Target: Jurnal Manajemen Bisnis Volume 8 Issue 1 .
Page 45-56 e-issn: 2715-9361 id/index.
php/target Social Media Influencers and Purchase Intention: The Role of Parasocial Relationships on TikTok Renasabylla Putri Asriliorentika* .
Ratih Amelia Universitas Negeri Surabaya.
Surabaya.
Indonesia Submitted 29th May 2026.
Revised 20th June 2026.
Accepted 23rd June 2026.
Published 28th June 2026 Abstract The rapid growth of IndonesiaAos beauty industry and TikTokAos dominance as a digital marketing platform present both opportunities and challenges for brands seeking to understand how consumer behavior is influenced by influencers.
Although the role of social media influencers in driving purchase intent has been extensively studied, the psychological mechanisms linking the two through parasocial relationships remain underexplored on the TikTok platform in Indonesia.
This study aims to analyze the influence of social media influencers on the intention to purchase beauty products, with parasocial relationships serving as a mediating variable among followers of the TikTok account @catlovers29 in Surabaya.
A quantitative approach using an associative survey method was employed.
Data were collected via an online questionnaire, and after undergoing purposive sampling and data screening, 192 valid respondents were obtained and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS version 4.
The results show that social media influencers have a positive and significant effect on both purchase intention and parasocial relationships, and that parasocial relationships, in turn, have a positive and significant effect on purchase intention.
Parasocial relationships were found to partially mediate the effect of social media influencers on purchase intention.
These findings imply that the emotional bond formed between followers and influencers plays a more dominant role in driving interest in purchasing beauty products than content quality alone does.
therefore, beauty brands need to prioritize selecting influencers capable of building strong parasocial bonds with their follower Keywords: Beauty Products.
Parasocial Relationship.
Purchase Intention.
Social Media Influencer.
TikTok.
Copyright A 2026 Authors.
Published by Universitas Bumigora Under the Creative Commons Attribution 4.
0 International License Corresponding author.
Tel: 62-823-3483-6378.
E-mail: renasabylla.
23067@mhs.
DOI: 10.
30812/target.
How to cite:
Asriliorentika.
, & Amelia.
Social Media Influencers and Purchase Intention: The Role of Parasocial Relationships on TikTok.
Target: Jurnal Manajemen Bisnis, 8 .
, 45-56.
45 | Accredited SINTA 4 .
ttps://sinta.
id/journals/profile/10.
46 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 Introduction IndonesiaAos beauty industry has experienced significant growth in recent years.
According to data from the Ministry of Industry of the Republic of Indonesia for 2024, the sectorAos revenue value increased from USD 8.
09 billion in 2023 to a projected USD 9.
17 billion in 2024.
The number of businesses grew by more than 46 percent over three years, from 819 companies in 2021 to over 1,200 companies by the end of 2024.
This growth encompasses seven categories of beauty products: skincare, sun care, hair care, decorative cosmetics, body care, oral care, and perfumes (Ainapure, 2.
This trend reflects a shift in consumer behavior, with beauty products now considered an essential part of daily life.
Digital platforms play a dominant role in shaping consumer behavior toward beauty products.
According to the Data Reportal Digital 2024 report.
TikTok is the social media platform with the largest user base in Indonesia, reaching 157.
6 million users as of July 2024, with the majority aged 18 to 34.
The TikTok Shop feature, which enables in-app transactions, reinforces the platformAos position as an efficient marketing and transaction space for the beauty industry.
Hung et al.
found that influencer characteristics such as physical attractiveness and trust have a significant positive relationship with consumersAo interest in purchasing cosmetic products on TikTok.
These findings are further supported by Sutiono et al.
, who demonstrate that the credibility dimensions of influencers, including reliability and expertise, significantly influence followersAo purchase intentions on the same platform.
Ao et al.
through a meta-analysis of 62 studies, concluded that the entertainment value of influencer content is the strongest predictor of consumer engagement among all other factors.
Barta et al.
further found that humor-based content among TikTok influencers enhances followersAo hedonic experiences, which in turn strengthens consumer engagement and purchase intent.
A social media influencer (SMI) is an individual who can influence othersAo purchasing decisions through the authority, knowledge, and rapport they have built with their audience on digital platforms (Leung et al.
, 2.
SMIs build their reputation through consistent content, so the messages they convey are often perceived as more authentic and trustworthy by their followers (Ao et al.
, 2.
Liu and Zheng .
identified three main dimensions that determine the effectiveness of SMI content:
informativeness, authenticity, and entertainment.
Cheung et al.
reinforced this by finding that SMI content attributes, particularly the dimensions of prestige and expertise, significantly influence followersAo online brand-related activities including purchase intention.
Sokolova and Kefi .
found that both the credibility and parasocial interaction of beauty influencers on YouTube and Instagram have a positive and significant effect on followersAo purchase intention, where followers who are more parasocially attached to the influencer tend to be more motivated to make a purchase.
Amelia et al.
assert that the effectiveness of digital marketing is not solely determined by how widely content is disseminated, but by the extent of trust and engagement built between the information source and the audience.
A parasocial relationship (PR) is defined as a persistent, long-term emotional bond between a follower and a media persona, even when no active interaction is taking place (Aw & Chuah, 2.
is formed cumulatively through repeated exposure to content and fosters a sense of familiarity, trust, and care toward the influencer, much like in real-life friendships (Yuan & Lou, 2.
PR is measured through three dimensions: affective closeness, trust and utility, and emotional care (Liu & Zheng, 2.
Sokolova and Kefi .
, who studied active social media users, found that social attractiveness and attitudinal similarity are the primary antecedents of PR formation, which significantly drive decisions to purchase beauty products.
Su et al.
concluded that PR is a stronger predictor than trust or expertise alone of consumer purchase intent, as it mediates the direct influence of influencer credibility on purchase intent.
Purnamaningsih et al.
, who studied the context of beauty products in Indonesia, found that parasocial interaction and influencer credibility play a significant role in driving consumer purchase Taher et al.
demonstrated that PR significantly mediates the relationship between influencer reputation signals and purchase intent for beauty products.
Dwiphala and Astuti .
found that among beauty influencers on Instagram in Indonesia.
PR significantly influences both purchase intention and consumersAo electronic word-of-mouth regarding beauty products.
In the context of beauty influencersAo followers in Indonesia.
PR serves as a consistent socio-cognitive pathway linking the quality of SMI content to consumersAo purchase intentions (He et al.
, 2024.
Muhmin et al.
, 2.
Liu and Zheng .
demonstrated that the informative value, authenticity, and shared values displayed by influencers positively influence PR, which in turn drives consumer purchase intention.
Purchase intention is defined as a consumerAos psychological tendency to purchase after evaluating Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
47 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 a product or service (Yao et al.
, 2.
Purchase intention can be measured across four dimensions:
transactional, referential, preferential, and exploratory (Septyadi et al.
, 2.
This study uses the three most relevant dimensions: transactional, referential, and exploratory interest.
Sokolova and Kefi .
demonstrated that parasocial interaction with beauty influencers on video platforms contributes positively and significantly to consumer purchase intention, with the influence of parasocial interaction being even more dominant than credibility among most follower groups.
Previous findings consistently indicate that the quality of SMI content positively affects purchase intent, both directly and through PR mediation, in which the emotional bond between followers and influencers serves as a bridge that transforms content exposure into concrete purchase intent (Rawajbeh et al.
, 2023.
Trisnawati et al.
, 2.
Octabella et al.
, who directly studied the consumer behavior of beauty product users in Surabaya, found that reviews from beauty influencers are a key determinant in the decision-making process for purchasing cosmetic products, particularly among women aged 20 to 40.
Strong parasocial relationships (PR) between followers and influencers not only directly drive purchase intent but also enhance customer equity as a long-term brand asset, while SMI credibility further strengthens PR, which partially mediates this (Almaida et al.
, 2021.
Tripopsakul & Hoonsopon, 2.
Parasocial interaction with highly popular influencers has also been shown to foster positive attitudes toward the brand that drive purchase intent, with PR remaining a consistently significant predictor across various social media platforms (Lin et al.
, 2.
Sokolova and Kefi .
note that research on PR in the context of beauty influencers in Indonesia remains very limited, particularly studies specifically focused on the TikTok platform, where the personalized For You Page (FYP) algorithm reinforces parasocial mechanisms through repeated content exposure that builds familiarity and trust in the influencerAos persona (Horton & Richard Wohl.
Therefore, this study aims to analyze the influence of social media influencers on purchase intention for beauty products, with parasocial relationship as a mediating variable, among followers of the @catlovers29 TikTok account.
Theoretically, this study contributes to the literature on parasocial relationships in the context of short-form video platforms, while, practically, it provides evidence-based guidance for beauty brands to evaluate influencer partnerships by the strength of parasocial bonds with their follower communities, rather than solely on content reach.
II.
Method This study employs a quantitative approach using an associative survey method to test the causal relationships among variables.
The population in this study consists of all TikTok users in Surabaya who follow the @catlovers29 account (Alma Nax Buara.
The minimum sample size was calculated using the 10-times rule formula by Hair and Alamer .
, which is at least 10 times the number of indicators in the most complex model .
y 12 = 120 respondent.
The sampling technique used was purposive sampling with the following inclusion criteria: respondents residing in Surabaya, actively using TikTok, and consuming content from the @catlovers29 account.
From the data collection and screening process, 192 respondents who met all predetermined criteria were selected as the study sample.
The research model identifies Social Media Influencers (SMI) as the independent variable.
Purchase Intention (PI) as the dependent variable, and Parasocial Relationship (PR) as the mediating variable.
The structural model is formulated in two path equations: P R = 1 SM I 1 and P I = 2 SM I 3 P R 2 , where the mediating effect (H.
is estimated via the product of the indirect coefficients, 1 y 3 .
The conceptual framework of the study is shown in Figure 1.
Figure 1.
Conceptual Framework Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
48 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 Each item in the instrument was measured using a 1Ae5 Likert scale.
This research instrument was developed through adaptations of several validated sources, as shown in Table 1.
Table 1.
Research Variables Variables Dimensions References Social Media Influencer (SMI) Parasocial Relationship (PR) Informativeness.
Authenticity.
Entertainment Affective Closeness.
Trust & Utility.
Emotional Care.
Exploratory.
Referential, & Transactional interest (Liu & Zheng, 2.
(Liu & Zheng, 2.
Purchase Intention (PI) (Septyadi et al.
, 2.
Data analysis using PLS-SEM via SmartPLS version 4 included testing of the outer model .
alidity and reliabilit.
and the inner model through bootstrapping of 5,000 subsamples (Hair & Alamer, 2.
All constructs were measured reflectively using the following criteria: outer loading Ou 0.
AVE Ou 0.
HTMT < 0.
CR Ou 0.
and CronbachAos Alpha Ou 0.
70, at a 95% confidence level ( = 0.
Results and Discussion Respondent Profile The demographic profile of 192 TikTok users in Surabaya who follow the @catlovers29 account is presented in Table 2.
The majority of respondents were women, totaling 167 people .
%).
119 people .
0%) were in the productive age range of 21Ae24 years.
and they exhibited a high level of engagement, with the majority, 113 people .
9%), stating they AuoftenAy watch cosmetic content.
This demographic composition is considered highly representative and relevant to the contemporary beauty industryAos focus, while also confirming that the sample has sufficient exposure to content to form a parasocial relationship.
Table 2.
Respondent Profile Characteristic Frequency Percentage Gender Female Male Sum Age < 17 years 17Ae20 years 21Ae24 years 25Ae30 years Sum Content Viewing Frequency Rarely Sometimes Often Very Often Sum Outer Model Test Outer-model testing was conducted to evaluate the measurement instrumentAos validity and reliability through three stages: convergent validity, discriminant validity, and composite reliability.
Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
49 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 Convergent Validity Convergent validity is assessed through the outer loading values of each indicator and the Average Variance Extracted (AVE) values of each construct.
An indicator is considered valid if it has an outer loading of Ou 0.
70, while a construct meets convergent validity if its AVE value is Ou 0.
50 (Hair & Alamer.
Figure 2 presents the results of the PLS-SEM model, along with the outer loading values for all Figure 2.
Model SEM-PLS Validity Test (Outer Loadin.
The test results show that all indicators have outer loadings ranging from 0.
706 to 0.
874, exceeding the minimum threshold of 0.
The outer loadings for each indicator are presented in Table 3.
Table 3.
Loading Factor Validity Test (Outer Loadin.
Variables Item
Loading Factor
Social Media Influencer (SMI) SMI1
SMI2
SMI3
SMI4
SMI5
SMI6
SMI7
SMI8
SMI9
SMI10
SMI11
SMI12
Parasocial Relationship (PR) PR1
PR2
PR3
PR4
PR5
PR6
PR7
Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
50 | Renasabylla Putri Asriliorentika
Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361
Variables Purchase Intention (PI) Item
Loading Factor
PR8
PR9
PR10
PR11
PR12
PI2
PI4
PI5
PI6
PI7
PI8
PI9
Average Variance Extracted (AVE) Validity Test The AVE values for all three constructs exceeded the 0.
50 threshold, indicating that convergent validity was met, as shown in Table 4.
Table 4.
Average Variance Extracted (AVE) Validity Test Variables AVE Description Social Media Influencer (SMI) Parasocial Relationship (PR) Purchase Intention (PI) Valid Valid Valid Discriminant Validity Test Discriminant validity was tested using the Fornell-Larcker criterion by comparing the root-meansquare error of estimation (RMSE) for each construct with its correlation with other constructs (Hair & Alamer, 2.
The results are presented in Table 5.
Table 5.
Discriminant Validity Test Variables SMI Parasocial Relationship (PR) Purchase Intention (PI) Social Media Influencer (SMI) Contruct Reliability Test The square root of the average variances (AVE) on the main diagonal of each construct is higher than its correlation with other constructs, thus satisfying the criterion of discriminant validity.
Construct reliability was further evaluated using CronbachAos Alpha (CA) and Composite Reliability (CR), with a minimum threshold of 0.
70 for each (Hair & Alamer, 2.
, as presented in Table 6.
Table 6.
Construct Reliability Test Variables CronbachAos Alpha Composite Reliability Description Social Media Influencer (SMI) Parasocial Relationship (PR) Reliable Reliable Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
51 | Renasabylla Putri Asriliorentika Variables Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 CronbachAos Alpha Composite Reliability Description Reliable Purchase Intention (PI) All variables had CronbachAos Alpha values ranging from 0.
918 to 0.
949 and CR values ranging 934 to 0.
955, all well above the minimum threshold of 0.
The AVE values for the three latent constructs were consistently above the 0.
50 threshold: SMI .
PR .
, and PI .
These results demonstrate that the instrument possesses strong reliability and internal consistency and meets the minimum criteria for convergent validity.
Inner Model Test R-Square Test The predictive ability of the structural model was evaluated using the R-squared value (R2 ), as shown in Table 7.
Table 7.
R-Square Test R-square R-square Adjusted The R2 value for PR is 0.
047, indicating that SMI explains 4.
7% of the variation in PR.
The R2 value for PI of 0.
171 indicates that SMI and PR together explain 17.
1% of the variation in PI, while the remainder is influenced by other factors outside the model.
Direct Effect Hypothesis Test Hypothesis testing was conducted using a bootstrapping procedure with 5,000 subsamples, with the relationship deemed significant if the T-statistics > 1.
96 and the P-value < 0.
05 (Hair & Alamer.
The results of the direct effect test are presented in Table 8.
Table 8.
Direct Effect Hypothesis Test Relationships Between Variables SMI Ie PI SMI Ie PR PR Ie PI Path Coefficient T-Statistics P-Value Description Significant Significant Significant Mediation Test (Specific Indirect Effect.
The following section presents the results of the mediation analysis using specific indirect effects.
The results of the mediation analysis using specific indirect effects are presented in Table 9.
Table 9.
Mediation Test (Specific Indirect Effect.
Relationships Between Variables SMI Ie PR Ie PI Indirect Coefficient Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
T-Statistics P-Value Description Partial Mediation Significant 52 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 An indirect coefficient of 0.
073, with a T-statistic of 2.
011 and a P-value of 0.
044, indicates that PR mediates the relationship between SMI and PI.
Since both the direct effect of SMI Ie PI and the indirect effect of SMI Ie PR Ie PI are significant, the mediation is classified as partial mediation.
Social Media Influencers and Purchase Intention A path coefficient of 0.
174, a T-statistic of 2.
370, and a P-value of 0.
018 indicate that social media influencers have a direct, positive, and significant effect on purchase intention.
H1 is The researchers argue that within the TikTok ecosystem, influencersAo ability to consistently blend informative, authentic, and entertaining content means that the product recommendations they share are no longer perceived as marketing messages but rather as advice from a figure trusted by their followers.
It is this dynamic that makes @catlovers29Aos content effective as a direct stimulus for purchase interest, as the established trust reduces the audienceAos resistance to commercial messages.
This finding aligns with Hung et al.
, who demonstrated that content appeal and trust in influencers substantially contribute to consumersAo purchase intentions for cosmetic products on TikTok.
Chen et .
confirmed that the quality of an influencerAos content directly shapes purchase intent through the audienceAos perception of credibility, while Amelia et al.
emphasized that trust and engagement between the information source and the audience are the primary determinants of digital marketing Social Media Influencers and Parasocial Relationships A path coefficient of 0.
216, a T-statistic of 2.
624, and a P-value of 0.
009 indicate that social media influencers have a positive and significant effect on the formation of parasocial relationships.
H2 is accepted.
The researchers argue that TikTokAos For You Page (FYP) algorithm plays a distinct role compared to other platforms in forming these parasocial bonds.
Unlike Instagram or YouTube, where users must actively search for or subscribe to content, the FYP proactively and repeatedly presents @catlovers29Aos content to users without any active effort on their part, making the exposure feel more organic and familiar.
It is this accumulation of organic exposure that gradually builds a sense of familiarity, comfort, and closeness toward Alma Nax Buaran, even in the absence of actual reciprocal interaction.
This finding aligns with Aw and Chuah .
, who argue that repeated exposure is a core mechanism in the formation of parasocial bonds, and is supported by Aw et al.
, who emphasize that content authenticity and a sense of personal interactivity are the primary drivers of parasocial bond formation on social media.
Taher et al.
also found that influencer reputation cues, particularly perceived expertise, are important antecedents for the formation of parasocial relationships among consumers of beauty products.
Parasocial Relationship and Purchase Intention A path coefficient of 0.
339 with a T-statistic of 4.
288 and a P-value of 0.
000 confirms that parasocial relationships have a positive and significant effect on purchase intention.
H3 is accepted.
The researchers believe that these findings reveal an important implication that is often overlooked by marketing practitioners: namely, that in the beauty industry, consumersAo purchasing decisions are not driven solely by the quality of the information conveyed by influencers, but rather by the depth of the emotional bond that has been established between followers and those influencers.
This is reflected in the magnitude of the parasocial relationship coefficient .
, which is far more dominant than the direct influence of social media influencers .
This implies that investing in influencers with strong parasocial bonds with their follower communities will be far more effective at driving purchase interest than relying solely on content quality.
Yuan and Lou .
support this view by arguing that parasocial relationships are a more robust predictor of purchase intent than other cognitive variables.
Dwiphala and Astuti .
reinforce these findings in the context of beauty influencers in Indonesia, while Sokolova and Kefi .
demonstrate a similar effect on beauty channels on YouTube and Instagram, where the influence of parasocial interaction is even stronger than the influencerAos credibility among most follower Volume 8.
Issue 1.
June 2026.
Page 45-56 DOI: 10.
30812/target.
53 | Renasabylla Putri Asriliorentika Target: Jurnal Manajemen Bisnis | e-issn: 2715-9361 Parasocial Relationships Mediate the Effect of Social Media Influencers on Purchase Intention A positive mediation coefficient of 0.
073, with a T-statistic of 2.
011 and a P-value of 0.
044, confirms that parasocial relationships significantly partially mediate the influence of social media influencers on purchase intention.
H4 is accepted.
The researchers argue that this partial nature of mediation actually carries important strategic implications: high-quality content still has its own direct influence on purchase intention, but that potential influence will be far greater if it is first grounded in an emotional bond consistently built through parasocial relationships.
In other words, beauty brands that focus solely on content quality without considering the depth of the relationship between the influencer and their follower community are, in fact, utilizing only a small fraction of the full potential of influencer marketing.
This finding aligns with Lin et al.
and Sharkasi and Rezakhah .
, who demonstrated that parasocial relationships mediate the relationship between social media influencers and purchase intention.
Sokolova and Kefi .
as well as Dwiphala and Astuti .
confirm that this pattern of partial mediation is widely applicable in the context of beauty influencer marketing, and is consistent with Purnamaningsih et al.
and Taher et al.
, who specifically demonstrated the mediating role of parasocial relationships in the context of beauty influencers in Asia.
IV.
Conclusion This study demonstrates that influencer marketing strategies on TikTok effectively drive purchase intent for beauty products among followers of the @catlovers29 account in Surabaya, both directly and through the partial mediation of parasocial relationships.
Content quality that is informative, authentic, and entertaining has been proven to provide a tangible stimulus that transforms viewers into potential Consistent content exposure also successfully builds an intimate PR between followers and Alma Nax Buaran without requiring reciprocal interaction.
The key findings of this study indicate that the strength of emotional bonds formed through parasocial relationships plays a more significant role in driving purchase intent than content quality The success of sales conversions on social media depends on an influencerAos ability to transform their content into a space of trust and lasting emotional connection in their followersAo minds.
Beauty brands are advised to evaluate influencers not only by reach or content quality but also by the strength of the parasocial bonds they have established with their community.
Further research is recommended to expand the model by including variables such as persuasion knowledge and brand trust, as well as examining other platforms like Instagram Reels or YouTube Shorts to obtain a more comprehensive cross-platform comparison.
Acknowledgment The author would like to thank all respondents who participated in this study, the academic advisor for the guidance and support provided throughout the research process, and friends who offered their support and encouragement.
The author also extends gratitude to Surabaya State University for the academic support provided.
Declaration The authors declare that they have no financial conflicts of interest or personal relationships that could influence the research results reported in this article.
Each author contributed proportionally to the design, data analysis, and writing of the article.
References