, the impact of content usefulness on purchase behaviour will be higher.
H5b. Influencer-follower congruence positively moderates the relationship between perceived influencer trustworthiness and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's trustworthiness on purchase behaviour will be higher.
H5c. Influencer-follower congruence positively moderates the relationship between perceived influencer attractiveness and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's attractiveness on purchase behaviour will be higher.
H5d. Influencer-follower congruence positively moderates the relationship between perceived influencer experience and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's experience on purchase behaviour will be higher.
The conceptual model is presented in Fig. 1.

Sample and data collection
This research paper applied a quantitative market research approach to gain deeper insights into the importance of follower-influencer congruence in the relationship between influencer marketing and purchase behaviour.Furthermore, the Probability Sampling Method was chosen for this research, given that this method makes it possible to acquire a sample that is representative of the target audience and to generalize research results for the entire population (Taherdoost, 2016).Simple random sampling in this research imparts equal opportunities for many social media users in the entire selected target population.
The target audience and participants of this research were recruited from the social network "Instagram", on the account of one of the wellknown Lithuanian influencer, famous for the content she creates in the context of lifestyle, fashion, and motherhood.Given that this study intends to address the relationship between a specific influencer and her followers, this approach was selected as the most suitable.According to the report of the Instagram platform, the Instagram account that was used to garner data for this research, at the time of conducting the study, had approximately 160,000 followers (it had more than 205,000 followers by the end of 2022), 81% of whom were female, and 21% -male.
The online survey was released on the Qualtrics platform on April 16, 2020.The survey was shared through the Instagram account of a wellknown influencer and data gathering took 2 days, from the 17th to 19th of April 2020.It was shared through the Instagram Stories function, where followers and other Instagram users viewing the stories were asked to fill in the online questionnaire.Stories' 'Swipe up' feature on Instagram is exclusively available to business accounts with 10,000 followers and more.Over two days of the data collection process, the online survey was opened about 80,000 times with 11,273 respondents completing it.A total of 7667 of the participants fully finished the survey the response rate being 68,01%.Such a sample size is exceptionally big, thus low sampling error is expected.

Sample profile and demographics statistics
The majority of the 7667 research respondents were female (98.1%), and only 88 respondents (1.1%) were male.This gender distribution is in line with the audience of the influencer.As mentioned before, according to Instagram data, at the time of conducting research, the distribution of the influencer consisted of 81% women, and 19% men.
The biggest age groups of the respondents were 'Under 25' (3655) and ' 25-34' (3637).According to this statistical data, our respondents are adults and are believed to have independence in purchase behaviour meaning that in most cases they can decide whether to make a purchase or not by themselves.Moreover, another question regarding income supports the proposition regarding respondents' ability to make independent purchase decisions.More than half (55.7%) of the respondents indicated receiving 501-1499 euros per month.33.8% of respondents (2594) indicated that their income is 501-999 euros per month, while 21.9% (1677) of respondents received 1000-1499 euros per month.The average net salary in Lithuania in the second quarter of 2020, was 883 euros.Most of the respondents receive income close to or above the average level in Lithuania.
The following question aimed to recognize the location of the respondents.Most of the respondents answered that they live in one of the two major Lithuanian cities: Vilnius (28.1%) and Kaunas (21.2%).
In addition, the question 'How long have you followed this influencer Instagram account?' reflects the length of time of following this influencer.Approximately 53% of respondents had followed the author of this survey for a couple years or more, which reflects a certain degree of trust and dedication to this particular Instagram account.It means that these followers are dedicated to this Instagram profile.About 41% of respondents had followed this Instagram account for about a month.According to this statistical data, about 94% of respondents were familiar with the influencer and her content.

Measurement scale
5-point Likert scales for influencer credibility: trustworthiness, attractiveness, and experience were adapted from Munnukka et al. (2016), with previous research adopted from Ohanian (1990).Moreover, a 5-point Likert scale for the influencer's content usefulness was adopted from Soh et al. (2009) scale named Ad Trust (usefulness).Thirdly, purchase behaviour scale items were adapted from Taylor and Bearden (2002), which were adapted from previous research by Aaker and Keller (1990).Moreover, the purchase behaviour measurement scale was adopted from Voorhees et al. (2006) measurement scale, named Purchase experience.Furthermore, the questionnaire's congruence measurement scale was adapted from the previous research of Johnson et al. (2006) that measured congruence (self with brand perspective).
To test the hypotheses, the following data analysis methods were used.Primarily, the descriptive statistics were analysed and will be discussed in the upcoming section of this paper.After that, the reliability of scales and internal consistency will be analysed using Cronbach's Alpha.Also, the next part of the analysis will focus on a regression analysis.In short, all these steps will be performed in the SPSS Statistics software program to examine the relationships between the independent and dependent variables that were presented in the conceptual model of this paper.

Reliability
The collected data was tabulated and analysed to estimate the Cronbach's Alpha value which is measured to estimate the internal consistency and understand how closely related a set of items are as a group.
The reliability test was implemented for all study variables and the results are overviewed in Table 1.The results show that all scales have good internal consistency and accuracy, as indicated by the Cronbach's Alphas falling within the range of 0.7 ≤ α < 0.9 (Streiner, 2003).
Additionally, the composite reliability (CR) of all constructs, ranging from 0.738 to 0.888, exceeds the recommended value of 0.6 (Fornell and Larcker, 1981).The values of average variance extracted (AVE) range from 0.416 to 0.599.It is worth noting that the construct Usefulness falls short of the recommended value of 0.5, while the other constructs have an AVE value above 0.5.However, Usefulness still exhibits acceptable reliability with a Cronbach's Alpha of 0.803 and CR of 0.739, in line with the criteria outlined by Fornell and Larcker (1981) and Lam (2012).
The Average Variance Extracted (AVE) analysis was employed as a method to assess the discriminant validity (Zait and Bertea, 2011).Table 2 presents the square root of AVE on the main diagonal, along with the bivariate correlations off the diagonal.Upon examination, it was concluded that the overall discriminant validity in this research is satisfactory.

Descriptive statistics
Descriptive statistics define the data in a sample across a variety of overview procedures and statistics.The 5-point Likert scale, where 1 means strongly disagree and 5 stands for strongly agree, was used to evaluate all the study variables, including the dependent variable, independent variables, and a moderator.Descriptive statistics of research variables show that the lowest mean score is determined for the dependent variable purchase behaviour (M = 2.68, SD = 0.85) signalling I feel this influencer is competent to make assertions about the products she promotes in social media posts I consider this influencer an expert on the products she promotes I consider this influencer sufficiently experienced to make assertions about the products she promotes Purchase Behaviour After seeing the post by this influencer, I have the intention to buy product X 0.891 6 0.573 0.888 After seeing the post by this influencer, I recommend product X to other people in the future After seeing the post by this influencer, I want to buy this product After seeing a recommended product on the influencer's social media, I purchase the product all the time I have good experience with purchased products, which was recommended in the influencer's social media After seeing a recommended product on the influencer's social media, I have purchased the product in the past

Usefulness of content
The content of this influencer: 0.803 4 0.416 0.739 Is valuable for me and fits my values very well Is good Is useful for me and I can find beneficial advice, recommendation, and acumen in my interested fields Helps a lot of people make the best decision in searching for desired products Congruence This influencer reflects my personal lifestyle 0.824 5 0.599 0.816 This influencer and my personality fit very well I can identify myself with this influencer I would like to take this influencer for dinner If this influencer had a brand, I would like to buy its products D. Venciute et al. that, on average, respondents answered neither disagree nor agree.The results show that the highest mean score is determined for one of the influencer's credibility attributesattractiveness (M = 4.39, SD = 0.54), showing respondents' agreement with this value in influencer's credibility and influencer marketing.The second highest mean score (M = 4.28, SD = 0.61) was found to be for the trustworthiness of an influencer.According to standard deviation data, in all cases, the SD is small (0.54 ≤ SD ≤ 0.69) and clustered closely around the mean.

Regression analysis and hypothesis testing
Linear regression was applied to check the hypotheses and the strength of the relationship of the variables.Two regression models were estimated to test the relationship between dependent, independent variables, and moderators.The first (1) regression model estimates the direct relationships and consists of the dependent variable Purchase Behaviour and the following independent variables: The second (2) regression model includes the estimates of moderators, thus the products of multiplication between independent variables (Trustworthiness, Attractiveness, Experience, and Usefulness) and moderator (Congruence) were included in the equation.This model has the same dependent variable and following additional right-hand variables: The results of both models are presented in Table 3. Direct effects.Overall, the first regression model was found to be statistically significant (at 95% confidence level) (F₁(5,7661) = 1089,856, p < 0.001).R Square (R 2 ) shows that in the first regression model independent variables explain 41.6% of the variability of the dependent variable (Purchase Behaviour).It was found that Usefulness (b = 0,316, p < 0,001) and Experience (b = 0,286, p < 0,001) have a positive and statistically significant (at 95% confidence level) effect on Purchase Behaviour, which supports hypotheses H1 and H4 respectively.However, Trustworthiness (b = − 0,018, p = 0,284) and Attractiveness (b = − 0,008, p = 0,634) have no statistically significant impact on Purchase Behaviour and leads to the rejection of hypotheses H2 and H3 respectively.
Moderated effects.Overall, the second regression model was also found to be statistically significant (F₂(9,7657) = 616,644, p < 0.001).The second regression model has R square equal to 42%, its adjusted R square is 42% and it is higher compared to the first model (41,5%).Therefore, the Second (2) regression model, with the moderator effect, explained more variability of the dependent variable Purchase Behaviour.The estimates of the second model show that Usefulness (b = 0,056, p = 0,023) and Experience (b = 0,063, p = 0,006) have a positive and statistically significant (at 95% confidence level) effect on Purchase Behaviour when moderated by Congruence.This finding supports hypotheses H5a and H5d respectively.However, Trustworthiness (b = − 0,033, p = 0,206) and Attractiveness (b = 0,026, p = 0,294) have no statistically significant impact on Purchase Behaviour even when moderated by Congruence and leads to the rejection of hypotheses H5b and H5c.To sum up, the empirical research reveals that Usefulness and Experience have a direct impact on Purchase Behaviour, both when it is and is not moderated by Congruence.However, Trustworthiness and Attractiveness have neither a direct nor moderated by Congruence impact on the dependent variable.The summary of hypothesis indications is provided  in Table 4.

Discussion
First, regarding the descriptive part of the results, it was shown that the lowest mean score is determined for Purchase Behaviour, signalling that on average respondents answered neither disagree nor agree.The results show that the highest mean score is determined for perceived influencer attractiveness.Previous studies in the context of traditional advertising have addressed that celebrities' attractiveness increases the effectiveness of the advertised products (e.g.Mathys et al., 2016).Also, based on Silvera and Austad (2004), an influencer who is perceived as credible and attractive can affect the purchase behaviour and willingness to buy of the target groups more, as possessing higher levels of competence and trustworthiness.One of the highest mean values was identified in perceived influencer trustworthiness.In line with the previous research study of Lee and Koo (2015), trust is one of the most important parts between a source and their audience, thus it is no surprise that it scored high.
When it comes to hypothesis testing, H1 was supported indicating that the perceived experience of an influencer has a positive direct impact on followers' purchase behaviour.This is in line with previous studies showing that consumers follow a competitive endorser's advice online before buying (Elwalda and Lu, 2016).Besides that, as discussed earlier in this paper, influencers usually tend to have expertise in certain areas and gain followers due to the specific content they create (Lou and Yuan, 2019).Therefore, the experience of an influencer when it comes to certain product categories of the sponsored brands is more likely to positively influence consumers' attitudes and subsequently impact purchase behaviour.
H2 was rejected, and the perceived trustworthiness of an influencer was not found to have a positive direct impact on followers' purchase behaviour.While previous studies considered trustworthiness as one of the most important aspects between endorsers and their audience (Lee and Koo, 2015;Zogaj et al., 2021) and found it has a positive impact on purchase behaviour (Mishra et al., 2015), the current study was not in line with previous research.This may be explained by the fact that in the context of influencer marketing, followers may be overwhelmed by the usefulness of the influencer's content as well as experience and may not care about how trustworthy an influencer is in relation to buying the products or services that they promote.
The attractiveness of an influencer was not found to have a positive direct impact on followers' purchase behaviour and therefore H3 was rejected.Given that no positive direct impact was found, it could be implied that followers do not need to perceive an influencer as attractive to buy the products or services they promote.It could imply that followers care more about how useful the content of an influencer is or what their competencies are instead of following content by someone who is attractive only.Previous research suggests different findings.While the attractiveness of an endorser is considered an important feature in terms of physical characteristics and the advertising message, and is considered a relevant factor in purchasing decisions (Chao et al., 2005), our research aligns with Hani et al. (2018) who suggest that celebrity endorser attractiveness does not have a significant effect on consumers' purchase intention.Our findings support this by demonstrating that attractiveness may not be a necessary factor in determining purchase behavior in influencer marketing.
H4 was supported indicating that the perceived usefulness of an influencer's content has a positive direct impact on followers' purchase behaviour.Based on a previous study by Hall (2016), the usefulness of content was indicated as a very important feature in influencer marketing that allows differentiation and is seen as a signal of trustworthiness that helps to become a valuable message sender for customers seeking information about products.This finding is also in line with previous research by Rui et al. (2013) who identified that when consumers make purchasing decisions, they are shown to rely heavily on the information, advice, and expert opinions they receive from people in their own network, which results in sales increase from positive purchase behaviour.Following Müller and Christandl (2019), content usefulness is very important in social media platforms, because it can influence customer belief and purchase behaviour.
Next, the results of the hypotheses considering the moderating effect are discussed.H5a was supported indicating that influencer-follower congruence positively moderates the relationship between perceived usefulness of the influencer's social media content and purchase behaviour, such that for followers who are congruent with an influencer, the impact of content usefulness on purchase behaviour will be higher.This finding aligns with the conclusions of Daneshvary and Schwer (2000), who emphasized the positive impact of the matchup between consumers and endorsers on purchase behavior.It also indirectly reflects earlier research indicating that congruence between the brand and the consumer has implications for brand devotion, brand satisfaction (Achouri and Bouslama, 2010), preference for the brand (Branaghan and Hildebrand, 2011), trust, and purchase behavior (Farhat and Khan, 2011).Furthermore, building on the work of Albert et al. (2017) and Pradhan et al. (2016), it is evident that the match-up between the brand and endorser image significantly influences advertisement positions and brand attitude.The current study extends the existing body of literature by highlighting that congruence can further enhance the effect of content usefulness on purchase behavior.Given the support for H1, this research shows that the perception of an influencer's content is significant for followers, both when there is an influencer-follower congruence and when there is not.
H5b was rejected indicating that influencer-follower congruence was not found to have a positive moderating effect on the relationship between perceived influencer trustworthiness and purchase behaviour, such that for followers who are congruent with an influencer, the impact of an influencer's trustworthiness on purchase behaviour will be higher.

Table 4
The summary of the hypothesis testing. No.
Hypothesis Rejected/ accepted H1 Influencer's experience has a positive direct impact on followers' purchase behaviour.

H2
The trustworthiness of an influencer has a positive direct impact on followers' purchase behaviour.

H3
The attractiveness of an influencer has a positive direct impact on followers' purchase behaviour.

H4
The usefulness of an influencer's created social media content has a positive direct impact on followers' purchase behaviour.

H5a
Influencer-follower congruence positively moderates the relationship between perceived usefulness of the influencer's social media content and purchase behaviour, such that for followers who are congruent with an influencer, the impact of content usefulness on purchase behaviour will be higher.

H5b
Influencer-follower congruence positively moderates the relationship between perceived influencer trustworthiness and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's trustworthiness on purchase behaviour will be higher.

Rejected H5c
Influencer-follower congruence positively moderates the relationship between perceived influencer attractiveness and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's attractiveness on purchase behaviour will be higher.

Rejected H5d
Influencer-follower congruence positively moderates the relationship between perceived influencer experience and purchase behaviour, such that for followers who are congruent with an influencer, the impact of the influencer's experience on purchase behaviour will be higher.

D. Venciute et al.
This finding shows that influencer-follower congruence is not a determining factor when it comes to the impact trustworthiness may have on purchase intention and contradicts earlier findings that suggest that closer relationships as an effect of congruence should result in higher acceptance of persuasive messages (Phua, 2016).
Similarly, H5c was also rejected showing that influencer-follower congruence does not have a moderating effect on the relationship between perceived influencer attractiveness and purchase behaviour.This means that even when there is congruence between the influencer and the follower, attractiveness does not significantly influence purchase behaviour.Even though previous research by Till and Busler (2000) discussed that the attractiveness of an endorser has a favourable effect on a customer's purchase behaviour when the target audience identifies themselves with endorsers in social media content.However, the present study did not find a significant impact of attractiveness on purchase behaviour, regardless the congruence level, showing that attractiveness is not a significant predictor in the context of influencer marketing and purchase behaviour.A holistic approach should be adopted regarding the concept of attractiveness instead of relying solely on it alone.This holistic approach should consider a comprehensive set of other factors (content quality, relevance, influencer expertise, …) that can enhance the effectiveness of the influencer marketing campaigns.
Finally, H5d was also supported showing that influencer-follower congruence positively moderates the relationship between perceived influencer experience and purchase behaviour, such that for followers who are congruent with an influencer, the impact of an influencer's experience on purchase behaviour will be higher.While people tend to rely on the opinion of the experience source and this likely leads to a positive attitude towards the promoted product or brand and purchase intention (Ohanian, 1990), this research also shows that the more a follower is congruent with an influencer, the higher the impact of their experience on purchase behaviour will be.

Managerial implications
The results suggest that influencers' experience and the usefulness of the influencer's created social media content are key elements to consider when selecting endorsers for marketing campaigns.When it comes to experience, influencers should be chosen based on their knowledge in a specific field.The congruence between an influencer and their followers is also an important element in the relationship between influencer marketing factors and purchase behaviour.This research is helpful for company or brand representatives when choosing influencers for their campaigns, as it suggests selecting those who align with the values of their target audience.Given the competitive market and the abundance of available influencers, representatives should find it easier to make decisions by assessing the congruence between a potential influencer and the target audience (Belanche et al., 2021).High congruence between the influencer and their followers will strengthen the relationship between experience, trustworthiness, and willingness to purchase the advertised products.In general, companies and brands should allocate resources to the influencer selection process by conducting qualitative or quantitative research with their target audience or by using the services of an influencer marketing agency.
When it comes to managerial recommendations for influencers, the findings of this research are evident in portraying the importance of content usefulness and experience, showing that it is important for consumers to see the value in the content they follow and to perceive its source as having expertise in a certain area.Instead of attempting to be perceived as experts in numerous categories, influencers should focus on selecting specific niches where they can demonstrate experience to their followers.Furthermore, when collaborating with brands, influencers should consider their target audience and align themselves only with those brands they believe are compatible.This aligns with previous research suggesting that such alignment can significantly impact their overall credibility (Evans et al., 2017).

Limitations and directions for future research
While previous research studies have looked into brand-endorser and brand-consumer congruence, studies on endorser-consumer congruence have been limited, which represented certain challenges when measuring and analysing the moderating effect of influencer-follower congruence in the relationship between influencer marketing and purchase behaviour.Previous research in this context mainly used a qualitative and experimental research design, whereas quantitative studies using questionnaires in this context were rare (Woods, 2016).Future research studies about influencer marketing and congruence between influencer and followers could use mixed research design approaches such as both qualitative and quantitative studies.Also, as the data only reflected how respondents felt about one particular influencer, the results should not be over-interpreted.One of the key limitations of this research study is the sample, which was composed of mainly female respondents and this unequal distribution among genders within the sample could have caused a higher potential for sampling error.However, this sample is in line with the audience that follows the influencer whose audience was questioned for this research.In the future, it would be interesting to compare the findings between male and female samples given that certain factors may be more important for one gender compared to the other.Also, the inclusion of only individuals of Lithuanian nationality restricts the generalizability of the findings to a broader population.However, despite these limitations, it is reasonable to assume that the overall patterns of social media users' behaviour would likely exhibit similarities across advanced countries.Future research could compare findings between respondents from different geographic regions as they might be influenced by different values and cultural characteristics.Another limitation to consider is that the study focused on an influencer who also served as a co-author.However, this decision was intentional as the study aimed to explore the relationship between a specific influencer and her followers.To ensure unbiased results, the anonymity of the respondents was strictly maintained in accordance with privacy regulations, mitigating any potential influence the influencer's involvement may have had on the respondents' perceptions.Therefore, it is strongly recommended that future research replicates this study with a wider range of influencers.

Declaration of competing interest
I am submitting the paper titled "The Impact of Perceived Influencer Credibility on Purchase Behaviour in Social Media Marketing" to the Journal of Retailing and Consumer Services for consideration for publication.I declare that I have no conflicts of interest that could potentially influence the research findings presented in this manuscript.
I understand that full disclosure of any potential conflicts of interest is important in ensuring the transparency and credibility of scientific research, and I have provided all relevant information in this regard.

Fig. 2
demonstrates how Congruence influences the impact of Usefulness and Experience on Purchase Behaviour.A unit increase in Congruence leads to a 0,056 change in the effect of Usefulness on Purchase Behaviour, while a unit change in Congruence results in a 0,063 change in the effect of Experience on Purchase Behaviour.Consequently, when Congruence is high, the effect of Usefulness or Experience on Purchase Behaviour could potentially increase by up to 0,4.

Table 1
Scale reliability with Cronbach's Alpha, AVE, and CR results.

Table 3
Results of regression analysis.