Original Research

Examining the influence of social media influencers on beauty product purchase intentions amongst South African consumers

Farzaana Motara, Nicole Cunningham, Christine De Meyer-Heydenrych

Received: 27 Oct. 2025; Accepted: 24 Feb. 2026; Published: 29 Apr. 2026

Copyright: © 2026. The Author(s). Licensee: AOSIS.
This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license (https://creativecommons.org/licenses/by/4.0/).

Abstract

Background: Social media influencers (SMIs) are critical information sources for beauty and personal care product purchases, yet limited research exists on their effectiveness in emerging markets like South Africa. Existing studies focus on the developed markets and peripheral persuasion factors, neglecting central routes (argument quality) to information adoption.

Objectives: This study investigates how SMIs impact South African consumers’ beauty and personal care product purchase intentions by extending the information adoption model (IAM) to include central and peripheral route factors.

Method: A quantitative cross-sectional survey design was employed using online questionnaires distributed via social media platforms. Non-probability snowball sampling yielded 242 usable responses from South African social media users aged 18–65 years who followed beauty influencers. Structural equation modelling was conducted to test the hypothesised relationships.

Results: Relevance, expertise, trustworthiness, similarity and likeability have significant positive relationships with perceived usefulness. Comprehensiveness does not have a significant relationship with perceived usefulness. Perceived usefulness significantly influences information adoption, which positively affects purchase intention.

Conclusion: Central and peripheral factors influence SMI effectiveness, with expertise demonstrating the strongest impact. The non-significant comprehensiveness effect challenges information richness theory, suggesting that focused content outperforms exhaustive information in social media environments.

Contribution: This research extends the IAM by incorporating source attractiveness constructs and challenges established information processing theories, as well as validates the model in an emerging market context.

Keywords: social media influencer; information adoption model; consumer behaviour; purchase intention; beauty products; South Africa.

Introduction

The digital revolution has fundamentally transformed the landscape of consumer behaviour and marketing communication, creating opportunities for brands to connect with consumers through innovative channels like social media (Verplancke & Gelati 2022). Social media platforms have evolved from simple networking tools into sophisticated information ecosystems where consumers actively seek, evaluate and share product-related information. This behaviour has prompted the use of and reliance on social media influencers (SMIs) who bridge the gap between brands and consumers, offering authentic, relatable and trusted perspectives on products and services (Spörl-Wang, Krause & Henkel 2025). This is especially prevalent in the beauty industry, where Mya et al. (2025) reported that 36% of surveyed consumers actively search for beauty and personal care product information from SMI content, a finding drawn from an online consumer study which reflects broader patterns of digital information-seeking behaviour in this sector. This is supported by Verplancke and Gelati (2022), who stated that beauty and personal care products are used in a visual way where they signal personal expression and self-image, creating a symbiotic relationship between SMIs and the beauty industry.

The beauty and personal care industry in South Africa represents a substantial and rapidly growing economic sector that is crucial to the country’s consumer economy. With an expected revenue value of $4.51 billion in 2025, the industry demonstrates remarkable resilience and growth potential, with an estimated annual growth rate of 3.24% that significantly outpaces general economic growth rates (Statista 2025). This growth trajectory positions the beauty and personal care industry as a key sector attracting consumer spending within South Africa. The sector includes the full range of beauty and personal care products, consisting of cosmetics, personal care products and fragrances (Research and Markets 2024; Statista 2025). The industry is highly competitive with established brands like Unilever, Colgate-Palmolive, and Procter & Gamble holding the majority of the market share (Euromonitor 2023). However, emerging and independent beauty brands, namely, those outside the portfolio of large multinationals such as Unilever, Colgate-Palmolive and Procter & Gamble, have leveraged social media and SMIs to reach consumers. The traditional focus on mass-market products has given way to increased demand for personalised, niche and specialised offerings that cater to diverse consumer needs and preferences (Kenang & Kasetty 2024; Singh, Telukdarie & Mongwe 2024; Spörl-Wang et al. 2025). This shift reflects growing consumer desire for products tailored to individual skin types and tones, reflecting different cultural contexts. Social media influencers have emerged to fill this gap by providing authentic product demonstrations, honest reviews, and relatable application scenarios that help consumers make more confident purchase decisions (Research and Markets 2024). This has prompted brands to continuously search for innovative ways to reach consumers and use social media to do so (Singh et al. 2024).

Social media is used as the main source of information when consumers require guidance in making a purchasing decision (Moodley & Machela 2023). This is because consumers regard SMI posts on social media as trustworthy and more authentic than a brand’s claim. Moodley and Machela (2023) concurred, stating that 60% of consumers in Africa purchase SMI-suggested products because they trust the content more. However, despite the use of influencers, there is a limited understanding relating to how SMIs should be utilised effectively. Current research predominantly centres on developed market contexts, particularly North America and Europe, with a limited focus on emerging markets like South Africa, which may exhibit different cultural values, economic conditions and technological adoption patterns (Singh et al. 2024; Spörl-Wang et al. 2025; Munaro et al. 2025). This geographic bias in existing research creates significant limitations for understanding SMI marketing effectiveness across diverse cultural and economic contexts. The assumptions, theoretical relationships, and practical recommendations derived from developed market research may not translate directly to emerging market contexts where consumer behaviour patterns, technological infrastructure, and cultural norms differ significantly. This limitation is particularly problematic given the rapid growth of social media adoption and digital marketing investment in emerging markets worldwide, such as South Africa. Additionally, existing research demonstrates a pronounced bias towards examining peripheral routes to persuasion, focusing primarily on source attractiveness, credibility and social influence factors, whilst neglecting central routes involving systematic evaluation of argument quality and information content (Kenang & Kasetty 2024; Liu & Gao 2025). Therefore, the predominant focus on peripheral factors reflects broader trends in influencer marketing research that have emphasised the social and emotional aspects of SMI influence, whilst paying insufficient attention to the cognitive and rational processes that may also drive consumer decision-making.

The information adoption model (IAM) (Sussman & Siegal 2003) offers a more comprehensive framework for understanding how individuals evaluate and adopt information from online sources (Cheung, Lee & Rabjohn 2008). The model’s integration of both central (systematic evaluation of argument quality) and peripheral (reliance on source characteristics and contextual cues) routes provides a more complete theoretical foundation for examining SMI influence processes. The model’s core proposition suggests that information adoption is influenced by both central (systematic evaluation of argument quality, including relevance and comprehensiveness) and peripheral (reliance on source characteristics, including credibility and attractiveness) routes. This dual-route approach acknowledges that consumers may engage in careful, systematic evaluation of SMI content or rely on heuristic cues depending on their motivation, ability and opportunity to process information thoroughly (Zhang & Watts 2008).

The central route to information adoption emphasises the importance of argument quality, including the relevance of information to consumer needs and the comprehensiveness of coverage provided (Sussman & Siegal 2003). In SMI marketing contexts, argument quality translates to the extent to which influencer content addresses specific consumer concerns, provides useful information for decision-making, and offers complete coverage of relevant product attributes and considerations (Nadlifatin et al. 2022). The peripheral route emphasises source characteristics that serve as heuristic cues for information quality and reliability (Rahim et al. 2015). Source credibility, including expertise and trustworthiness, offers consumers shortcuts for evaluating information quality when they lack the motivation or ability to engage in systematic evaluation. Source attractiveness, including similarity and likeability, influences information adoption through affective and identification mechanisms that operate independently of rational evaluation processes (Argyris et al. 2020).

The IAM’s emphasis on information adoption as a mediating mechanism between source and content factors and behavioural intentions provides important insights into the psychological processes underlying SMI influence (Zhang & Watts 2008). Understanding these intermediate processes can inform more sophisticated marketing strategies and measurement approaches that track the complete conversion funnel from exposure through adoption to behavioural intention (Erkan & Evans 2018). Furthermore, the rapid evolution of social media platforms, changing algorithm dynamics and shifting user behaviour patterns create additional complexity for brands seeking to develop effective SMI marketing strategies (Djafarova & Rushworth 2017). The need to adapt continuously to platform changes, whilst maintaining authentic relationships with influencers and audiences, requires sophisticated strategic capabilities that many brands continue to refine (Spörl-Wang et al. 2025; Munaro et al. 2025). Whilst some established brands have developed robust influencer marketing frameworks, the pace of platform innovation means even experienced marketers must continuously adapt their approaches. Consumers have become more sophisticated in recognising sponsored content and may discount influencer recommendations that appear overly commercial or inauthentic (Ilieva et al. 2024). This study extends the traditional IAM by incorporating source attractiveness constructs that have been less extensively examined in previous applications of the model. The addition of similarity and likeability as predictors of perceived usefulness acknowledges the unique characteristics of social media environments where parasocial relationships and emotional connections play crucial roles in information evaluation and adoption (Djafarova & Rushworth 2017). The similarity construct captures the extent to which consumers perceive SMIs as resembling themselves in terms of demographics, values, preferences, or lifestyle characteristics. This perceived similarity may enhance information relevance and adoption through identification mechanisms that make SMI experiences and recommendations more personally applicable and trustworthy (Djafarova & Rushworth 2017). The likeability construct reflects the affective responses that consumers develop towards SMIs based on personality characteristics, communication style and overall appeal. Likeability may influence information adoption through mood congruence effects and positive association mechanisms that enhance receptivity to SMI communications independently of rational evaluation processes (Djafarova & Rushworth 2017).

The integration of these attractiveness constructs with traditional credibility and argument quality factors provides a more comprehensive understanding of SMI influence mechanisms. This theoretical enhancement is imperative for understanding SMI marketing effectiveness in contexts where emotional and social factors may be as important as rational and informational considerations (Cheung et al. 2008; Erkan & Evans 2018). Therefore, the study contributes to theory by, first, extending the IAM through the incorporation of source attractiveness constructs, such as likeability and perceived similarity – factors that have received limited attention in previous IAM applications, but are central in social media environments characterised by parasocial interaction and emotional resonance (Djafarova & Rushworth 2017). Second, this study advances the understanding of dual-route information processing by simultaneously exploring central (argument quality) and peripheral (credibility and attractiveness) routes to information adoption. This integrative approach allows for a richer explanation of how rational and affective processes interact in shaping consumer behaviour within digital influencer ecosystems (Zhang & Watts 2008).

From a practical standpoint, the findings provide marketers with evidence-based recommendations for optimising influencer campaign design. In particular, identifying which SMI characteristics most significantly impact information adoption can assist brands in influencer selection and content development to maximise engagement and conversion (Nadlifatin et al. 2022). Moreover, the study contributes to marketing measurement practices by proposing intermediate variables, such as perceived usefulness and information adoption, as more diagnostic metrics than traditional surface-level engagement indicators (e.g. likes or shares). These variables offer a deeper understanding of how consumer attitudes and intentions form over time in response to influencer content (Erkan & Evans 2018).

Literature review and hypothesis development

This section presents the theoretical foundation and hypotheses development for the study. First, the IAM is discussed as the overarching theoretical framework. Subsequently, the central route factors of argument quality, specifically relevance and comprehensiveness, are reviewed and hypothesised. This is followed by a discussion of peripheral route factors, namely source credibility (expertise and trustworthiness) and source attractiveness (similarity and likeability). Finally, the relationships amongst perceived usefulness, information adoption and purchase intention are examined, culminating in the full set of hypotheses.

Information adoption model

The IAM provides a robust theoretical framework for understanding how individuals evaluate and adopt information in online environments (Sussman & Siegal 2003). The model addresses critical limitations of its predecessor theories by integrating elements from both the technology acceptance model (TAM) and the elaboration likelihood model (ELM). The IAM provides a more comprehensive framework – whilst the TAM focuses primarily on technology acceptance, it overlooks social processes and cannot explain why different individuals are influenced differently by the same message (Liao et al. 2022), and the ELM fails to incorporate information usefulness in the influence process (Sussman & Siegal 2003). As discussed above, the IAM’s dual-route approach acknowledges that consumers process information through central (systematic evaluation of argument quality, including relevance and comprehensiveness) or peripheral routes, relying on heuristic cues drawn from source characteristics, specifically source credibility (expertise and trustworthiness) and source attractiveness (similarity and likeability), as examined in the present study. As Sussman and Siegal (2003) explained, when individuals receive influential information, they engage in different levels of cognitive processing depending on their motivation and ability. This theoretical flexibility makes the IAM suitable for examining SMI marketing effectiveness, where consumers navigate between careful content evaluation and quick judgements based on source characteristics. Applications of the IAM in digital contexts support its validity. Cheung et al. (2008) found that information usefulness is a critical determinant of adoption behaviour, whilst Erkan and Evans (2018) successfully extended the model to social media and digital contexts, demonstrating strong relationships between perceived usefulness and purchase intentions.

Argument quality

Building on the IAM framework, argument quality constitutes the central route through which consumers systematically evaluate the persuasive strength of information. In social media environments characterised by information overload, argument quality becomes crucial for differentiating valuable content from the noise on social media (Kenang & Kasetty 2024). Argument quality represents the central route to information processing and encompasses the persuasive strength of information within messages. In social media environments characterised by information overload, argument quality becomes crucial for differentiating valuable content from the noise on social media (Kenang & Kasetty 2024). This distinction is particularly important in the beauty and personal care industry, where consumers actively seek reliable information to minimise purchase risks and enhance their social status by purchasing beauty and personal care products that are socially acceptable (Singh et al 2024).

Additionally, consumer behaviour theory supports the importance of argument quality. The consumer decision-making process involves information search as a critical stage, where individuals consult various sources, including reference groups like SMIs (eds. Erasmus & Mpinganjira 2020). As mentioned, research indicates that 36% of beauty consumers specifically seek online information through SMI tutorials and product reviews before making purchases (Mya et al. 2025), highlighting the significant role of information quality in purchase decisions.

Relevance and perceived usefulness

Information relevance refers to the degree to which content addresses consumers’ specific needs and pain points (Nadlifatin et al. 2022). The theoretical foundation for this relationship draws from social learning theory, which suggests that individuals are more likely to learn from and be influenced by sources providing information aligned with their specific needs (Bandura & Walters 1977). When SMIs provide relevant information that resonates with consumer requirements, they capture attention and enhance their persuasive potential.

The importance of relevance extends beyond mere attention capture. Research demonstrates that personalised, relevant information facilitates efficient decision-making by reducing the time and cognitive effort required for product evaluation (Shim & Jo 2020; Wang, Huang & Davison 2021). In the beauty industry context, relevant information encompasses specific details about ingredients, product applications for different skin types, and solutions to particular beauty concerns. Research confirms that relevance significantly impacts consumers’ perceptions of information usefulness, particularly in contexts where personal application is important (Matute, Polo-Redondo & Utrillas 2016; Shim & Jo 2020; Wang et al. 2021). Collectively, these studies demonstrate that when SMI content directly addresses individual consumer needs and concerns, perceived usefulness is substantially enhanced. Consequently, the following hypothesis is proposed:

H1: Relevance has a significant positive relationship with the perceived usefulness of SMI information.

Comprehensiveness and perceived usefulness

Comprehensiveness encompasses the completeness and breadth of information provided about products or services. This dimension of argument quality ensures that consumers receive sufficient detail to make informed decisions. Comprehensive information increases consumer understanding and awareness of product benefits, whilst incomplete information may lead to misinterpretation and reduced utility (Saranya & Joji 2024).

However, the relationship between comprehensiveness and usefulness is complex in digital environments. Whilst detailed information satisfies consumer informational needs, excessive detail may overwhelm processing capacity, particularly in mobile-first contexts. Therefore, factual objective information reduces ambiguity, but the optimal level of detail varies by context and consumer preferences (Kim & Lee 2015). Within the personal care and beauty industry, consumers rely on the comprehensive nature of the information shared to identify new trends and remain abreast with the latest brands and products (Mya et al. 2025). Therefore, the following hypothesis is proposed:

H2: Comprehensiveness has a significant positive relationship with the perceived usefulness of SMI information.

Source credibility

Whilst argument quality addresses the content of information, the IAM’s peripheral route emphasises source characteristics as heuristic cues that consumers rely on when systematic processing is limited. Source credibility represents a peripheral cue that consumers utilise when lacking motivation or ability to systematically process information (Liu & Gao 2025). In online environments where sensory evaluation of products is limited, credibility becomes particularly crucial (Handoyo 2024; Lin, Jan & Chuang 2019). The theoretical foundation for credibility’s importance draws from social exchange theory, which posits that consumers engage with SMIs to maximise benefits and minimise risks in purchase decisions (O’Reilly et al. 2018).

Credibility comprises multiple dimensions, with expertise and trustworthiness consistently identified as primary components. These dimensions work simultaneously to reduce perceived risk and uncertainty, providing consumers with confidence in their decision-making processes (Rahim et al. 2015). Recent research by Low, Goh and Lim (2025) confirms that source credibility remains a critical factor in fashion and beauty contexts, directly influencing brand loyalty and purchase intentions.

Expertise and perceived usefulness

Within the IAM framework, source credibility functions as a peripheral heuristic cue, and expertise constitutes its primary operational dimension in the context of SMI marketing. Expertise represents the extent to which SMIs possess relevant knowledge and experience in their domains, serving as a security signal for consumers and reducing uncertainty in decision-making (Rahim et al. 2015). Social learning theory provides theoretical support for the expertise-usefulness relationship, suggesting that individuals preferentially learn from competent and knowledgeable sources (Bandura & Walters 1977).

In beauty and personal care contexts, expertise manifests through demonstrated product knowledge, application techniques, and understanding of diverse skin types and concerns. Studies by Wang et al. (2022), Chen, Chen and Pan (2024) and Cheng et al. (2024) confirm that perceived expertise significantly influences information adoption. Specifically, Wang et al. (2022) found that expert SMIs generate higher engagement and conversion rates, whilst Chen et al. (2024) demonstrated that expertise moderates the relationship between product attributes and purchase intentions. Cheng et al. (2024) further established that expertise credibility directly impacts consumer responses in live-streaming situations. Considering the above, it is hypothesised that:

H3: Expertise has a significant positive relationship with the perceived usefulness of SMI information.

Trustworthiness and perceived usefulness

Trustworthiness reflects consumers’ beliefs in the honesty, integrity and reliability of SMI communications (Rahim et al. 2015). This dimension assumes particular importance in reducing perceived uncertainty and risk associated with online purchases (Ilieva et al. 2024). Social media influencers cultivate trustworthiness through consistent, transparent communication and authentic relationship building with their audiences (Yuxuan 2025). The South African context adds unique dimensions to trustworthiness concerns. With SMI marketing still maturing, characterised by limited regulatory oversight, inconsistent disclosure practices for sponsored content, and the prevalence of fraud and fake followers, trust becomes a critical differentiating factor. A recent comprehensive meta-analysis by Spörl-Wang et al. (2025) found that trustworthiness consistently predicts SMI marketing effectiveness across cultures. Masuda, Han and Lee (2022) demonstrated that trustworthiness mediates the relationship between influencer characteristics and purchase intentions, highlighting its central role in the influence process. Accordingly, it is proposed that:

H4: Trustworthiness has a significant positive relationship with the perceived usefulness of SMI information.

Source attractiveness

Beyond credibility, source attractiveness constitutes an additional peripheral cue within the IAM framework, encompassing the appeal and likeability of SMIs, functioning as a peripheral cue when consumers engage in heuristic processing (Sokolova & Kefi 2020). Attractive sources demonstrate enhanced persuasive power within digital contexts. The construct’s relevance is amplified in beauty and personal care industries where product categories naturally align with appearance and attractiveness concepts (Lili et al. 2022; Patel & Basil 2018). Research suggests that micro-celebrity influencers leverage attractiveness differently than mainstream celebrities, emphasising relatability over aspirational appeal (Spörl-Wang et al. 2025).

Similarity and perceived usefulness

Similarity represents the degree to which consumers perceive SMIs as resembling themselves in demographics, values, preferences or lifestyle characteristics (Martensen, Brockenhuus-Schack & Zahid 2018). Social identity theory provides theoretical grounding, suggesting that individuals show a preference for in-group members and find their communications more persuasive (Argyris et al. 2020). This perceived similarity enhances relatability and authenticity, distinguishing SMIs from traditional celebrities (Santiago & Serralha 2022; Scholz 2021). Spörl-Wang et al. (2025) suggest that SMIs who share personal experiences and interact directly with followers create stronger similarity perceptions than celebrities. Similarity encourages consumers to imitate SMIs’ behaviour. Hence, it is hypothesised:

H5: Similarity has a significant positive relationship with the perceived usefulness of SMI information.

Likeability and perceived usefulness

Likeability encompasses consumer admiration for SMIs based on physical and personality characteristics (Martensen et al. 2018). Individuals typically seek to emulate admired figures through similar consumption patterns. This emulation process operates through ideal self-congruity, where consumers perceive likeable SMIs as aspirational representations of their desired selves (Xu & Pratt 2018).

Social learning theory reinforces this relationship by emphasising that attention is enhanced when observers find models attractive and worthy of imitation. Physical attractiveness significantly influences behavioural intentions, and likeability drives product adoption through identification processes. Likeable SMIs fulfil consumer needs for information and aspiration, making their content particularly valuable (Ki et al. 2020). Thus, the following hypothesis has been developed:

H6: Likeability has a significant positive relationship with the perceived usefulness of SMI information.

Perceived usefulness, information adoption and purchase intention

Perceived usefulness represents consumers’ evaluations of whether SMI information will enhance their decision-making and life outcomes (Erkan & Evans 2018). According to the IAM framework, perceived usefulness serves as the critical mechanism linking information characteristics to behavioural outcomes (Sussman & Siegal 2003). This relationship is supported by rational choice perspectives, suggesting that individuals adopt information based on cost-benefit evaluations (Erkan & Evans 2018).

Information adoption constitutes an internal cognitive process whereby consumers accept and internalise SMI communications as valid and actionable (Zhang & Watts 2008). This process involves not merely receiving information, but actively incorporating it into decision-making frameworks. Purchase intention represents the culmination of the influence process, reflecting consumers’ willingness and plans to acquire recommended products (Cheung et al. 2008). Munaro et al. (2025) found that information adoption mediates influencer effects across product categories, whilst Al-Muani et al. (2023) demonstrated that these relationships hold in emerging market contexts. In the African context, Ezenwafor, Olise and Ebizie (2021), in a study of social media users in a developing economy, reported that 60% of consumers had purchased products based on SMI recommendations, with over 50% expressing greater trust in influencer content compared to brand communications. Whilst these figures are context-specific, they align with broader patterns of SMI influence documented in emerging market research. Consequently, it is hypothesised:

H7: The perceived usefulness of SMI information has a significant positive relationship with information adoption.

H8: Information adoption from SMIs has a significant positive relationship with consumer purchase intention.

Research methods and design

Research design

This study adopted a quantitative research approach, employing a descriptive cross-sectional survey design. The quantitative approach was deemed appropriate for testing the hypothesised relationships between constructs and enabling the generalisation of findings to the broader population (Kline 2016). The cross-sectional design facilitated data collection at a single point in time, which is suitable for examining consumer attitudes and behavioural intentions (Malhotra 2019).

Population and sampling

The target population comprised active social media users residing in South Africa, operationalised as individuals who accessed social media platforms at least once a month via any digital device. This aligns with established social media usage metrics (Statista 2020). In addition, the inclusion criteria specified South African residents aged 18–65 years who followed at least one beauty or personal care SMI.

Non-probability snowball sampling was employed to recruit participants. Whilst this approach introduces potential selection bias, it was justified by the need to access a specific population of social media users engaged with beauty influencer content. Initial participants were recruited through social networks on Instagram, Facebook and LinkedIn, with subsequent recruitment occurring through participant referrals. The minimum sample size was calculated using the Kline (2016) guidelines for structural equation modelling (SEM), suggesting 10–20 observations per estimated parameter. With 45 parameters in the measurement model, a minimum of 225 participants was required. The final sample comprised 242 valid responses, exceeding this threshold.

Measurement instrument

A structured questionnaire was developed incorporating validated scales from prior research, with all constructs measured using five-point Likert-type scales (1 = ‘strongly disagree’, 5 = ‘strongly agree’) to ensure consistency and facilitate statistical analysis. The measurement of argument quality was operationalised through two dimensions: relevance (six items) and comprehensiveness (five items), adapted from Cheung et al. (2008) and Erkan and Evans (2018). Source credibility comprised expertise (six items) and trustworthiness (six items), with items drawn from Hu, Zhang and Wang (2019); Ki et al. (2020); Park and Lin (2020); and Xiao, Wang and Chan-Olmsted (2018). Source attractiveness was measured through similarity (seven items) and likeability (nine items), adapted from Ki et al. (2020), Sokolova and Kefi (2020), and Xiao et al. (2018). The dependent variables, namely perceived usefulness (five items), information adoption (six items), and purchase intention (eight items), were adapted from Cheung et al. (2008), Erkan (2016), and Sokolova and Kefi (2020), respectively. The questionnaire underwent pretesting with 30 respondents to assess face validity and comprehension, with minor modifications to item wording implemented based on feedback.

Data collection procedures

The data were collected using Google Forms. The online format was appropriate given the target population’s digital engagement. As mentioned, the questionnaire link was distributed through social media platforms with instructions for snowball recruitment. Participants provided informed consent before accessing the questionnaire, and no incentives were offered to minimise response bias.

Data analysis

Data analysis employed a two-stage SEM approach using SPSS 27 and AMOS 27 (Anderson & Gerbing 1988). Confirmatory factor analysis evaluated the measurement model’s psychometric properties. Model fit was assessed using multiple indices: Chi-square/degrees of freedom ratio (χ²/df < 3), root mean square error of approximation (RMSEA < 0.08), comparative fit index (CFI > 0.90), Tucker-Lewis index (TLI > 0.90), goodness-of-fit index (GFI > 0.90) (Hu & Bentler 1999). Reliability was assessed through Cronbach’s alpha (α > 0.70) and composite reliability (CR > 0.70). Convergent validity was established through average variance extracted (AVE > 0.50) and factor loadings (> 0.70). Discriminant validity was evaluated using the Fornell-Larcker criterion (Fornell & Larcker 1981; Motara 2022). The structural model examined the hypothesised relationships (H1–H8) between the constructs. Path coefficients and their significance were evaluated using maximum likelihood estimation. The model fit indices identical to those used for the measurement model were applied.

Ethical considerations

Ethical clearance to conduct this study was obtained from the University of Johannesburg School of Consumer Intelligence and Information Systems, College of Business and Economics Ethics Committee (Ref. No. 2021SCiiS005). Participants were informed of the study’s purpose, their voluntary participation, their right to withdraw, data confidentiality, and that personal identifiers would not be collected to ensure anonymity.

Results

This section presents the results of the study.

Sample characteristics

The study’s final sample (n = 242) was predominantly female (86.4%), reflecting the gender distribution typical of beauty-focused social media engagement. The majority of respondents were aged 18–29 years (79.3%), with representation from multiple ethnic groups: Indian people (48.8%), Coloured people (25.6%), black African people (16.9%) and white people (6.6%). Educational attainment was relatively high, with 71.0% holding tertiary qualifications.

Measurement model evaluation

The measurement model demonstrated acceptable fit: χ²/df = 2.045, RMSEA = 0.066 (90% confidence interval [CI]: 0.062–0.070), CFI = 0.929, TLI = 0.923, GFI = 0.742, AGFI = 0.707. Whilst GFI and AGFI fell slightly below the 0.90 threshold, other indices indicated satisfactory fit (Schreiber et al. 2006). All constructs (Table 1) exhibited satisfactory reliability, with Cronbach’s α ranging from 0.922 to 0.973 and CR values from 0.923 to 0.973 (Motara 2022). It is acknowledged that Cronbach’s α values exceeding 0.90 may indicate potential item redundancy or high semantic overlap amongst items (Tavakol & Dennick 2011). However, this should be interpreted in the context of the present study for the following reasons. First, the constructs under investigation (expertise, trustworthiness, likeability, and perceived usefulness) are theoretically narrow and unidimensional by design, and high inter-item correlations are therefore an expected property of well-constructed unidimensional scales rather than evidence of redundancy (Tavakol & Dennick 2011). Second AVE values ranged from 0.669 to 0.878 – all exceeding the minimum threshold of 0.50, and factor loadings ranged from 0.685 to 0.955. Had items been semantically identical, AVE values would be expected to approach 1.0; the present values indicate that items capture meaningfully differentiated variance within each construct. Third, the measurement items were adapted from multiple distinct validated instruments across prior studies (Cheung et al. 2008; Ki et al. 2020; Sokolova & Kefi 2020; Xiao et al. 2018), providing conceptual breadth within each construct. Consequently, whilst the potential for item overlap cannot be entirely dismissed, the convergent validity evidence supports the conclusion that the scales demonstrate strong integrity.

Discriminant validity was established, as the square root of AVE for each construct exceeded its correlations with other constructs, as shown in Table 2. The highest inter-construct correlation (r = 0.841) occurred between expertise and trustworthiness, which was expected given their conceptual relationship as credibility dimensions.

Structural model results

The structural model demonstrated acceptable fit: χ²/df = 1.883, RMSEA = 0.061 (90% CI: 0.057–0.065), CFI = 0.941, TLI = 0.935, GFI = 0.762, AGFI = 0.726 (Motara 2022). As with the measurement model, the GFI (0.762) and AGFI (0.726) fell meaningfully below the 0.90 threshold, which is attributed to the sensitivity of these indices to model complexity and sample size (Schreiber et al. 2006). The χ²/df ratio (1.883 < 3.0), RMSEA (0.061, below the 0.08 threshold), CFI (0.941 > 0.90), and TLI (0.935 > 0.90) collectively indicate an acceptable model fit, supporting the structural model’s representation of the hypothesised relationships. The model explained substantial variance in the endogenous constructs: perceived usefulness (R² = 0.72), information adoption (R² = 0.71) and purchase intention (R² = 0.73).

Discussion

This study successfully extended the IAM by incorporating source attractiveness constructs and validating its effectiveness in the South African beauty and personal care market. The findings reveal several critical insights that support and challenge existing theoretical frameworks and industry practices.

Key findings

The most significant finding is expertise emerging as the strongest predictor of perceived usefulness (β = 0.411, p < 0.001) (Table 3), surpassing all other factors, including traditional attractiveness measures. This finding strongly supports social learning theory’s proposition that individuals preferentially learn from competent and knowledgeable sources (Bandura & Walters 1977) and aligns with recent research (Chen et al. 2024; Wang et al. 2022) demonstrating expertise’s critical role in digital influence contexts. However, this finding challenges prevailing industry practices that prioritise reach metrics, such as follower counts and engagement rates, over domain competence. The dominance of expertise over attractiveness factors contradicts the common assumption that beauty and personal care marketing rely primarily on aspirational and aesthetic appeals. Instead, it suggests that even in visually orientated industries, consumers prioritise credible information when making purchase decisions.

Perhaps the most theoretically challenging finding is comprehensiveness’s non-significant relationship with perceived usefulness (β = −0.156, p = 0.194), directly contradicting information richness theory’s core proposition that more complete information enhances decision-making quality (Daft & Lengel 1986). This paradox suggests that the traditional assumption of ‘more information equals better decisions’ may not hold in attention-constrained social media environments. This finding diverges from previous IAM applications by Cheung et al. (2008) and Erkan and Evans (2018), where comprehensiveness typically showed positive relationships with information adoption. The contradiction can be explained through three theoretical lenses: cognitive load theory, suggesting information overload in mobile-first contexts (Sweller 2011); uses and gratifications theory, indicating preference for efficient, targeted content; and cultural communication preferences in high-context cultures, favouring focused messaging over exhaustive detail (Katz, Blumler & Gurevitch 1973).

The study validates dual-route information processing in the South African context, with both central (relevance: β = 0.341, p = 0.002; expertise: β = 0.411, p < 0.001) and peripheral (similarity: β = 0.159, p = 0.001; likeability: β = 0.110, p = 0.026; trustworthiness: β = 0.265, p = 0.019) route factors significantly predicting perceived usefulness. This finding supports the IAM’s theoretical foundation, whilst extending its applicability to emerging market contexts (Motara 2022). The dominance of central route factors challenges assumptions that emerging market consumers rely primarily on heuristic processing because of lower information literacy or technological sophistication. Instead, South African consumers appear to engage in systematic evaluation of SMI content, particularly for discretionary purchases like beauty products. This aligns with the Singh et al. (2024) research on South African digital consumer behaviour, but contradicts broader emerging market stereotypes.

The significant but modest effects of similarity and likeability empirically support social identity theory’s application in digital contexts (Argyris et al. 2020). These findings validate theoretical arguments that social media influence operates through identification mechanisms beyond traditional credibility assessments (Djafarova & Rushworth 2017). However, the relatively small effect sizes compared to expertise and relevance suggest that emotional and relational factors, whilst important, serve complementary rather than primary roles in purchase-related decisions. This finding extends the Ki et al. (2020) research on parasocial relationships by demonstrating that their contextual importance varies based on decision-making scenarios and cultural contexts.

The strong relationships between perceived usefulness and information adoption (β = 0.846, p < 0.001), and between information adoption and purchase intention (β = 0.855, p < 0.001), validate the IAM’s theoretical proposition that information adoption mediates the relationship between source and/or content factors and behavioural outcomes (Motara 2022). This finding supports the Zhang and Watts (2008) conceptualisation of information adoption as an active cognitive process and extends the Munaro et al. (2025) research by demonstrating consistent mediation effects across cultural contexts.

Strengths and limitations of the research

This research demonstrates several methodological and theoretical strengths that enhance confidence in the findings. The rigorous analytical approach using SEM with established fit indices (χ²/df = 1.883, RMSEA = 0.061, CFI = 0.941) and comprehensive reliability assessments (Cronbach’s α = 0.922–0.973) ensures robust statistical foundations. The theoretical integration of IAM with source attractiveness constructs provides a more comprehensive framework for understanding SMI influence mechanisms than previous studies that examined these factors in isolation. The emerging market context addresses a significant gap in existing literature, which has predominantly focused on developed markets, particularly North America and Europe. This geographical contribution is particularly valuable, given the rapid growth of social media adoption and digital marketing investment in emerging markets worldwide. The beauty and personal care industry focus is strategically appropriate, given this sector’s high SMI engagement rates and visual nature that align with social media platform characteristics. The study’s sample size (n = 242) exceeds minimum requirements for SEM, whilst the model’s substantial explanatory power (R² > 0.70 for key constructs) demonstrates strong predictive validity. The use of established, validated measurement scales from multiple sources enhances construct validity and facilitates comparison with previous research.

Despite these strengths, several limitations constrain the generalisability and interpretation of results. The non-probability snowball sampling approach, whilst appropriate for accessing specific populations of SMI followers, introduces potential selection bias and limits the representativeness of the broader South African population. The sample’s demographic concentration amongst young female respondents (86.4% female, 79.3% aged 18–29) may not reflect diverse consumer behaviour patterns across age groups, genders, and socioeconomic segments. The cross-sectional research design precludes causal inference and fails to capture the dynamic evolution of SMI-follower relationships over time. Social media influencer influence likely develops through repeated exposure and relationship building, which are processes that single-point-in-time measurements cannot adequately assess. Additionally, the study’s focus on the beauty and personal care industry limits generalisability to other product categories where different influence mechanisms may operate. The study does not account for platform-specific differences in SMI influence mechanisms, as consumer behaviour may vary across Instagram, TikTok, YouTube, and other social media platforms. The measurement of SMI characteristics relied on consumer perceptions, rather than objective assessments, introducing potential response bias. Social desirability bias may have influenced responses, particularly regarding purchase intentions and information adoption behaviours. Finally, the study does not examine moderating factors, such as product involvement, purchase experience, or individual differences in information processing preferences that may influence the observed relationships.

Implications and recommendations

This research opens several avenues for future investigation that could significantly advance understanding of social media influence mechanisms. The comprehensiveness paradox challenges traditional information richness theory and necessitates research examining the curvilinear relationship between information quantity and perceived usefulness across different platforms, demographics and cultural contexts to identify optimal information load thresholds. The successful integration of source attractiveness constructs into the IAM demonstrates the need for more comprehensive theoretical frameworks, suggesting future studies should explore additional relational factors, such as authenticity, responsiveness and community engagement, that may influence information adoption in social media contexts. Longitudinal research is critical for understanding how SMI influence develops over time, requiring studies that track relationship evolution through repeated exposure, content consistency and interactive engagement, whilst employing experimental designs that manipulate specific SMI characteristics to provide stronger causal evidence. Cross-cultural validation of the extended IAM is essential, particularly examining how cultural values, communication preferences, and technological infrastructure moderate observed relationships to determine whether the expertise-dominance finding represents universal patterns or context-specific phenomena. Additionally, future research should investigate platform-specific variations in influence mechanisms, examine moderating factors, such as product involvement and individual processing preferences, and develop more sophisticated measurement approaches that capture the dynamic nature of parasocial relationships in digital environments.

The study’s findings suggest fundamental shifts in influencer marketing strategy, moving from reach-based to competence-based approaches that prioritise expertise over follower metrics. Consequently, the following recommendations are made from the findings of the study:

• Brands should develop comprehensive expertise assessment frameworks for influencer selection. These frameworks should evaluate domain knowledge depth through content analysis, track record assessment, and audience feedback evaluation. Brands must prioritise partnerships with beauty professionals, make-up artists, dermatologists and skincare specialists over general lifestyle influencers, even when the latter have larger follower bases.
• Content strategies should shift from comprehensive product presentations to focused, insight-driven approaches. Brands should develop modular content frameworks that help influencers deliver targeted solutions to specific consumer problems, rather than exhaustive product coverage. This approach aligns with short-form content trends and attention-span constraints, whilst maximising information value.
• Portfolio approaches should balance different influencer types to leverage both expertise and attractiveness factors. Brands must engage expert influencers for credibility and information provision, whilst incorporating relatable influencers for emotional connection and brand affinity. This strategy recognises that different influencers serve varying functions across the consumer decision journey.
• Market-entry strategies should prioritise credibility building through partnerships with recognised industry professionals and educational content development. Brands must invest in content that addresses emerging market-specific concerns, such as product authenticity, skin-type suitability, cultural appropriateness and value for money, rather than focusing solely on brand promotion.
• Trust-building initiatives should address the unique challenges of emerging markets, including concerns about product quality, delivery reliability and customer service. Influencer partnerships should emphasise transparency, honest reviews and authentic product experiences, rather than purely promotional content.
• Brands should develop platform-specific content guidelines that optimise the expertise-relevance combination for different social media environments. Short-form platforms like TikTok may require highly focused, problem-solution content, whilst longer-form platforms like YouTube can accommodate more detailed educational content without overwhelming audiences.
• Cross-platform influencer strategies should leverage platform strengths, using expertise-focused content for information-seeking contexts, whilst incorporating attractiveness factors for entertainment-oriented platforms. This approach recognises that consumer motivations and information processing preferences vary across digital environments.

Conclusion

This study significantly advances understanding of SMI marketing effectiveness by extending the IAM to include attractiveness constructs and validating its application in an emerging market context. The findings challenge prevailing industry practices and established theoretical assumptions, particularly regarding the value of comprehensive information in digital environments.

The theoretical contributions include demonstrating the importance of relational factors in information adoption, whilst confirming the primacy of competence and relevance in purchase-related contexts. The practical implications suggest fundamental shifts in influencer selection criteria and content strategy approaches, with particular relevance for brands operating in emerging markets.

Most importantly, the study reveals that effective SMI marketing requires balancing multiple influence pathways, whilst adapting to platform constraints and cultural contexts. The dominance of expertise over attractiveness factors suggests that competence-based approaches may be universally important, whilst the comprehensiveness paradox indicates that ‘more information’ strategies may be counterproductive in attention-scarce digital environments.

These insights provide the foundation for theoretical development and practical strategy refinement, contributing to a more nuanced understanding of how information adoption processes operate in contemporary digital marketing contexts. The research demonstrates that emerging markets offer valuable contexts for testing and extending marketing theories, whilst revealing important cultural and economic factors that shape consumer behaviour in globally diverse digital environments.

Acknowledgements

The authors acknowledge the participants who contributed their time to complete the survey questionnaire and the College of Business and Economics at the University of Johannesburg for providing ethical clearance and institutional support for this research.

This article includes content that overlaps with research originally conducted as part of Farzaana Motara’s master’s dissertation entitled, ‘Effects of social media influencers on consumer purchase intentions of beauty products’, submitted to the College of Business and Economics, University of Johannesburg in 2022. The dissertation was supervised by Christine De Meyer-Heydenrych and Nicole Cunningham. Portions of the data, analysis, and/or discussion have been revised, updated, and adapted for journal publication. The original dissertation is publicly available at: https://hdl.handle.net/10210/504201. The authors affirm that this submission complies with ethical standards for secondary publication, and appropriate acknowledgement has been made of the original work.

The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

Farzaana Motara: Conceptualisation, Data curation, Methodology, Writing – original draft. Nicole Cunningham: Conceptualisation, Methodology, Software, Supervision, Writing – original draft. Christine De Meyer-Heydenrych: Conceptualisation, Formal analysis, Methodology, Software, Supervision, Writing – review & editing. All authors reviewed the article, contributed to the discussion of results, approved the final version for submission and publication, and take responsibility for the integrity of its findings.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

The data that support the findings of this study are available from the corresponding author, Christine De Meyer-Heydenrych, upon reasonable request.

The views and opinions expressed in this article are those of the authors and are the product of professional research. It does not necessarily reflect the official policy or position of any affiliated institution, funder, agency, or that of the publisher. The authors are responsible for this article’s results, findings, and content.

References