TikTok Live Shopping has experienced tremendous expansion since its inception and currently serves as a crucial component that combines live streaming with e-commerce[cite: 8]. The platform allows users to purchase products in real-time while simultaneously interacting with live streaming content[cite: 9]. In Indonesia, TikTok Shop has garnered significant recognition, making Indonesia a key market for TikTok in Southeast Asia, boasting over 125 million TikTok users and demonstrating a conversion rate for live shopping sales three times higher than that seen in conventional e-commerce (ContentGrip, 2025)[cite: 10]. As of July 2024, Indonesia had approximately 157.6 million TikTok users, establishing itself as the country with the highest user base globally during that timeframe[cite: 11]. This figure continues to rise, and by 2025, Indonesia is expected to have 194.37 million users, indicating significant growth in the app's penetration among Indonesians[cite: 12]. This indicates that TikTok enjoys significant popularity within the Indonesian demographic, which includes both youth and adults[cite: 13]. According to a report from Databoks in November 2023, Indonesia ranked second in the world in terms of TikTok users, second only to India, which had more users at the time[cite: 14]. TikTok usage time in Indonesia is also very high compared to other countries[cite: 15]. According to data from RRI, the average TikTok user in Indonesia spends around 44 hours and 54 minutes per month on the app, making it the country with the highest usage time in the world (Goosstats, 2024)[cite: 16]. This figure demonstrates a very strong level of user engagement, with Indonesian users spending more time exploring TikTok content each month than the global average[cite: 17]. User engagement in the Indonesian context encompasses not only app utilization but also participation in a myriad of creative endeavors, including content creation and engagement with the platform's rapidly growing viral phenomena[cite: 18]. According to data from KOL ID (2024), it has been documented that TikTok users in Indonesia spend approximately 41 hours and 35 minutes per month on the platform, a figure that exceeds the global average for TikTok engagement[cite: 19]. TikTok's emergence in Indonesia can be explained by several factors[cite: 20]. First, the platform offers a diverse range of content that appeals to a wide demographic, spanning various age groups and interests[cite: 21]. Second, TikTok offers creators the opportunity to monetize through established features, including advertising, affiliate marketing, live streaming, and brand collaborations[cite: 22]. Third, TikTok uses a highly efficient algorithm that curates content aligned with individual user preferences, thereby increasing app engagement[cite: 23].

Figure 1. the 10 countries with the most TikTok users in the world as of July

The image above shows data on the 10 countries with the most TikTok users in the world as of July 2025. Based on the graph, Indonesia ranks first with the most users, namely 194.37 million. This indicates that Indonesia is the largest market for TikTok, reflecting the platform's popularity among Indonesians. With rapid growth and high engagement rates, TikTok in Indonesia shows great potential as a platform for digital marketing. For companies and brands, TikTok offers a highly effective marketing channel to reach young audiences and increase brand awareness globally. TikTok isn't just an entertainment-based app; with its business opportunities, it's also expanding into sales features similar to those found on other marketplaces. Along with these features, TikTok has also introduced features specifically for content creators, such as the yellow basket. Many content creators can leverage this feature to generate income. Content creators compete to create the best content to earn income through content in the yellow basket, commonly known as affiliate marketing, or through live streaming, which features the yellow basket.

TikTok Live Shopping has experienced tremendous growth since its inception and currently stands as a key feature integrating live streaming with e- commerce. The platform facilitates users in acquiring products in real-time while engaging with live streaming content. In the Indonesian context, TikTok Shop has achieved significant popularity, positioning Indonesia as TikTok's key market in Southeast Asia, boasting over 125 million TikTok users, and demonstrating live shopping sales conversion rates that are up to three times higher than those observed in traditional e-commerce (ContentGrip, 2025). Data also shows that the success rate of live shopping on TikTok is highly dependent on real-time interactions and direct promotions, such as discounts and limited-time offers, which trigger impulsive purchasing decisions. In Indonesia, for example, 55.2% of purchasing decisions in the fashion segment can be explained by interactions during live streaming and the discounts offered during the session (Faturrohman, 2025). Thus, TikTok Live Shopping provides a significant opportunity for businesses, especially in Indonesia, to capitalize on a large audience and high engagement rate, increase sales conversions, and strengthen the relationship between brands and consumers through a more personalized and interactive approach.

Not only through live shopping, content creators can also utilize the yellow basket as a source of income, commonly known as affiliate marketing. With the current affiliate feature, content creators are increasingly competing to create optimal content, reviewing products clearly and explaining the benefits of using the products they review. Affiliate marketing is a performance-based marketing model in which an affiliate (publisher) earns a commission for every transaction or action taken through the promotion of products or services provided by merchants (advertisers). In this model, affiliates promote products or services through unique affiliate links, and they receive a commission when consumers make a purchase or take a desired action, such as registering or clicking. This process offers benefits to all parties involved, especially since companies only pay affiliates when the desired results are achieved (Shofify.com, 2025).

Affiliate marketing is a feature of TikTok Shop, a platform for content creators to create their best content, not just for entertainment but also to generate income. To become an affiliate marketer, TikTok sets several rules that content creators must meet, one of which is a minimum number of followers. To be able to display the yellow basket feature, content creators must have a minimum of 600 followers. Furthermore, affiliate marketing has various types, including pay- per-sale (commission based on sales), pay-per-click (commission based on ad clicks), and pay-per-action (commission based on specific actions such as registration). This model allows businesses to save on marketing costs because they only pay for measurable results, which is very profitable, especially for small businesses with limited budgets (Thepowermba.com, 2025). For companies, the primary benefit of affiliate marketing lies in its ability to expand market penetration without incurring significant risk, given that compensation is based on performance metrics. As a report from AffiliateWP shows, this framework also allows companies to engage a more targeted demographic and increase sales conversion rates in an efficient and scalable manner. Many companies are now even creating affiliate communities as marketing tools and willing to pay per piece of content. One example is Paragon, which created a community of affiliates and pays affiliates between 5,000 and 15,000 rupiah per piece of content.

The features provided by TikTok Shop are very beneficial for micro, small, and medium businesses. With the features described above, many audiences who initially view content simply for entertainment end up making sudden purchases or impulse buys. Impulse buying, commonly referred to as impulse buying, indicates a transaction decision made spontaneously without prior consideration, often triggered by emotional triggers or external influences such as promotions or advertising. As articulated by Kotler and Keller (2016) in Siswanto's work (2024), impulse buying tends to occur when consumers acquire a product without careful consideration or thorough assessment of its price and quality, which is usually a consequence of the atmosphere or stimuli present in the surrounding environment. Consumer trust in a product and its endorsers is a crucial factor in reducing barriers to impulse buying behavior. When consumers have assurances about product quality and the reliability of a platform or influencer, they are generally more likely to engage in unplanned purchases. Consequently, methodologies such as affiliate marketing and live streaming have the potential to increase consumer trust, thereby increasing impulse buying rates. However, over time, TikTok Shop has tightened the rules governing its live shopping and affiliate features. Within a month, TikTok's algorithm has undergone rapid changes. This was felt by one business owner, Broguy Fashion, who noted the rapid changes in TikTok Shop's regulations and algorithm, resulting in delays in adaptation for content creators and hosts. This can be demonstrated as follows:

The table above concludes that Broguy Fashion has experienced a decline in annual revenue. The owner of Broguy Fashion explained that this was due to the lack of responsiveness of live hosts and affiliates to changes made by the TikTok Shop app. For example, changes in pricing by hosts or content creators resulted in accounts being charged with violations and resulted in reduced views or viewership during live streams. Content creators' content was not on the FYP or was not subject to the store's advertising GMV MAX. This research is highly appropriate given the rapid development and regulations of the TikTok Shop app. This study can help business owners identify the causes of the decline in revenue and increase revenue beyond their stated targets.

Live Shopping

Live shopping is a contemporary digital marketing strategy in which merchandise is marketed directly to consumers through live streaming sessions[cite: 73]. Throughout the broadcast, sellers or influencers actively promote the merchandise, offer detailed explanations, and respond to questions in real time (Siswanto et al., 2023)[cite: 74]. This approach utilizes interactivity and a sense of urgency to stimulate impulse buying behavior among consumers[cite: 75]. On TikTok Shop, live shopping works by allowing sellers or influencers to broadcast live, promote their products, and provide consumers with the opportunity to interact directly with them[cite: 76]. Viewers can ask questions or leave comments during the broadcast, as well as purchase products directly through features embedded in the broadcast (Taufiq et al., 2025)[cite: 77]. Live shopping influences impulsive buying because it creates a sense of urgency and high emotional engagement in consumers[cite: 78]. Putri & Masnita (2024) showed that interactions during live shopping can strengthen spontaneous purchasing decisions, especially when there are limited-time offers or special discounts that are only valid during the live broadcast[cite: 79]. Research by Amin & Fikriyah (2023) shows that live shopping on social platforms like TikTok and Instagram is effective in increasing sales, primarily due to the social engagement factor and limited-time offers that create urgency among consumers (Taufiq et al., 2025)[cite: 80].

The dimensions and indicators of live shopping in this study encompass three main aspects: interactivity, offer urgency, and emotional engagement. Interactivity is reflected in the level of two-way communication between sellers and viewers during the live session, as evidenced by the number of comments submitted by viewers, the speed and accuracy of seller responses to those comments, and activities such as quizzes conducted to increase audience participation. Furthermore, offer urgency relates to sellers' efforts to encourage quick purchasing decisions, demonstrated by offering discounts or limited-time promotions valid only during the live session, reminders to make purchases immediately, and exclusive special offers during the live session. Emotional engagement, on the other hand, describes the extent to which consumers feel emotionally connected during live shopping, as measured by the level of consumer excitement or enthusiasm, the perception of excitement during the live session, and the emergence of positive responses influenced by the interactive atmosphere created during the activity.

Affiliate Marketing

Many fashion brands, such as Broguy Fashion, use affiliate marketing to expand their market reach on TikTok Shop. Influencers who promote fashion products through short videos and live shopping can increase brand credibility, thus encouraging more consumers to make impulse purchases (Putri & Masnita, 2024). Badi'ah et al. (2025) identified that affiliate marketing has a significant impact on impulse buying behavior, especially in the e-commerce realm, because it fosters deep trust between consumers and influencers. The dimensions and indicators of affiliate marketing in this study encompass four main aspects: affiliate credibility, content quality, social influence, and ease of purchase links. Affiliate credibility is reflected in a good reputation, honesty in conveying information, product understanding, and consumer trust in recommendations. Content quality relates to an attractive, easy-to-understand presentation, and the ability to clearly explain and illustrate product benefits, thereby increasing consumer interest. Affiliate social influence is evident in the ability to build closeness and personal relationships with audiences, provide a sense of comfort, and influence consumer decisions. Meanwhile, ease of purchase links is demonstrated through easy product access, a practical purchasing process, and helping consumers avoid the need to manually search for products.

Impulse Buying

Impulse buying refers to acquisitions that occur spontaneously and without prior consideration. Various determinants that impact impulse buying behavior include emotional dynamics, the urgency of promotional offers, and sociocultural influences (Stern, 1962). Direct shopping mechanisms and affiliate marketing can increase impulse buying by leveraging these determinants. The Stimulus-Organism-Response (SOR) framework proposed by Mehrabian & Russell (1974) serves to explain how external stimuli, such as promotional offers or product endorsements, influence consumers' internal cognitive processes, ultimately culminating in impulse buying behavior. In the context of fashion products, impulse buying is often triggered by visual factors, emotional engagement, and influencer recommendations. Siswanto et al. (2023) showed that live shopping and affiliate marketing can trigger impulsive buying because they create urgency, visual appeal, and strong social influence. Psychological factors such as time pressure, discounts, and social interaction can encourage impulsive buying. Time pressure arises when there are limited offers during live shopping, while discounts and social interactions, such as influencer influence, increase consumers' tendency to purchase products without prior planning (Taufiq et al., 2025).

This study uses a quantitative approach with an associative research type that aims to analyze the relationship and influence between the variables Live Shopping (X₁), Affiliate Marketing (X₂), and Impulse Buying (Y). This approach was chosen because it is able to measure the strength of the relationship between variables using numerical data and statistical analysis (Sugiyono, 2019). The population in this study were all 1,330 Broguy Store customers who had made purchases through TikTok Shop. The sampling technique used the Slovin formula with a 10% error rate, resulting in a minimum sample of 93 respondents. To increase validity and anticipate incomplete data, the number of samples was set at 100 respondents. The types of data used consisted of primary and secondary data. Primary data were obtained through distributing online questionnaires to respondents, while secondary data were obtained from literature, scientific journals, and supporting data related to e-commerce. The data collection technique used a survey method with a questionnaire instrument based on a Likert scale of 1–5, ranging from strongly disagree to strongly agree. The collected data were then analyzed using statistical techniques to test the relationship between the variables in this study.

Measurement Model (Outer Model)

This examination aims to assess the legitimacy of the questionnaire used in the research. This assessment includes convergent validity, discriminant validity, and reliability. Presented here are the results of the validity evaluation related to the research questionnaire statements.

Figure 2. Measurement Model (Outer Model)

Convergent Validity

Validity assessment using convergent validity is based on the correlation between statement indicators, scores, or score components that have been estimated through the application of the SmartPLS 4.0 software suite. The magnitude of each indicator is considered significant if it shows a correlation exceeding 0.7 (Hair et al., 2021). However, according to Ghozali et al. (2015), external loading values ranging from 0.5 to 0.6 are still considered acceptable during the development stage of measurement scales in research, where the evaluation of convergent validity is illustrated in the following table:

According to the findings obtained from the external model evaluation using Structural Equation Modeling with Partial Least Squares (SEM-PLS), it has been determined that each indicator related to the constructs of Affiliate Marketing (AM), Impulse Buying (IB), and Direct Shopping (LS) has an external loading value exceeding the threshold of 0.70. This observation indicates that each individual indicator is adept at depicting the latent construct it embodies in a robust and consistent manner. In the Affiliate Marketing construct, the indicator outer loading values range from 0.727 to 0.901, indicating that all indicators contribute significantly to explaining the Affiliate Marketing variable. No indicators scored below the threshold, thus all indicators were deemed valid and worthy of retention. The Impulse Buying constructs showed comparable results, with external loading values fluctuating between 0.740 and 0.837. These values indicate that the Impulse Buying indicator exhibits a strong correlation with the corresponding construct and has met the convergent validity criteria. Meanwhile, the Live Shopping construct exhibits relatively high external loading values, ranging from 0.787 to 0.920. These increasing values indicate the indicator's exceptional ability to accurately assess the Live Shopping construct. In short, the results of the external loading examination prove that the external model in this study has met the convergent validity criteria, thus validating all indicators and allowing development to the structural model testing phase (inner model) to evaluate the relationships between latent variables.

Specifically, the Affiliate Marketing variable shows an AVE value of 0.667, the Impulse Buying variable 0.602, and the Live Shopping variable 0.735. Since the AVE values for these three variables are well above the 0.50 threshold, it can be concluded that convergent validity at the construct level has been met. This confirms that each latent variable is able to explain more than 50% of the variance of its constituent indicators, thus the construct is valid.

The results obtained from the discriminant validity assessment using the Heterotrait-Monotrait Ratio (HTMT) methodology indicate that all constructs in the model conform to the specified standards. The HTMT values related to all pairs of constructs remain below the strict threshold of 0.85. Specifically, the correlation between Impulse Buying and Affiliate Marketing is 0.710. Furthermore, the correlation between Live Shopping and Affiliate Marketing is 0.842, and the correlation between Live Shopping and Impulse Buying is 0.748. Since all these HTMT values are below 0.85, it can be concluded that discriminant validity has been met. Furthermore, there is an evaluation of the Fornell Larcker-Criterion which suggests that the AVE of a construct should be compared with other constructs, furthermore the model construction should not be greater than the AVE and can be said to be good or valid with a value < 0.7. Table 4. shows the Fornell Larcker value.

According to the table mentioned above, it is proven that the basic value of AVE Affiliate Marketing is 0.816, Impulse Buying is 0.776, and Live Shopping is 0.857. So it can be seen that each variable has a root AVE value that is higher than the correlation value between variables or is worth < 0.7. So it can be concluded that the evaluation with Fornell Larcker for discriminant validity of all variables is declared good or valid.

It is known that the Cronbach alpha value of the Affiliate Marketing variable is 0.961, the Impluse Buying variable is 0.934 and the Live Shopping variable is 0.948 so that all variables are reliable or accurate with good reliability. Furthermore, the composite reliability (rho_c) value of the Affiliate Marketing variable is 0.965, the Impluse Buying variable is 0.943, and the Live Shopping variable is 0.957 so that all variables are reliable or accurate with good reliability.

In the table mentioned above, it is evident that the Impulse Buying variable produces an R Square value of 0.546. The findings indicate that the Impulse Buying variable is significantly influenced by the Live Shopping and Promotion variables on Social Media, categorized as moderate, while the remaining influence comes from other factors not included in the research model.

According to the table above, the F-square value relating to the relationship between Affiliate Marketing and Impulse Buying is measured at 0.067, which is classified as substantial. Conversely, the F-square value regarding the ratio of Live Shopping to Impulse Buying is recorded at 0.175, categorizing it as significant.

Based on the findings obtained from the Q² test, it can be concluded that the structural model shows adequate predictive ability regarding the Impulse Buying construct, as evidenced by the Q² value of 0.292, which is greater than zero, meaning the model has predictive power. In contrast, the Q² value of 0.000 associated with the Affiliate Marketing and Direct Shopping constructs does not warrant further interpretation (weak), considering that both constructs are classified as exogenous variables. Consequently, the SEM-PLS model used in this investigation shows quite good predictive relevance regarding the main endogenous variables.

Hypothesis Testing

Hypothesis testing is conducted by examining direct and indirect effects. The direct effect assesses the impact of the independent variable on the dependent variable alongside the mediating variable. Furthermore, the indirect effect evaluates the effect of the independent variable on the dependent variable through the mediating variable. Furthermore, this analysis can also be conducted by assessing probability values and t-statistics. For probability values and p- values with a 5% alpha level, the threshold is set at >0.05. The critical t-table value for a 5% alpha is <1.96. Consequently, the hypothesis acceptance or rejection criteria stipulate that the alternative hypothesis (Ha) is accepted and the null hypothesis (Ho) is rejected if the t-statistic exceeds the t-table value of 1.96 and the p-value is less than 0.05.

H1 = The correlation coefficient for Affiliate Marketing on Impulse Buying was measured at 0.296, accompanied by a T-statistic of 2.137 and a P-value of 0.033. The observation of a T-value below 1.96 and a P-value below the threshold of 0.05 indicates that the effect of Affiliate Marketing on Impulse Buying has statistical significance. Consequently, the hypothesis affirming the influential nature of Affiliate Marketing on Impulse Buying is substantiated. This finding implies that activities related to affiliate marketing have the capacity to directly stimulate the emergence of impulsive buying behavior among consumers. H2 = The relationship between Live Shopping and Impulse Buying shows a strip coefficient value of 0.479, accompanied by a T-statistic value of 3.718 and a P-value of 0.000. A T-statistic value exceeding 1.96, along with a P-value less than 0.05, indicates that Live Shopping has a positive and statistically significant effect on Impulse Buying. Consequently, the hypothesis stating that Live Shopping influences Impulse Buying is confirmed. These results imply that live shopping initiatives have the potential to generate attractive stimuli, thus encouraging consumers to engage in spontaneous purchases, without prior consideration.

H3: Based on the results of the structural model evaluation (inner model), the R-Square value was obtained at 0.548 and the Q-Square test results showed a value greater than zero, which indicates that the research model has good predictive relevance. Thus, based on the R-Square and Q-Square values, it can be concluded that Hypothesis 3 (H3) is accepted, namely there is a simultaneous influence of Live Shopping and Affiliate Marketing on Impulse Buying among Broguy Store customers on TikTok Shop. The Effect of Live Shopping on Impulse Buying The empirical findings indicate that Live Shopping exerts a beneficial and statistically significant influence on Impulse Buying, as evidenced by the path coefficient value of 0.479, accompanied by a T-statistic of 3.718 and a P-value of 0.000. These results imply that the more efficacious the implementation of live shopping, the greater the tendency of consumers to engage in impulsive buying behavior at Broguy Store. The conceptual framework underlying these findings can be explained through the key attributes of live shopping, which include instant interaction, direct visual engagement with the product, and reciprocal communication between vendor and consumer. These conditions create a captivating and emotionally resonant shopping experience, thereby diminishing consumers' rational reasoning processes. In this situation, consumers tend to respond spontaneously to stimuli, ultimately leading to impulse buying.

Furthermore, live shopping is often accompanied by time pressure, such as limited-time offers, special discounts during the broadcast, or claims of limited stock availability. These components serve to increase the perception of immediacy and the fear of possible exclusions, which, from a psychological perspective, forces consumers to make instant purchasing decisions without careful consideration. This aligns with the concept of impulse buying, which is triggered by situational and emotional stimuli. Within the Broguy Store framework, live shopping serves not only as a promotional tool but also as a mechanism to foster trust and intimacy with consumers. Direct interaction between the host and the audience allows consumers to obtain instant product information, see demonstrations of product use, and receive immediate responses to questions. This situation increases consumer confidence and accelerates the purchasing decision-making process. These findings corroborate the conclusions drawn by Badi'ah et al. (2025) and Ugroho et al. (2024), who assert that affiliate marketing has a significant influence on impulse buying behavior.

The results of this study indicate that affiliate marketing and live shopping both have a positive and significant effect on impulse buying behavior among consumers on TikTok Shop, particularly at Broguy Store. Affiliate marketing is proven to directly encourage spontaneous purchases through promotional activities carried out by affiliates, while live shopping shows a stronger influence due to its interactive, engaging, and emotionally appealing nature. The combination of real-time interaction, product visualization, and time-limited offers in live shopping creates a sense of urgency and reduces rational decision- making, leading consumers to make immediate purchases. Additionally, the model used in this research demonstrates good explanatory and predictive power, indicating that these two variables play an important role in shaping impulse buying behavior.

Based on these findings, it is recommended that online businesses optimize the use of live shopping by creating more interactive and attractive content, as well as utilizing urgency strategies such as limited-time promotions to stimulate impulse buying. Companies should also collaborate with credible affiliates to expand reach and build consumer trust. Furthermore, integrating affiliate marketing and live shopping strategies can enhance marketing effectiveness. For future research, it is suggested to include additional variables such as consumer trust, hedonic motivation, or self-control, and to expand the research scope to different platforms or industries in order to obtain more comprehensive insights.

For advanced research, future studies are encouraged to explore a more comprehensive model by incorporating additional variables such as consumer trust, hedonic shopping motivation, self-control, and perceived value to better understand impulse buying behavior. Researchers may also examine different e- commerce platforms beyond TikTok Shop, apply comparative or longitudinal approaches, and use larger, more diverse samples to enhance the generalizability and depth of findings in digital consumer behavior.