ISSN 2809-929X (Prin.
ISSN 2809-9303(Onlin.
Journal of Social Commerce Vol.
5 No.
4, 2025 (Page: 484-.
DOI: https://doi.
org/10.
56209/jommerce.
The Influence of Visual Content on TikTok Social Media Toward Customer Engagement in the Food and Beverage Industry on Product Purchase Decisions Dwi Erlanda Anggara1.
Tri Inda Fadhila Rahma1.
Nursantri Yanti1 State Islamic University of North Sumatra.
Indonesia Article History Keywords Visual Content Customer Engagement Purchase Decision TikTok JEL Classification M31.
M37.
D12.
D91 Abstract The purpose of this research is to provide insights into the extent to which visual content on TikTok influences customer engagement, how significantly customer engagement affects purchasing decisions, and the relationship between visual content on TikTok and customer engagement in making product purchase decisions.
This study employs a quantitative method with an explanatory research approach.
Data were collected through questionnaires distributed to 80 consumers of Sinar Utama chicken noodles.
The analysis technique used is Path Analysis via SmartPLS software version 3.
The measurements applied include the outer model assessment, mediation test, and inner model assessment.
The findings from the field demonstrate that: .
Attractive visual content significantly enhances customer engagement because strong visuals are able to capture attention, thereby encouraging interactions that increase interest in the product.
Customer engagement plays a crucial role in influencing purchasing decisions by building trust, loyalty, and emotional connections, which can drive customers to buy.
Customer engagement serves as a mediator between attractive visual content and purchasing decisions, as emotional involvement and customer interaction with visual content strengthen the decision-making process to purchase the product.
Introduction The development of digital platforms has revolutionized marketing strategies, particularly in the realm of digital business.
One of the fastest-growing platforms is TikTok, a short-video sharing application that has now become an essential part of digital marketing strategies (Lubis et al.
, 2023.
Wu, 2021.
Su, 2.
TikTok enables its users to create creative and viral content through engaging visual features such as video effects, music, and highly personalized recommendation algorithms.
This makes TikTok an effective promotional medium, especially for businesses in the food and beverage industry targeting young and modern consumers (Oktaheriyani et al.
, 2020.
Kulkarni, 2025.
Pieter et al.
, 2021.
Winzer et al.
, 2022.
Dupuis et , 2025.
Elliott et al.
, 2.
1Corresponding Author: Dwi Erlanda Anggara.
Email: dwierlandaanggara630@gmail.
Address Jl.
William Iskandar Ps.
Medan Estate.
Kec.
Percut Sei Tuan.
Kabupaten Deli Serdang.
Sumatera Utara 20371 Copyright A 2025.
Journal of Social Commerce is licensed under Creative Commons Attribution-ShareAlike 0 International License .
ttp://creativecommons.
org/licenses/by-sa/4.
Celebes Scholar pg Journal of Social Commerce Indonesia ranks as the number one country in the world with the highest number of TikTok This can be seen in Figure 1 below:
Figure 1.
Countries with the Largest TikTok Users in the World, 2025 A study conducted by Populix 2025 revealed that 86% of Indonesians have shopped through social commerce platforms, with an average monthly expenditure of IDR 275,000.
Among these transactions.
TikTok Shop dominated with 46%, followed by WhatsApp Business at 21%.
Facebook Shop and Instagram Shop at 10% each, while the remainder came from Telegram.
Pinterest, and other platforms.
The types of products purchased were diverse, with clothing taking the top position .
%), followed by beauty products .
%), and food and beverages .
%).
This phenomenon proves that appealing visual content can act as a trigger for consumersAo instant purchasing decisions.
In the digital era, visual content has become one of the key elements in marketing.
Visual content, such as photos and videos, has the ability to deliver messages more quickly and attractively compared to text (Lubis, 2020.
Sunarso & Mustafa, 2023.
Bashirzadeh et al.
, 2022.
Onyekuru, 2.
User engagement with digital platform content such as likes, comments, and shares is an important indicator of the success of marketing campaigns.
Content that achieves high engagement is often able to increase brand awareness and build emotional connections between brands and consumers (Appriliani Rizki & Adlina, 2.
Customer engagement is now considered a primary driver in creating meaningful brand experiences.
This engagement reflects a two-way marketing strategy in which customers are not merely recipients of information but actively participate in shaping brand perceptions through digital interactions (Nabila et al.
, 2023.
Tong & Chan, 2023.
Bozkurt et al.
, 2.
High engagement serves as the foundation for building consumer loyalty and trust, which are crucial in influencing purchasing decisions, particularly in industries that heavily rely on visual appeal, such as food and beverages (Aryawan & Valdez, 2024.
So et al.
, 2016.
Sarkis et al.
Ultimately, high engagement can significantly influence consumer purchasing decisions, especially in the food and beverage sector, where product visuals strongly determine attractiveness (Zahra et al.
, 2.
The development of digital technology and platforms has transformed the way businesses promote products and interact with customers.
One of the emerging platforms gaining massive popularity, particularly among younger generations, is TikTok.
TikTok has become a highly popular platform because it allows users to create and share short, creative, engaging, and easily consumable videos with audiences in a short amount of time.
This uniqueness and simplicity Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce have made TikTok one of the most effective digital marketing platforms, particularly in the food and beverage industry.
In Indonesia, the trend of using TikTok in business continues to grow, both in major cities and in regions such as Medan.
Many entrepreneurs in the food and beverage sector utilize TikTok as a tool to promote their products through appealing visual content, such as food preparation videos, taste reviews, aesthetically pleasing product displays, and trending formats adapted into business promotions.
Through short yet creative videos, entrepreneurs aim to capture usersAo attention, create interaction .
, and drive purchase decisions.
A concrete example of this phenomenon is the local business Mie Ayam and Kopi Sinar Utama in Medan, which leverage TikTok to introduce their products to a wider audience.
producing various visual content such as noodle-making videos, the unique taste of their coffee, and customer testimonials, this business seeks to attract attention and build connections with consumers, particularly young people who are active TikTok users.
However, there is a phenomenon where high engagement on digital platforms such as numerous likes, comments, views, and shares does not always align with an actual increase in purchasing decisions.
Many food and beverage businesses have viral content but do not experience significant growth in This raises the question of whether attractive visual content on TikTok truly influences consumer behavior or merely creates brand awareness without driving purchases.
Moreover, in todayAos digital era, consumers have become increasingly selective, critical, and intelligent in evaluating content.
They are not only influenced by appealing visuals but also consider other factors such as brand trust, product quality, consumer testimonials, and content This demonstrates that while TikTok provides great opportunities for generating engagement, its effectiveness in influencing purchasing decisions still requires deeper This phenomenon is particularly important to study in the food and beverage industry in Medan, so that business actors can understand whether and how visual content on TikTok can be optimally utilized to enhance customer engagement and simultaneously influence purchasing decisions.
This research is also expected to provide insights into more effective digital marketing strategies that align with the evolving trends of digital platforms.
The influence of visual content on purchasing decisions in this sector has become increasingly significant, as consumers tend to be more quickly attracted to appetizing food or beverage visuals (Pradnyani et al.
, 2.
Based on this data, it can be concluded that one of the strategies employed by Sinar Utama Medan Chicken Noodles to increase consumer purchasing decisions is the consistent application of content marketing and maintaining interaction with customers.
This is because product marketing and sales through social commerce, such as TikTok Shop, are highly dependent on the role of content marketing.
This study aims to understand and examine the influence of visual content on the TikTok digital platform on customer engagement in the business of Mie Ayam and Kopi Sinar Utama Medan.
Furthermore, the study also seeks to understand how customer engagement created through TikTok interactions affects purchasing decisions.
The research intends to investigate whether visual content published on TikTok directly influences consumer decisions to purchase food and beverage products without customer engagement, or whether engagement acts as an important mediating factor.
In addition, this study aims to identify the elements of visual content that are most effective in increasing engagement while simultaneously driving purchasing decisions.
Through this study, it is expected that business actors will gain deeper knowledge regarding the effectiveness of using TikTok as a promotional medium and be able to design more targeted marketing strategies to enhance competitiveness and product sales in the food and beverage industry.
Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce Previous studies such as (Appriliani Rizki & Adlina, 2.
, entitled AuThe Effect of Content Marketing on Customer Engagement and Purchase Decisions of Avoskin Your Skin Bae Series Products on TikTok Shop (A Study on Consumers in Medan Cit.
Ay, (Dewi et al.
, 2.
, entitled AuThe Effect of TikTok Content on Consumer Purchase Decisions in TikTok ShopAy, and Anggaran et al.
, entitled AuThe Effect of Food Review Content by TikTok Influencers on the Development of Micro Enterprises (Street Food Stall.
in Kebon Kacang.
Central JakartaAy, highlight that the existence of TikTok has facilitated MSMEs in increasing sales.
This study is relevant because there has not been much research that specifically discusses the effect of visual content on TikTok on customer engagement and purchase decisions in the food and beverage industry.
Therefore, this study is expected to provide insights into the extent to which visual content on TikTok affects customer engagement levels, the extent to which engagement influences consumer purchasing decisions, and the relationship between TikTokAos visual content and customer engagement in shaping purchasing decisions.
Literature Review Uses and Gratifications The Uses and Gratifications Theory provide a framework for examining audience behavior in media consumption.
This perspective emphasizes how individuals deliberately choose media channels to fulfill specific personal needs (Imsar & Nurhayati, 2.
As noted by Dangoran .
, this theoretical model represents an important paradigm in mass communication studies, distinguished by its audience-centered orientation rather than a media-centered This theory explains why individuals use certain media and what they expect from it.
In this study, the theory helps explain why customers engage with TikTok content and what types of gratification they seek from visual content on the platform.
Digital Marketing Theory In the modern era, digital marketing has become increasingly popular as business actors realize the importance of building a strong online presence.
By utilizing digital technology to promote products or services, companies can reach wider markets, interact directly with consumers, and achieve these goals at a relatively efficient cost.
With the rapid advancement of technology, digital marketing strategies are becoming even more vital for businesses seeking to survive in todayAos highly digital world (Nabila et al.
, 2.
This theory encompasses various aspects of digital marketing strategies, including how digital platform content influences consumer It is useful in analyzing how appropriate visual content strategies can enhance customer engagement and loyalty.
AIDA (Attention.
Interest.
Desire.
Actio.
AIDA stands for Attention.
Interest.
Desire, and Action.
This concept is a fundamental model in marketing and marketing communication.
The AIDA model describes the psychological stages that consumers go through before deciding to purchase or take action related to a particular product or service (Andirwan et al.
, 2.
This theory illustrates the consumerAos decision-making process.
TikTok visual content can be analyzed using this model to determine how such content captures attention, generates interest, stimulates desire, and ultimately drives action from customers.
Visual Communication This theory explains how visual elements in communication such as images, videos, colors, and layouts influence audience perceptions and reactions.
In the context of TikTok, it can be applied to analyze how displayed visual content impacts customer perceptions and engagement Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce (Nurbaiti et al.
, 2.
Visual communication practices represent an innovative methodology that combines artistic expression with technological applications to effectively convey conceptual messages (Harahap et al.
, 2.
Professional designers employ various visual elements, with imagery and typography as key components, to facilitate message delivery between businesses and their target audiences (Anggaran et al.
, 2.
Visual content itself refers to materials presented in visual form, such as images, videos, graphics, animations, or other visual elements.
According to Anggaran et al.
, visual content has several dimensions: Attractiveness: Visual content uses design to present information in an engaging and captivating way.
Knowledge: The human brain can process visual information quickly and Retention: The brainAos visual processing system works with long-term memory, making visual content easier to remember.
Engagement Theory This theory focuses on how users engage with digital content, including the elements that drive engagement such as interactivity, content relevance, and visual appeal (Siti Nurhalita & Imsar.
It is highly relevant for understanding the factors influencing customer engagement on digital platforms (Rahmayati et al.
, 2.
Customer engagement is a business strategy employed by companies to build continuous interaction with customers.
Its goal is to create valuable experiences while fostering emotional bonds that encourage loyalty, motivating consumers to make repeat purchases due to psychological attachment to a particular brand (Nabila et al.
, 2.
Based on various expert opinions, the concept of customer engagement can be divided into three main elements (Wiranti & Nugraha, 2.
Cognitive, mental activities that enable individuals to connect, assess, and evaluate choices.
Emotional, emotional conditions that are beyond oneAos control and can influence daily life.
Behavioral, behaviors that are shaped by rewards, recognition, or reinforcement from the environment.
Decision-Making Theory The decision-making process is a crucial mechanism that significantly influences individual behavior and therefore requires deep understanding (Nisrina Salwa, 2.
Desmita defines decision-making as a cognitive activity that produces a decision as the final output (Muhammad Baga, 2.
Meanwhile.
Anzizhan views it as a procedure of selecting an option from various alternatives to determine an action that achieves specific objectives.
According to Ibnu Syamsi (Supranto, 2.
, the indicators of decision-making are: Goals:
Goals must consider relevance to needs, clarity of formulation, and the feasibility of Identification of Alternatives: This involves generating several solution options, from which the most appropriate alternative is selected to achieve the goal.
Unforeseen Factors:
There are unpredictable elements where the success of an alternative choice can only be realized after the decision is implemented.
Future uncertainty highlights the importance of a leaderAos predictive competence as a key determinant of decision success.
Facilities for Evaluating Outcomes: Each alternative must be accompanied by output measurement criteria, including Positive and negative impacts.
Probability of uncontrollable events.
Resource Estimated costs for each combination of alternatives.
Methods This study adopted a quantitative research design with an explanatory orientation in order to examine the structural relationships among visual content, customer engagement, and purchase decision within the context of TikTok based marketing in the food and beverage sector.
explanatory approach was selected because the primary objective of the research was not Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce merely to describe consumer responses, but to test theoretically grounded causal relationships and mediation mechanisms among variables that have been conceptually linked in prior digital marketing and consumer behavior literature.
By employing this design, the study sought to empirically verify how visual content functions as a stimulus that shapes customer engagement and, subsequently, influences purchasing decisions.
The research was conducted at a local food and beverage business that actively utilizes TikTok as a promotional platform.
This setting was selected purposively, as the business represents a typical micro to small enterprise that relies heavily on visually driven digital content to attract and engage consumers.
The unit of analysis in this study was individual consumers who had been exposed to the businessAos TikTok content and had direct purchasing experience with the This focus allowed the study to capture consumer level perceptions and behavioral responses resulting from digital visual exposure.
Primary data were collected through a structured questionnaire administered directly to consumers of Mie Ayam and Kopi Sinar Utama Medan.
The population consisted of daily visiting consumers, which fluctuated between approximately sixty and one hundred Given the naturalistic field setting and the absence of a complete sampling frame, a quota sampling technique was employed.
This non probability approach was considered appropriate to ensure practical access to respondents who met the research criteria, namely consumers who were familiar with the businessAos TikTok content.
A total of eighty respondents were included in the final sample, a size that is considered adequate for Partial Least Squares Structural Equation Modeling, particularly in exploratory and explanatory models with mediation structures.
The research instrument was developed to measure three latent variables, namely Visual Content.
Customer Engagement, and Purchase Decision.
Visual Content was operationalized through three reflective indicators representing attractiveness, knowledge, and retention, which capture the extent to which TikTok content is visually appealing, informative, and memorable to consumers.
Customer Engagement was measured using reflective indicators encompassing cognitive, emotional, and behavioral dimensions, reflecting consumersAo mental involvement, emotional attachment, and observable actions toward the brand.
Purchase Decision was measured through indicators related to goal clarity, identification of alternatives, previously unrecognized influencing factors, and evaluative considerations in making purchasing choices.
All indicators were measured using a five point Likert scale ranging from strong disagreement to strong agreement, allowing respondents to express the intensity of their perceptions and Prior to hypothesis testing, the measurement instrument was evaluated for validity and reliability to ensure the robustness of the constructs.
Validity testing was conducted using item total correlation analysis, while reliability was assessed through internal consistency measures.
These preliminary assessments were necessary to confirm that the indicators adequately represented their respective latent constructs before proceeding to structural analysis.
Data analysis was performed using Partial Least Squares Structural Equation Modeling with SmartPLS version 3.
PLS SEM was chosen due to its suitability for predictive and explanatory research, its ability to handle complex mediation models, and its robustness when working with relatively small sample sizes.
The analytical procedure followed a systematic sequence beginning with the evaluation of the measurement model through convergent validity, discriminant validity.
Average Variance Extracted, and composite reliability.
This was followed by evaluation of the structural model through analysis of path coefficients, coefficient of determination, predictive relevance, and mediation effects.
The mediation role of Customer Engagement was examined by comparing direct and indirect effects within the structural Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg This approach enabled the study to determine whether visual content influences purchase decisions directly or whether its influence operates primarily through customer Hypothesis testing was conducted using bootstrapping procedures to assess the statistical significance of all structural paths.
Results and Discussion This study demonstrates that visual content disseminated through TikTok plays a substantive role in shaping consumer behavior within the food and beverage sector by operating through both direct and indirect mechanisms.
The overall findings indicate that visually engaging content significantly enhances customer engagement, which in turn exerts a strong influence on purchase decisions.
Moreover, the results confirm that customer engagement functions as a partial mediating mechanism, suggesting that visual content does not merely attract attention but also fosters cognitive, emotional, and behavioral involvement that strengthens consumersAo decision-making processes.
Data Instrument Testing Validity Test The validity test in this study was conducted using the Product Moment correlation.
instrument is considered valid if its total score reaches a value of Ou 0.
The validation process was performed with the assistance of IBM SPSS 25 software, and the results are presented in Table 1 below.
Based on the output, all items obtained correlation coefficients Ou 0.
3, therefore all items are declared valid.
Table 1.
Validity Test Output
Variable Visual Content
(X)
Customer Engagement (Z) Customer Decision (Y) Indicator Attractiveness Knowledge Retention Cognitive Emotional Behavioral Goal
Alternative Identification Previously Unrecognized Required Facilities for Output Calculation Item
Correlatio
,893
,714
,845
,637
,768
,913
,602
,694
P0,000
Value 0,000 0,000 0,000 0,000 0,000 0,000 0,000 Description Valid
Valid
Valid
Valid
Valid
Valid
Valid
Valid
,625
0,000
Valid
,845
0,000
Valid Reliability Test This study verified the reliability of the measurement instruments through reliability analysis.
The evaluation criterion used was CronbachAos Alpha coefficient with a minimum threshold of The testing was carried out using IBM SPSS version 25 software.
As shown in Table 2, all research variables met the reliability standard with CronbachAos Alpha values exceeding 0.
indicating that the measurement instruments used have adequate internal consistency.
Table 2.
Reliability Test Output Variable Visual Content (X) CronbachAos Alpha Conclusion Reliable Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg Customer Engagement (Z) Customer Decision (Y) Reliable Reliable Classical Assumption Test To test the linearity assumption, this study applied the compare means method to evaluate the linear relationship between independent and dependent variables.
The results of the test are presented in Table 3.
Table 3.
Classical Assumption Test Output No Variable Relationship Deviation from Linearity (Sig.
X*Y
009 < 0.
Z*Y
002 < 0.
Conclusion Meets criteria Meets criteria Structural Equation Modeling with PLS Approach The testing covered three aspects: .
hypothesis testing on the outer model, .
testing, and .
hypothesis testing on the inner model.
Outer Model Testing (Loading Factor Result.
The outer model analysis in this study was conducted using six evaluation criteria through SmartPLS software, including: .
convergent validity, .
discriminant validity, .
composite reliability, .
Average Variance Extracted (AVE), .
structural model .
nner mode.
testing, and .
Goodness of Fit test.
Convergent validity for the research model with reflective indicators was measured based on the correlation between item scores evaluated using SmartPLS.
For reflective indicators, a correlation value above 0.
70 is considered ideal.
However, according to Chin .
, in prior studies a loading factor value between 0.
50Ae0.
is still acceptable.
Based on this consideration, the minimum loading factor threshold in this study was set at 0.
Convergent Validity In SmartPLS analysis, the outer model calculates the contribution of indicators through two approaches: .
outer loading for reflective indicators .
value > 0.
7 demonstrates good representatio.
, and .
outer weight for formative indicators.
These outputs indicate the extent to which the indicators explain the observed latent variables.
The convergent validity test demonstrated that all indicators for each variable had values > 0.
The full results can be seen in the following figure:
0,672
0,608
0,648
0,702
0,708
0,668
0,562
0,672
0,577
0,762
0,788
0,768
0,762
Figure 2.
Structural Diagram in the PLS Model Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg For a more detailed explanation, each variable is presented as follows:
Visual Content Factor (X) The Visual Content variable was measured using reflective indicators.
The Visual Content loading factor output is as shown in the table below: The Visual Content variable was measured using reflective indicators.
The loading factor results for Visual Content are presented in the following table:
Table 4.
Visual Content Factor Test Output Indicator Loading Factor Mean P-Value The Visual Content variable consists of three indicators: attractiveness, knowledge, and Each indicator is represented by one item, resulting in a total of three items.
The data analysis results show that all three indicators significantly shape the Visual Content variable, as indicated by the loading factor values in Table 4.
Since all p-values are < 0.
05, the indicators are significant.
Among them.
Retention has the highest loading factor .
, making it the most dominant component of Visual Content.
This indicates that the retention aspect reflecting the ease of recalling Sinar Utama Medan products through TikTok content is the most reflective element of this variable.
Customer Engagement Factor (Z) The Customer Engagement variable was measured using reflective indicators.
The loading factor results for Customer Engagement are presented below:
Table 5.
Customer Engagement Factor Test Output Indicator Loading Factor Mean P-Value Customer Engagement is formed by three main indicators: cognitive, emotional, and Each indicator is represented by three items.
Based on the analysis, all three indicators have a significant influence on Customer Engagement, with their respective loading factors shown in Table 5.
The highest loading factor .
was observed in the behavioral indicator, making it the most dominant in shaping Customer Engagement.
This confirms that customer engagement is most strongly reflected through behavioral aspects, such as choosing Sinar Utama Medan products based on TikTok reviews.
Customer Decision Factor (Y) The Customer Decision variable was measured using reflective indicators.
The loading factor values for Customer Decision are shown in the following table:
Table 6.
Customer Decision Factor Test Output Indicator Loading Factor Mean P-Value Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg The Customer Decision variable consists of four main indicators: goal, alternative identification, previously unconscious factors, and the need for facilities to evaluate outcomes.
The data analysis shows that all four significantly shape Customer Decision, as reflected in the loading factor values in Table 6.
The highest loading factor .
corresponds to the need for facilities to evaluate outcomes, making it the most dominant indicator of Customer Decision.
This indicates that customer decisions are strongly influenced by alternative evaluation, reflected in the purchase of Sinar Utama Medan products due to positive assessments from other customers.
Discriminant Validity Discriminant validity ensures that each latent variable in this study is unique and clearly distinguishable from others.
The main criterion is that the loading factor of an indicator on its original variable must be higher than its loading on other variables.
Table 7.
Discriminant Validity Test (Cross Loadin.
Indicator The results confirm that all indicators meet this criterion: their loading factor values are always higher on their original constructs than on other constructs.
Thus, the model satisfies strict discriminant validity requirements, indicating that each latent variable accurately represents a distinct construct.
Average Variance Extracted (AVE) Validity and Reliability of the Constructs were evaluated based on the level of reliability and the Average Variance Extracted (AVE) values.
According to standard criteria, a construct is considered reliable if its AVE value exceeds 0.
The complete AVE values of all constructs are presented in Table 8:
Table 8.
Output of Average Variance Extracted (AVE) Variable AVE The data in the table demonstrate that all constructs in this study meet the reliability standard.
This criterion is fulfilled since the AVE values of all constructs are above 0.
50, in line with the requirements of research methodology.
Evaluating Composite Reliability Furthermore, to test reliability, the Composite Reliability value was used.
A construct is considered reliable if its value exceeds 0.
70, as shown in Table 9 below:
Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg Table 9.
Output of Composite Reliability Variable Composite Reliability Conclusion Reliable Reliable Reliable The analysis of the table above confirms that all constructs in this study fulfill the requirements of composite reliability.
This is evident from the Composite Reliability values of each construct, all of which are above 0.
70, thereby meeting the recommended standard criteria in Testing the Structural Model (Inner Mode.
This study applied a structural model to examine the relationships among constructs, with the evaluation procedure including: .
analysis of R-Square as an indicator of the modelAos predictive strength, and .
statistical significance testing through t-tests on the structural path Table 10.
Output of R-Square Values Endogenous Variable R Square (RA) The research design involved two dependent variables (Customer Engagement and Purchase Decisio.
, which were influenced by independent variables.
The determination analysis output (Table .
reveals that the model explains 54.
6% of the variance in Customer Engagement (RA = 0.
7% of the variance in Purchase Decision (RA = 0.
Goodness of Fit Testing The evaluation of the structural modelAos goodness of fit was conducted using Q-Square Prediction Relevance (QA) in the inner model.
The results show RA values of 0.
546 for Customer Engagement (Z) and 0.
567 for Purchase Decision (Y).
The predictive QA value was then calculated based on the established formula.
Q2 = 1 Ae .
- 0,.
- 0,.
Q2 = 1 Ae .
Q2 = 1 Ae .
,196.
Q2 = 0,803418 Based on the calculations, the Predictive Relevance value obtained was 0.
803418, equivalent This score demonstrates that the model employed possesses relevant and adequate predictive competence.
More specifically, the value of 80.
34% indicates that the model is capable of explaining 80.
34% of the total data variation, while the remaining 19.
66% is explained by other variables excluded from the model.
These findings confirm that the constructed PLS model meets the criteria of a good model, as it can accommodate and explain more than 80% of the information contained in the research data.
The predictive accuracy level 34% proves the strength of the model in predicting the studied phenomenon.
Mediation (Intervenin.
Test Results The next stage of analysis involves testing the mediation effect by considering two aspects: .
the magnitude of the path coefficient, and .
the statistical significance level of the variable Dwi Erlanda Anggara et al.
Journal of Social Commerce Celebes Scholar pg
0,608
0,672
0,702
0,648
0,708
0,668
0,562
0,672
0,577
0,762
0,788
0,768
0,762
Figure 3.
Mediation Test Output Using Algorithm The mediation analysis in Figure 3 reveals two significant positive relationships: .
from Visual Content to Customer Engagement ( = 0.
, and .
from Customer Engagement to Purchase Decision ( = 0.
To test the mediation effect, a comparative analysis was conducted by comparing the model that includes the mediator variable with the model that excludes it.
0,528
0,656
0,596
0,657
0,647
0,843
0,682
0,740
Figure 4.
Mediation Test Output Using Algorithm Without Customer Engagement The path analysis reveals that: .
the direct effect of Visual Content on Purchase Decision is .
after including Customer Engagement as a mediator, the coefficient increased to This finding confirms the existence of a partial mediation effect in the relationship (Figure .
Hypothesis Testing Results (Inner Mode.
The statistical significance criteria in this study are based on the parameter values presented in the inner weight results output, which represent the estimation of the structural model as shown in Table 11.
These parameters are used to evaluate the relationships among the variables in this Table 11.
Inner Weight Output Values Relationship
XIeY
XIeZ
ZIeY
Original Sample (O) Sample Mean
(M)
Std.
Deviation Statistics Values The hypothesis testing procedure in this study was carried out through a statistical simulation approach to examine each hypothesized variable relationship.
H1: The Effect of Visual Content on Customer Engagement The analysis of the first hypothesis confirms the existence of a significant positive effect of Visual Content (X) on Customer Engagement (Z) with a path coefficient of 0.
< 0.
Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce This finding indicates that higher quality visual content increases the level of customer Thus.
Hypothesis 1 is supported, as there is clear evidence of a significant positive relationship between the two variables.
Visual content essentially represents a creative expression that integrates art and technology to communicate ideas.
As explained by Anggaran et al.
, design practitioners employ various communication media to deliver brand messages to target audiences, with visual elements and text serving as the primary message carriers.
Visual content is a means of presenting specific meanings that can be perceived visually and attract customer attention.
This is consistent with previous research by (Wiranti & Nugraha, 2.
AuAnalysis of Customer Engagement Strategy on Loyalty at PT.
Nasmoco MagelangAy, and (Pradnyani et al.
, 2.
AuAnalysis of the Mediating Role of Customer Engagement in the Effect of Content Marketing on Purchase Intention.
Ay These prior studies reinforce the finding that Visual Content strongly influences Customer Engagement.
H2: The Effect of Customer Engagement on Purchase Decision The analysis of the second hypothesis reveals that Customer Engagement (Z) has a significant positive effect on Purchase Decision (Y), with a path coefficient of 0.
-value = 0.
This finding suggests that an increase in customer engagement leads to a stronger tendency to purchase.
Therefore.
Hypothesis 2 is validated, as the results demonstrate a significant positive relationship between the two variables.
Customer engagement is a strategic approach by companies to build ongoing interaction with consumers in order to create meaningful experiences and emotional bonds, ultimately driving loyalty and repeat purchases through brand attachment (Nabila et al.
, 2.
This finding is consistent with prior studies, including: Rizki .
AuThe Impact of Content Marketing on Customer Engagement and Purchasing Behavior of Avoskin Products on TikTok Shop (Study on Medan Consumer.
Ay Aryawan & Valdez .
AuThe Role of Customer Engagement in the Purchasing Process of Spotless Products on Instagram @Madformakeup.
Ay These studies affirm that Customer Engagement directly influences a companyAos Purchase Decision outcomes.
H3: The Effect of Visual Content on Purchase Decision Through Customer Engagement The mediation analysis results indicate that Customer Engagement plays a role as a partial mediator in the relationship between Visual Content and Purchase Decision.
This is supported by statistical significance values, which show that: Visual Content significantly affects Customer Engagement .
= 0.
000 < 0.
Customer Engagement significantly affects Purchase Decision .
= 0.
000 < 0.
Therefore.
Hypothesis 3 is accepted.
Decision-making is a critical mechanism that influences individual actions, making a comprehensive understanding of this process essential.
According to Desmita, decisionmaking is a cognitive activity that produces a decision as its output (Muhammad Baga, 2.
Furthermore, this process can be understood as a series of evaluation stages among various alternatives to determine the best course of action in achieving a given objective.
This conclusion is reinforced by prior studies, such as Nargis .
, who examined AuThe Impact of Service Quality on Consumer Loyalty with Customer Satisfaction as a Mediating Variable in PalembangAy, and Takaya .
, who investigated AuThe Effect of TikTok Marketing Content on Consumer Trust and Purchase Intention: A Study on Fashion Products.
Ay These earlier studies highlight that Customer Engagement significantly influences how visual content provided by content creators affects customer Purchase Decisions.
Dwi Erlanda Anggara et al.
Celebes Scholar pg Journal of Social Commerce Conclusion The field findings demonstrate that: .
Attractive Visual Content significantly enhances Customer Engagement because strong visuals can capture attention, deliver messages more effectively, and evoke emotions that drive interaction, thereby increasing interest in the product and strengthening the relationship between customers and the brand.
Customer Engagement plays a crucial role in influencing Purchase Decisions by building trust, loyalty, and emotional connections, which in turn encourage customers to buy.
Customer Engagement serves as a mediator between Attractive Visual Content and Purchase Decisions, as emotional involvement and customer interaction with visual content can reinforce purchasing behavior.
Based on the results of this study, it is recommended that future research expand the scope of analysis beyond a single MSME case, such as Mie Ayam and Kopi Sinar Utama Medan, to include various businesses within the food and beverage sector at the micro, small, and medium This would provide a more comprehensive understanding of the impact of visual content on digital platforms in relation to consumer engagement and purchase decisions.
Furthermore, future studies are encouraged to incorporate additional variables such as brand trust, influencer marketing, or perceived value in order to enrich the analytical model and explore other determinants of purchasing behavior.
Finally, since this study focused exclusively on TikTok, subsequent research may also consider comparing the effectiveness of visual content across other digital platforms, such as Instagram.
YouTube, or Facebook.
References