Understanding influence from the influencer’s perspective: examining leadership, attributes and dedication in social media influencers

Social media platforms are central spaces for building friendships and relationships through shared content. They allow users to exchange ideas and visual materials in an interactive environment. In this article, the term “social media” encompasses any interactive media technology that facilitates the exchange of ideas, information and various forms of expression across digital platforms (Tan et al., 2024). However, it’s essential to recognize that popularity within these platforms varies widely. Individuals who achieve significant popularity – commonly referred to as “influencers” – typically focus on specific areas such as fashion, cosmetics or lifestyle and create content that aligns with their expertise (Nosrati et al., 2023).

The momentum behind influencer marketing is primed for a substantial uptick in popularity within marketing strategies. Findings from a comprehensive survey of 1,300 marketing professionals reveal that an impressive 74% of companies are gearing up to invest in influencers over the next year (Dwidienawati et al., 2020). Beyond these statistics, recent reports by Hayes (2022) underscore the power of influencer endorsements on consumer behavior, indicating that a significant 71% of consumers are swayed towards making a purchase when products are promoted on social media. Moreover, a staggering 92% of consumers place their trust in influencer marketing reviews over those directly from brands themselves (Bokunewicz and Shulman, 2017). These influencers serve not only as conduits for brand promotion but also as opinion leaders and repositories of specialized knowledge. Delving deeper into the psychological mechanisms at play, Sweet and Adhikari (2020) shed light on the concept of social influence, elucidating how individuals’ behaviors, opinions and beliefs undergo metamorphosis under the influence of their network connections. This phenomenon often results in individuals aligning more closely with the views and preferences of their social circles, including those of influential figures within their online communities.

Recent studies emphasize the enduring relevance of influencer marketing as a vital component of digital marketing strategies. This is particularly evident in the increasing prominence of nano and micro-influencers, who, despite their smaller followings, achieve higher engagement levels due to their perceived authenticity and closer connections with their audiences (Vogue-Business, 2024). Moreover, brands are progressively acknowledging the necessity of fostering transparency in their collaborations with influencers, a practice aimed at preserving consumer trust and promoting ethical standards in advertising campaigns (Okonkwo and Namkoisse, 2023). Parallel to these developments, there is a noticeable cultural shift toward responsible consumption, reflecting more conscientious and ethical attitudes toward digital advertising practices (Cabeza-Ramírez et al., 2022). In addition, the emergence of corporate and virtual influencers further illustrates the evolution of this marketing approach, as it aligns with consumer expectations and explores innovative modes of interaction within digital ecosystems (Conti et al., 2022; Puccia et al., 2025). Together, these trends underscore the dynamic and increasingly sophisticated nature of influencer marketing in contemporary digital landscapes.

As highlighted by Kanaveedu and Kalapurackal (2022), the fundamental theories most frequently applied in influencer marketing have been three: source credibility model, persuasion knowledge model and social interaction theory. These frameworks have shaped the understanding of the characteristics of influencers and serve as a support for understanding from their perspective the different ways of exercising opinion leadership (OL) and influencing purchasing decisions. Accordingly, the objective of this study is to identify the combination of conditions and factors that contribute to high performance on social media (PSM) for influencers, from their perspective, using a qualitative comparative analysis with fuzzy sets (fsQCA). In addition, this document includes a case study analyzing these three influencers. One critical aspect of this research is to analyze the influencer from their own perspective rather than from the viewpoint of the consumer or follower. Ultimately, simple yet effective recommendations are expected to be provided for future influencers contemplating entering social media, as well as for the companies hiring them, as it provides guidance on what they should be like to achieve some success. To achieve this goal, the structure of this document includes: a literature review, methodology, analysis results, discussion and conclusions, presented in the corresponding sections.

Social media users engage with platforms to explore others’ lives, seek enjoyment and fulfil social needs (Houghton et al., 2020), while influencers’ success hinges on cultural capital and strong personal branding (Ki et al., 2020). Recent research examines how celebrities and traditional influencers shape purchase decisions (Arora et al., 2019; Bakirtas and Celik, 2024), whereas other work explores trust in AI influencers (Alboqami, 2023) and the shift to virtual influencers (Shao, 2024), and the impact of celebrity advertising on consumer behavior (Al-Heali, 2021) each focusing on specific factors like expertise, reliability, attractiveness or credibility. Influencer marketing operates as a persuasion process, where source characteristics (credibility, attractiveness, similarity and power) affect message receptivity (Erdogan, 1999; Farivar et al., 2022; Kang and Herr, 2006). In the knowledge-based economy, SMEs – including solo or small-team influencers – must leverage external knowledge to enhance competitiveness and optimize social media performance (Fidel et al., 2018; Wu et al., 2022). This study, therefore, tests five key conditions – knowledge-oriented leadership, OL, content attributes (CAs), personal attributes (PAs) and dedication to social media to determine their distinct roles in driving influencer success.

Research has shown that leadership is a crucial determinant influencing organizational performance, both directly and indirectly (Rehman and Iqbal, 2020). Knowledge-oriented leadership (K-OL) exemplifies a blend of transformational and transactional leadership styles (Donate and Sánchez de Pablo, 2015). This leadership approach aims to foster a culture of continuous improvement and innovation, thereby significantly impacting organizational success and adaptability in a knowledge-driven economy (Naqshbandi and Jasimuddin, 2018). In this regard, K-OL involves behaviors focused on constructing, promoting and sharing knowledge. It enhances learning experiences, facilitates external knowledge acquisition, improves communication, motivation and coordination skills, boosts moral and fosters a cohesive and supportive team environment (Donate et al., 2022). Influencers could play the role of knowledge leaders (K-OL) in their domains by using their platforms to share valuable information, facilitate access to external resources, encourage participation and foster collaboration among followers, focusing on education and enrichment rather than mere product promotion:

OL is characterized by personal involvement, familiarity and knowledge about a product, as well as uniqueness in the public sphere (Chan and Misra, 1990). Opinion leaders exhibit particular traits when influencing the decisions and behaviors of others (Chai et al., 2023). The ability of an opinion leader to exert influence is based on two essential qualifications: specific knowledge about topics or products and the intention to communicate with others (Kim et al., 2015).

Regarding influence on social media, OL can affect performance by influencing the perceptions, decisions and behaviors of the audience (Song et al., 2017). In digital environments, opinion leaders can gain followers, generate conversations and catalyze the adoption of ideas or products. However, effectiveness on social media will depend on the authenticity, relevance and consistency of their messages, as well as their ability to adapt to changing platform dynamics and audience expectations (Balaban and Szambolics, 2022). When influencers are seen as opinion leaders, their recommendations are more likely to be followed, significantly influencing consumer behavior and shaping public perceptions (Ki and Kim, 2019). Influencers who maintain authenticity, relevance and consistency in their messages may have a more solid performance. Consequently, the proposition is proposed:

User‐generated content – photos, videos, stories, animations and memes – is the core means by which influencers stay relevant and build credibility as opinion leaders through regular posts and follower interaction (Casaló et al., 2020; Childers et al., 2019; Lou and Yuan, 2019). In an environment of information overload on platforms like Instagram, TikTok and YouTube (Roetzel, 2019), truly unique, interactive, informative and visually appealing content drives engagement, boosts brand interaction and makes influencer marketing more effective and lasting (Ki et al., 2020; Tafesse, 2015). Consequently, the proposition is proposed:

Research in social media has highlighted the relevance of attractiveness in influencers (Farivar et al., 2022). Physical attractiveness refers to the appeal of the influencer’s appearance, while social attractiveness relates to the likability generated by their behavior (Sakib et al., 2020). Over time, it has been observed that media personalities with physical and social attractiveness receive greater attention and engagement from the audience, thus generating stronger connections (Bond, 2018). In influencer marketing research, conducted through large-scale online surveys, particular emphasis has been placed on relationship-based engagement, suggesting the positive impacts that influencers’ physical and social attractiveness can have on such relationships (Sokolova and Kefi, 2020). There is a natural tendency among individuals to prefer affiliating and socially interacting with attractive individuals, as this contributes to and enriches their own identity (Steger and Kashdan, 2009). Consequently, the proposition is proposed:

Social media influencers need to invest dedication, time and effort into developing their authentic personal brand, similar to how successful brands meticulously craft their identity (Gómez, 2019). For example, a fitness influencer might consistently share workout routines, meal plans and motivational content to build a reputation for expertise and inspiration in their niche. This ensures that the unique selling proposition and narrative resonate effectively with the intended audience (Khamis et al., 2017). Some influencers demonstrate a strong commitment to capturing attention in digital realms: they create engaging video tutorials, curate visually appealing Instagram posts, write informative blog articles, actively respond to viewer comments, collaborate with other influencers on podcasts and meticulously plan social media campaigns to maintain audience engagement (Brooks et al., 2021). The dedicated effort in attention management becomes an essential practice to establish oneself as an influencer.

As celebrities become more well-known, they often find ways to sustain their perceived authenticity. This authenticity is particularly appealing to brands and agencies when they collaborate on advertising campaigns with influencers. It signifies that influencers can maintain a genuine connection with their audience, enhancing the effectiveness of marketing efforts and bolstering brand credibility (Childers et al., 2019). Exclusive professional dedication to social media and high investment in improvements indicate a serious commitment, which could benefit performance. Daily dedication and increased income in the last year suggest consistent and successful activity on social networks. Consequently, the proposition is proposed:

Therefore, our research is based on the analysis of five conditions and their impact on social network performance (see Figure 1).

In this section, we introduce the approach used in this study to scrutinize and assess the suggested research framework. Building on the discourse above, we conducted an empirical investigation through a survey to explore how these attributes enable influencers to improve their effectiveness (performance) in social media (see  Appendix). Using a fsQCA, we delved into how combined circumstances culminate in a particular outcome (Ragin, 2018). This analysis, based on fsQCA, has recently been applied in the field of influencer marketing (see e.g. Bakirtas and Celik, 2024; Shao, 2024; Alboqami, 2023). As a complementary methodological approach, we carried out a case study analysis (Yin, 2018) focusing on three influencers and the variables detailed in Figure 1. The case study method was chosen because it enables an in-depth analysis of subjects in their real-world context, providing a nuanced understanding of complex phenomena. This approach proved particularly suitable for the objectives of this research, as it enabled the identification of distinct strategies and behavioral patterns used by each influencer to engage their audiences effectively and optimize their performance.

This research adopts a descriptive-exploratory design consistent with the main objective of the study. Currently, social networking platforms do not offer full access to public centralized databases of content creators (due to their privacy policies and commercial restrictions). Consequently, the application of probability sampling methods was ruled out (Le and Aydin, 2023). Based on this, Cartwright et al. (2022) was followed and theoretical purposive sampling was used (Eisenhardt and Graebner, 2007). A random selection of 100 potential participants was directly contacted through social networks (Twitter, Instagram and TikTok) in February 2024. A total of 26 responded, followed by two restrictive criteria aligned with the objective (followers/promotional experience). The measurement instrument was designed based on the literature review (see previous section), for its correct application, a pilot test was conducted with a manager of a marketing company, a researcher with experience in influencer research and an influencer. Based on the results, the order of the questions was regulated and. inconsistencies were corrected.

Our survey comprised 20 Spanish influencers actively promoting products or services on social media. We specifically targeted influencers with over 20,000 followers on their primary social network. To ensure response validity, the questionnaire began by confirming whether influencers engaged in promoting any products or services on their platforms (see Table 1). Only those who affirmed this were included in our sample. Data collection took place in 2024 via a validated survey reviewed by a panel of experts including professors, influencers and marketing professionals. The study’s sample may be affected by self-selection bias, as successful or confident influencers were likely more willing to participate, excluding lower-performing ones. Focusing on influencers with over 20,000 followers also omits valuable insights from micro-influencers and convenience bias from selecting participants based on proximity or relationships may limit the representativeness and generalizability of the findings.

The influencers were chosen based on practical considerations such as proximity, thematic relevance or existing relationships. To validate the measurement model, the data set underwent confirmatory factor analysis (CFA) using IBM SPSS. The results of the CFA confirmed the original five-factor structure. Those loadings with absolute values greater than 0.4 were considered relevant, indicating a substantial relationship between the variable and the latent factor. The conceptual model consists of five latent variables and one outcome variable, PSM.

All Cronbach’s alpha values are above 0.7, indicating adequate internal consistency (see Table 2). Composite reliability (ρ_a and ρ_c) (≥ 0.7): composite reliability evaluates the quality of the factor loadings for each construct. ρ_a and ρ_c for all constructs exceed the minimum threshold of 0.7, indicating that the latent variables are well-represented. Finally, average variance extracted (AVE) measures the proportion of variance explained by a construct compared to the error. An AVE ≥ 0.5 indicates good convergent validity. All constructs meet this criterion, demonstrating that the variance explained by the items is greater than the variance attributable to error.

The fsQCA (Ragin, 1998) methodology was used to identify causal configurations that explain PSM. We used the Fuzzy-Set/Qualitative Comparative Analysis 4.0 software (Ragin and Sean, 2022). This approach allows the analysis of combinations of conditions leading to a specific outcome, overcoming the limitations of traditional linear methods (de Diego Ruiz et al., 2023; Farooq Sahibzada et al., 2021; Liu et al., 2017). The five conditions analyzed were K-OL, OL, CAs, PAs and dedication (D). These conditions were selected based on previous studies highlighting their relevance to influencer performance.

Data collected through questionnaires was transformed into fuzzy values (ranging from 0 to 1) using the direct calibration method (Mattke et al., 2022). Specific anchors were established to determine membership levels: 0.95 for full membership, 0.05 for nonmembership and 0.50 as the crossover point. These scores, typically ranging from 1 to 7, are then converted into fuzzy sets, with values above 4 indicating full membership, 4 as the crossover point and values below 4 as nonfull membership, based on the methodology suggested by Ordanini et al. (2014).

In the second step of the analysis, a necessity analysis is conducted to determine which conditions are essential for a specific outcome (Mattke et al., 2022) (see Table 3). Consistency measures the extent to which a configuration consistently leads to the outcome. A threshold of 0.75 was used to ensure that the selected configurations had a sufficiently strong relationship with performance. Coverage indicates the percentage of cases that follow a specific configuration. Unique coverage was used to identify configurations that exclusively contribute to the outcome, excluding redundant ones with a unique coverage value of zero. The consistency and coverage thresholds were based on recommendations from the literature (Schneider and Wagemann, 2010) to ensure the reliability and robustness of the results. Consistency ≥ 0.75 ensures the configurations are sufficient to produce the outcome, while coverage ≥ 0.40 ensures their empirical relevance.

The analysis examines whether a particular condition consistently appears in all instances where the outcome is present, or whether its absence is noted in all instances where the outcome is absent (Nikou et al., 2022). Table 3 shows that OL consistently exceeds the 0.90 threshold, suggesting it significantly impacts the outcome (Fiss, 2007; Ragin, 2006). However, other conditions did not meet the threshold, suggesting they do not play a necessary role in influencing the outcome.

In the third phase of our analysis, we perform a sufficiency analysis to pinpoint all configurations that effectively result in a specific outcome (Nikou et al., 2022). Initially, the truth table (Fiss, 2011; Ragin, 2006, 2008) presents various potential configurations that could potentially induce the desired outcome. The truth table was generated to enumerate all possible combinations of conditions. Each configuration was evaluated to determine its ability to explain the desired outcome. Parsimonious, intermediate and complex solutions were generated using the Quine-McCluskey algorithm (Pappas and Woodside, 2021). The intermediate solution was selected as it balances simplicity and detail, facilitating interpretation without losing essential elements.

The fsQCA results, based on the causal configuration of five conditions (see Table 4), reveal 11 solutions (S1 to S11) aimed at analyzing performance (Ragin and Fiss, 2008). These conditions include K-OL, OL, CAs, PAs and dedication. In the representation, the black circles represent the presence of a condition (●), the white circles signify its absence (○) and blank spaces denote that the condition is irrelevant without affecting the outcome (Ragin and Fiss, 2008).

The final configurations were evaluated based on their consistency and coverage to ensure they were both statistically sound and practically relevant. Configurations with high values in both indicators (e.g. S10, with 83.3% consistency and 80% coverage) stood out as the most explanatory. All solutions have high consistency ≥0.8, indicating that each individual configuration is sufficient to achieve the desired operational performance (Ragin, 2006).

The selection of the following three influencers – La Cocina de Poblete, Teresa Guitar and Comerte Córdoba – is particularly suitable for a qualitative analysis due to their distinctive profiles and the unique value they bring to their respective fields. Each represents a specific niche – culinary arts, music and gastronomy – enriched with strong cultural, artistic or community-driven elements. These qualities make them ideal subjects for exploring how different variables influence their impact and engagement strategies (Table 5).

First, La Cocina de Poblete, a chef blending traditional recipes with humor, excels on YouTube (26 K subscribers), Instagram (68 K followers) and TikTok (124 K followers). Focused on classic Manchegan dishes, he has appeared on shows and is a key advocate for regional culinary heritage. Teresa Guitar, a flamenco guitarist (in Instagram 221 K followers), merges traditional flamenco with other genres. After offering online classes, she now performs internationally, collaborating with artists like DJ Martin Garrix, pioneering the fusion of flamenco and contemporary styles. Finally, Comerte Córdoba (in Instagram 28 K followers), a gastronomic initiative in Córdoba, Spain, celebrates local cuisine. Its March 2024 edition featured 16 food stalls, 5 beverage stands and performances at Castillo de la Albaida, promoting gastronomy and community collaboration. Next, we will study qualitatively these influencers in the variables analyzed.

Table 6 presents intermediate solutions (S1 to S11) for performance on social networks, evaluating the presence or absence of specific conditions and providing metrics related to coverage and consistency. These conditions are considered essential elements for explaining performance on social networks in the context analyzed. To better analyze the results obtained from the Quine-McCluskey test, we have synthesized them into the following table:

The analysis of these cases underscores that influencer success is not determined by a single factor but by the interplay of multiple elements. Each influencer uses these dimensions in a distinct manner, shaped by their specific context and audience characteristics. Common factors such as authenticity, content innovation and strong professional commitment emerge as pivotal contributors to their effectiveness. In the following section, we will assess the acceptance or rejection of the initial propositions, informed by the findings derived from fsQCA and the insights gathered from the case studies (Table 7). Emphasis will be placed on identifying the most consistent and broadly applicable conditions while considering the contextual factors that influence these outcomes.

This study has uncovered key factors that significantly impact influencer performance on social networks. By focusing on conditions such as K-OL, OL, CAs, PAs and dedication on social networks, crucial patterns and relationships have been identified. Different combinations of conditions explain social media performance, with OL and CA as key factors. Coverage and consistency metrics reinforce the effectiveness of specific solutions like S10, S8 and S7.

The findings emphasize the critical role of OL and content quality in influencer success, highlighting authenticity, relevance, uniqueness and aesthetic appeal as essential elements for effective influence. Companies can leverage this information to select influencers who possess these characteristics, thereby increasing the likelihood of success in their marketing campaigns. Furthermore, K-OL and personal traits, such as physical and social attractiveness, also play a significant role in influencer effectiveness. Companies and agencies can provide training to develop these skills. Ultimately, the use of the fsQCA methodology has proven valuable in assessing the complexity of these combined factors, offering a deeper understanding of their interaction and effect on influencer performance.

The theoretical implications of this study enhance our understanding of influencer PSM by incorporating various influential factors and validating methodologies like fsQCA to address this complexity. The identification of intermediate solutions offers a more detailed perspective on causal relationships in this context. This study offers significant insights into how brands and marketing professionals can effectively harness the potential of influencers. By identifying the factors that drive influencers’ success on social media, companies can optimize their marketing strategies and build more impactful collaborations.

For instance, in the food industry, an olive oil brand might partner with La Cocina de Poblete to sponsor a series of TikToks featuring traditional recipes. Such a collaboration would not only enhance brand visibility but also highlight its association with regional quality and cultural heritage. In the music and technology sector, a company specializing in music production software could sponsor Teresa Guitar’s international tours or develop co-branded content. This initiative would showcase how its technology supports the fusion of flamenco and electronic music, positioning the brand as a key innovator in the music industry. Similarly, in the promotion of gastronomic tourism, a regional tourism office could team up with Comerte Córdoba to organize a festival celebrating local cuisine and cultural heritage. This event would attract tourists, stimulate local businesses and spotlight the region’s culinary diversity. These examples demonstrate how brands can capitalize on the expertise and authenticity of influencers to create impactful campaigns that deliver value for all stakeholders involved.

Social media influencers shape not only consumer behavior but also public discourse and information quality. Their credibility and message quality boost adherence to health guidelines – evident during COVID-19 (Gupta et al., 2022) and their reach can set political agendas and influence policy debates (Iuliia et al., 2021). In youth mental health, rigorously sourced influencer content enhances health literacy and peer support (Adeane, 2024), yet there’s a danger of promoting unrealistic ideals that harm self-esteem and well-being (Engel et al., 2024). To harness their positive societal role while mitigating risks, collaboration among researchers, brands and health agencies is essential for crafting ethical, evidence-based guidelines.

The study’s limitations may include some subjectivity in the process of selecting intermediate solutions following the application of the Quine-McCluskey algorithm. In addition, the data is based on surveys targeting influencers with over 20,000 followers, which could limit the generalization of the results. The evaluation of the adequacy of fsQCA configurations is based on measures such as consistency and coverage, which may not fully capture the complexity of causal relationships in the environment of influencers on social media.

This study may have self-selection bias, as more successful influencers likely participated, potentially excluding those with lower performance. In addition, focusing on influencers with over 20,000 followers excludes micro-influencers and convenience bias may arise from selecting participants based on proximity or prior relationships. These limitations could affect the representativeness and generalizability of the findings.

As for future research, other factors not considered in this study could be explored, such as posting frequency and interaction with followers. Furthermore, investigating how specific marketing strategies can impact the success of influencers on social media would be interesting. Different methodological approaches for data analysis in this field could also be examined, such as the use of data mining techniques or social network analysis.

Future research could explore how content frequency affects engagement across niches, the impact of emotional connections and parasocial relationships on consumer behavior and whether influencers with specialized knowledge generate more trust. In addition, examining the long-term effects of influencer marketing on brand loyalty, comparing virtual vs human influencers in terms of trust and engagement and studying cross-platform performance could provide deeper insights into influencer marketing dynamics.