Exploring the Impact of Streamer Competencies and Situational Factors on Consumers’ Purchase Intention in Live Commerce: A Stimulus–Organism–Response Perspective

Abstract

Recently, the live commerce market has experienced rapid growth, accompanied by increasingly intense competition. To improve business performance in this dynamic environment, it is essential to foster competent streamers and create effective commerce environments. Therefore, this study developed a research model based on the stimulus–organism–response (S-O-R) framework, focusing on streamer competencies and the commerce environment, to explore ways to effectively enhance live commerce business performance. Data for this study were collected through a questionnaire and analyzed using statistical techniques with 390 respondents. The results revealed that streamers’ competencies (expertise, demonstration skills, and interactive ability) significantly influence consumers’ internal states (perceived functional value of products and perceived trust in product recommendations), which in turn significantly influence purchase intentions. Moreover, the physical surroundings of the studio and the social surroundings, including peers’ perceptions of live commerce, were found to moderate the relationships between consumers’ internal states and their purchase intentions. This study holds academic significance in that it presents a model that effectively understands the mechanisms influencing viewers’ purchase decisions in live commerce contexts. The findings and practical implications discussed in this study are expected to provide valuable insights for developing strategies to enhance the performance of live commerce.

1. Introduction

Live commerce is a cutting-edge e-commerce model based on real-time video streaming that provides a more intimate environment for consumer interaction and is rapidly gaining popularity, attracting significant attention from consumers [1,2]. According to DataHorizzon research [3], the global live e-commerce market size was valued at $187.5 billion in 2024 and is projected to reach $976.2 billion by 2033 at a compound annual growth rate (CAGR) of 20.1%. In this growth trend of global live e-commerce, China was evaluated to have the world’s largest live e-commerce market by far. As of 2024, China recorded a market size of $90 billion USD, followed by North America, which ranked second with $23.44 billion USD in live e-commerce market size. The number of live commerce consumers in China reached 620 million as of December 2024, an increase of 91.822 million from December 2023, accounting for 55.96% of all Chinese netizens [4].

The growth of live commerce can be attributed to its unique ability to replicate offline shopping experiences in an online environment, a feature that enhances consumers’ immersion and satisfaction [5]. Here, streamers serve as the core component in this replication, acting as product spokespersons and enhancing customer immersion [6,7,8]. The contribution of these streamers is the most significant advantage that the existing e-commerce sales method cannot provide, and it is the most crucial factor in the attractiveness of live commerce [9]. This attractive point has strongly appealed to consumers by effectively replacing offline shopping, especially in situations where people have been restricted from offline activities [10].

The rapid growth of live commerce has attracted numerous companies, leading to a surge in streamers. This situation has led to negative issues such as a deterioration in the quality of streamers [11]. Such issues negatively affect not only consumers’ online shopping experiences but also the overall image of the live commerce industry, ultimately raising concerns that trust in the industry may be undermined [12]. To address these concerns and enhance the competitiveness of live commerce, it is necessary to consider how streamers’ capabilities can be leveraged to strengthen business performance and sustain industry competitiveness.

Furthermore, to maximize the performance of live commerce, it is also crucial to consider environmental conditions that amplify the positive effects of streamers’ capabilities. Regarding environmental conditions, as explained above, considering that live commerce replicates offline shopping situations, physical surroundings—such as interior design and music—constitute important situational conditions. In addition, as traditional e-commerce is influenced by social situational factors like word of mouth [13], it is necessary to consider this social environment in live commerce as well. However, until recently, it has been difficult to find studies that address social situational factors in the live commerce context (see Appendix A).

Therefore, based on these motivations, this study establishes the following research questions (RQs): (RQ1) How can the performance of live commerce be improved through the capabilities of streamers? (RQ2) What are the ways to maximize the impact of streamer capabilities on live commerce performance? To address these research questions, this study sets the following research objectives. First, to answer RQ1, this study seeks to identify the specific mechanism through which streamers’ competencies influence consumers’ purchase intentions. To this end, this study approaches the topic from the perspective of the Stimulus–Organism–Response (SOR) model. The SOR model proposes that various stimulus factors influence the recipient’s perceptions or emotions, which in turn lead to behavioral responses [14]. The streamer, which is the focus of this study, is a stimulus factor that affects consumers’ purchasing desire and judgment. The purpose of this study is to analyze the mechanism by which the stimulus factor, the streamer, affects the purchase intention, which can be seen as the consumer’s behavioral response. In this way, the purpose of this study can be structured and explained more precisely and effectively through the logic of S-O-R. Therefore, this study employs the SOR model as a theoretical base. The SOR model has also been applied in other live streaming studies focusing on consumer behavior [9,15,16].

In this context, this study specifically seeks to analyze, from the S-O-R perspective, how streamer competencies as stimulus factors influence the organism factors, which refer to the recipient’s perceptions or emotions—internal states generated or shaped by the stimulus factors. This study, concerning consumers’ internal states, focuses on their perceived value of product features and their trust in streamers’ product recommendations. According to Scaglione and Mendola [17], the functional value of a product is the product’s ability to meet consumers’ needs in terms of performance and quality, which is the most basic starting point in consumers’ purchasing decisions. In the actual purchase decision process, most consumers decide to buy a product based on whether its functional value can satisfy their needs [18]. On the other hand, in the online transaction environment, consumers face increased uncertainty and transaction risks due to information asymmetry and the expansion of the transaction space; therefore, research on consumers’ online purchasing behavior tends to consider consumer trust as an essential factor [19,20]. Therefore, consumers’ perceived value of product features and trust in streamers’ recommendations can be considered key internal states that directly influence consumers’ purchase decision-making.

Second, to answer RQ2, this study seeks to examine whether the impact of consumers’ internal states—such as the perceived functional value of the product and trust in streamers’ recommendations—on purchase intention can be further enhanced by situational factors. This study considers situational factors from two primary perspectives: the physical environment of the studio and the social environment, which includes others’ perspectives or opinions regarding the product. Specifically, this study focuses on examining whether these physical and social situational factors moderate the relationship between consumers’ internal states and their purchase intention, which is the final response stage in the S-O-R model.

The following sections present the literature review, research model and hypotheses, methods, and data analysis and results. Finally, this study discusses the theoretical and practical implications, limitations, and future research directions.

2. Theoretical Background

2.1. Characteristics of Live Commerce

Live commerce, which can be regarded as a type of webcasting, operates by integrating various social business attributes with media functions for content delivery [21]. This integration incorporates real-time social interactions into e-commerce. There are two main approaches: first, embedding social live streaming into existing e-commerce platforms, as exemplified by Taobao [22], and second, integrating existing e-commerce functions into social live streaming, as observed on TikTok [23]. In essence, like traditional e-commerce, live commerce is also based on the core components of “people, products, and places” [24]. In live commerce, the “people” refers to direct interaction among various stakeholders, such as consumers, streamers, and product manufacturers. Consumers can convey various opinions or questions to streamers through chat windows during the broadcast, and streamers can respond to them. Product manufacturers invited to the live room can also interact with streamers or consumers. Sometimes, consumers interact with each other via chat windows. These diverse interactions create a community-centered shared experience and enhance consumers’ emotional engagement [25,26].

As shown in

Table 1. The different characteristics between live commerce and traditional e-commerce.

Aspect	Live Commerce	Traditional E-Commerce
Displayed content	Streamers’ interaction with users	Sales Page and graphic information
Interaction characteristic	Real-time interaction with streamers (One-to-one/one-to-many)	Chatting messages with sellers (one-to-one)
Format of the information	Image/text/video/live content/VR	Image/text/video
Source of information	Non-editable	Editable
Purchase method	Buy-while-watching streamers	Final purchase decision

, live commerce has the following characteristics compared to traditional e-commerce: it improves communication efficiency through real-time interactions between people [16], enhances consumers’ understanding by providing more diverse forms of information [1], and strengthens the perception of product authenticity by offering real-time, non-editable information [27]. These characteristics are closely related to streamers. In live commerce, streamers play an important role in several key aspects. Existing e-commerce primarily conveys product information graphically. However, graphic information often emphasizes what favors sellers. This can lead to unbalanced information and increase the risk of information asymmetry [21]. In contrast, streamers convey rich product information in real time. This allows consumers to collect comprehensive information quickly and reduces the risk of information asymmetry [28]. These characteristics of live commerce align better with consumer needs [29]. Another key aspect is that consumers can interact with streamers in real time [30]. Through this interaction, streamers can demonstrate products while answering consumers’ questions. This gives consumers an experience similar to trying the products themselves. Considering these points, it is necessary to focus on the role of streamers for developing strategies to improve live commerce performance.

2.2. Competencies of Live Commerce Streamers

As live commerce continues to grow rapidly, streamers have become increasingly important as a link between consumers and merchants (or products) [31,32]. Liao et al. [33] defined live commerce streamers as individuals who consciously craft and curate their online identities to sell and monetize through self-presentation or merchandising on live commerce platforms. They generate revenue by attracting followers and influencing consumer behavior [32]. To succeed in this role, streamers need strong live commerce capabilities. In this context, researchers have found that expertise and interactive ability positively affect consumers’ purchase intention [8,9,15,16].

Expertise refers to the level of understanding and knowledge the streamers have about the product [34]. In traditional e-commerce situations, consumers often spend significant time researching products on their own, especially to understand unfamiliar professional characteristics [31]. In contrast, in live commerce, consumers can quickly understand product or service characteristics due to streamers’ expertise [35]. Generally, the expertise of the information source influences the acceptability of the information to the receiver [36]. In this context, information receivers tend to trust expert opinions more than those of ordinary people [37].

In addition, streamers’ interactive ability is crucial for business performance in the live commerce environment [38]. Here, the interactive ability refers to streamers’ capacity to understand customers’ needs and effectively resolve their questions or issues through interactions [39]. Hassanein and Head [40] argued that human interaction in a medium makes users feel a sense of warmth. In traditional e-commerce, sellers are usually unable to guide customers in a timely manner, and this lack of interaction can make customers doubt product authenticity [9]. In contrast, live commerce enables streamers to offer real-time interactions, which can enhance consumers’ purchase intentions [38].

Another essential role of streamers is to demonstrate products to consumers through real-time interaction based on their expertise [31]. The streamer should enhance consumers’ perception of the product’s value by demonstrating how to use it effectively. However, Walton [41] noted that experts often struggle to convey their knowledge to non-expert audiences. Expertise is crucial for delivering messages, but does not always make them compelling or persuasive [42]. For instance, some streamers focus too much on technical details and fail to communicate in an understandable and engaging way [31]. This implies that streamers must also have the skill to demonstrate products in a way that meets audience needs.

Nevertheless, previous studies have focused simply on demonstrating effectiveness rather than on the streamers’ ability to demonstrate. For example, Elder and Krishna [43] argued that streamers enhance consumers’ perception of product authenticity by vividly showing product usage through visual and auditory methods. Guo et al. [21] pointed out that streamers enhance perceived authenticity by showing all aspects of a product. This compensates for traditional online shopping, where demonstrations may rely on embellished photos. Similarly, Akdevelioglu and Kara [44] argued that streamers effectively demonstrate product features, which improves audience understanding. Considering the effects on demonstration behavior, this study suggests that demonstration ability should be considered an essential skill for streamers. Drawing on Beverland et al.’s [45] concept of “pure (literal) authenticity”, which involves providing consumers direct, observable evidence of a product’s attributes in its natural state through unmediated cues, we define demonstration skills as a streamer’s ability to convey the necessity and usefulness of a product to viewers without embellishment.

To clarify the theoretical boundaries of streamer competencies, this study distinguishes demonstration skills from other concepts. As we discussed above, demonstration skills are central. However, this construct may be confused with related concepts such as presentation quality, vividness, and argument quality. To address this concern, we develop a detailed contrast table that clearly defines what demonstration skills uniquely capture (see Appendix B). In summary, demonstration skills distinguish themselves from related constructs by focusing on the active, authentic portrayal of a product’s necessity and usefulness. Unlike expertise, which is knowledge-based, demonstration skills are about the practical application and communication of that knowledge. While interactive ability emphasizes two-way audience engagement, demonstration skills are a proactive, product-centric performance. It differs from the overall broadcast aesthetics of presentation quality by focusing on the specific, micro-level act of showcasing the product. Lastly, it transcends imaginative simulation (vividness) and mere verbal persuasion (argument quality) by offering tangible proof of the product’s utility through actual use.

2.3. Situational Factors: Physical Surroundings and Social Surroundings

Situational factors are observable environmental factors, beyond individual attributes, that can affect individuals physiologically or psychologically and influence consumption behavior [46]. Belk [46] first proposed that situational factors impact consumer behavior. He categorized them into five types: physical surroundings, social surroundings, temporal perspective, task definition, and antecedent states. Physical surroundings refer to the spatial environmental elements where consumption activities occur, such as store decorations, background music, and lighting. Social surroundings are external factors related to individuals or groups, mainly peers, including companionship, peer personalities, and interactions. Temporal perspective involves time-related factors, like the sufficiency of purchase time. Task definition describes the buyer’s motivation, including the purchase purpose, such as for personal use or as a gift. Lastly, antecedent states refer to the consumer’s state before making a purchase decision, encompassing emotional or economic conditions like anxiety, excitement, or available funds.

According to Gehrt and Yan [47], it is not necessary to select all five situational factors but rather to choose those most relevant to the research purpose. Recently, researchers have selectively applied these five situational factors to study consumer behavior in offline stores and online shopping websites. For example, Rehman et al. [48] studied social and physical surroundings in offline clothing stores. They found that both factors positively affect consumers’ buying intentions. Hussain and Ali [13] found that physical surroundings, including cleanliness, scent, lighting, and display/layout, positively influence consumers’ willingness to buy in international retail chain shops in Karachi, Pakistan. Odeh and As’ ad [49] conducted a field study in a Jordanian shopping mall. They found that ambient conditions (air quality, lighting, music, scent), design factors (layout, furnishing, décor), and social factors (personal, noise, signs & symbols) significantly affect consumer purchasing behavior. Muhammad et al. [50] studied situational factors on online grocery shopping websites in Malaysia and found that antecedent states and temporal perspective have a positive relationship with online grocery shopping adoption.

However, research on situational factors in live commerce remains limited. Given the real-time, interactive, and digitally mediated nature of live streaming, this study focuses on physical and social surroundings. First, live streaming rooms combine curated visual and auditory elements with immediate social interactions, creating situational effects that differ from classic retail atmospherics [9]. According to Hussain and Ali [13] and Rehman et al. [48], the physical surroundings factors in offline stores can directly impact consumers’ perception of products. In live commerce, a well-decorated live streaming room that displays products in a visually appealing manner and features good music can have a direct positive effect on consumer perception. Hence, the physical environment factors being perceived in real-time are examined in this study.

In addition, consumers experience real-time chats with the streamer and peers, while asynchronous cues—such as friend recommendations, post-stream discussions, and social media sharing also influence consumers’ purchase behavior [9]. Except for accidental visits to live streaming rooms, consumers generally access them through friends’ recommendations. If a friend positively reviews a product or encourages an order, this social environment prompts more consumers to place an order [48]. Therefore, considering the characteristics of live commerce, where the influence of the ‘social environment’ is strongly exerted, these social surroundings factors are treated as significant variables in this study.

Meanwhile, although temporal perspective, task definition, and antecedent states are traditionally important situational factors, their influence in live commerce is relatively limited. In traditional online shopping, consumers often spend more time researching product information and comparing different products independently [15]. In contrast, in live streaming rooms, consumers can quickly acquire product information and make purchase decisions through streamers’ real-time demonstration [8]. Even during limited-time discounts, consumers can interact freely with streamers without feeling rushed [16]. Therefore, the temporal perspective factor is relatively weak in live-streaming contexts. Similarly, task definition appears to play a limited role in the live shopping context. Even with a clear purchasing goal, consumers often buy additional products impulsively or change their original decisions due to events like time-limited promotions [15]. Furthermore, antecedent states, which reflect consumers’ emotional or financial conditions before purchase [46], are difficult to measure accurately. Data collection often occurs long after purchase, introducing memory bias that makes it hard for consumers to recall pre-purchase emotions accurately. Additionally, some consumers may hesitate to share their true emotional or financial state, especially negative ones.

Based on the above discussion, this study focuses on physical and social factors. This study aims to verify whether these factors enhance consumer perceptions and purchase intentions in live commerce. In this study, physical surroundings refer to the live streaming room environment, including the visible interior design (e.g., layout, colors, and lighting) and other invisible factors (e.g., sound and background music) [51]. Social surroundings refer to the situation where consumers interact and empathize with friends about their experiences or opinions about live commerce shopping [51].

2.4. Stimulus–Organism–Response Model

The stimulus–organism–response (S-O-R) model, proposed by Mehrabian and Russell [14], explains how external stimuli shape cognitive and emotional states that in turn drive behavioral responses. In this model, “stimulus” refers to an external factor that influences an individual’s cognitive or affective processes, “organism” denotes the individual’s internal state, and “response” represents the resulting behavioral responses [14]. This model places an individual’s internal state between stimulus and final response, focusing on explaining the connection between the two [14]. This model is recognized for its value as it serves as a logical and useful framework for understanding how individuals are influenced by external stimuli and ultimately exhibit specific behaviors or results [52]. Specifically, the SOR model has been validated to successfully explain human behavior in various contexts, such as short video advertisements [52], hotel and tourism [53], VR shopping for sustainable management [54], and social commerce [55,56,57].

In live commerce, the S-O-R model has been applied to explain how streamer characteristics shape consumers’ purchase intentions. For example, Zhong et al. [9] investigated how stimuli such as expertise and interaction affect trust and, in turn, purchase intentions. Sun et al. [16] demonstrated that interactive stimuli from Internet celebrity anchors affect attitudes toward purchase intentions. Wang et al. [58] found that attractiveness and interactivity shape social presence and flow experience, thereby enhancing purchase intentions. Like these studies, this study focuses on analyzing consumer behavior in the context of live commerce and thus uses the S-O-R model as a research framework. However, unlike previous studies, this study specifically considers the capabilities of streamers as key stimuli in the SOR model. It has already been pointed out that streamers’ ability plays an important role in promoting consumers’ purchasing behavior [9]. In addition, this study conceptualizes trust and perceived product value as organismic states, reflecting consumers’ internal psychological processes. Through this SOR model-based approach, this study seeks to reveal another psychological mechanism through which consumers form purchasing intentions in the context of live commerce.

3. Research Model and Hypotheses

3.1. Research Model

The purpose of this study is to investigate how consumers’ perceptions of streamer competencies influence purchase intentions based on the S-O-R model. On this basis, it aims to explore strategies for enhancing streamer competencies and improving business performance. To achieve this, as shown in Figure 1, this study first developed a basic model that considers streamer competencies as the stimulus (S), consumers’ perceptions of the functional value of products and their trust in streamers’ product recommendations as the organism (O), and consumers’ purchase intentions as the response (R). Additionally, to explore how the physical and social situational factors discussed in Section 2.3 strengthen the effect on consumers’ internal states (i.e., the organism) on purchase intentions (i.e., the response), these situational factors are incorporated into the research model as moderating variables.

A detailed explanation of this research model is as follows. From the S-O-R perspective, streamer competencies can be seen as stimulus factors that influence consumers’ desires and judgments. First, this study aims to analyze how streamer competencies as stimulus factors influence the organism factors, which refer to the recipient’s perceptions or emotions—internal states shaped by the stimulus factors. In this study, consumers’ internal states focus on the perceived value of product features and trust in streamers’ product recommendations. According to Scaglione and Mendola [17], the functional value of a product is defined as its ability to satisfy consumer needs regarding performance and quality, serving as a fundamental basis for purchase decisions. In actual purchase processes, most consumers decide whether to buy a product based on whether its functional value satisfies their needs [18]. Meanwhile, in online transactions, consumers face greater uncertainty and transaction risks due to information asymmetry and the expansion of transaction spaces, so consumer trust is considered an essential factor in online purchase behavior studies [19,20]. Therefore, consumers’ perceived value of product features and trust in streamer recommendations can be regarded as key internal states that directly influence purchase decisions. This study aims to identify significant roles of competencies by verifying their effects on internal states.

Furthermore, this study seeks to examine whether situational factors can further enhance the impact of consumers’ internal states on purchase intention. In this study, situational factors are examined from two main perspectives: the physical environment of the studio and the social environment, including the perspectives or opinions of others regarding the product. This study seeks to test whether these physical and social situational factors have a moderating effect on the relationship between consumers’ internal states and purchase intention, which is the response in the S-O-R model.

3.2. Live Commerce Streamers’ Competencies and Consumers’ Internal States

Live commerce streamers‘ expertise significantly shapes consumers’ perceptions of product price and performance [34]. For example, streamers demonstrate their expertise by advising consumers on how to mix and match products and informing them of the maximum discounts available for different combinations [9]. As Lee and Chen [15] elucidated, professionalism involves streamers possessing accurate and comprehensive knowledge about products. However, when relevant and accurate product information is lacking, consumers may abandon their purchase intentions or delay their orders, underscoring the critical role of expertise in the purchase process [59,60]. In this context, Lou and Yuan [61] argued that greater domain-specific expertise enables influencers to deliver more accurate information, thereby enhancing consumers’ perceived functional value. Similarly, Chen et al. [62] suggested that streamers’ expertise enhances consumers’ perceived utilitarian value of products. Furthermore, the connection between streamers’ expertise and consumers’ perceived trust is supported by various findings. For example, Li et al. [63] demonstrated that streamers’ expertise during product promotion significantly increases consumers’ trust. Huanyu et al. [64] also pointed out that expert streamers who exhibited professional behavior in product recommendations strengthen consumers’ trust perception. Based on these discussions, we propose the following hypotheses:

H1a. 

The expertise of live commerce streamers positively affects consumers’ perceived functional value of products.

H1b. 

The expertise of live commerce streamers positively affects consumers’ perceived trust in streamers’ product recommendations.

In a live shopping environment, real-time product presentation by streamers enriches consumers’ perception of product information [28]. This real-time demonstration approach, highlighted by Sun et al. [16], bridges the gap between the digital realm and physical reality by providing consumers with a simulated experience of products. This customer experience can be understood as an experiential marketing process, and Herrera [65] argued that it significantly increases consumer engagement. From the perspective of experiential marketing, such demonstrations not only convey functional information but also provide vivid visual cues that generate strong sensory stimulation [66]. This stimulation increases consumers’ sense of immersion in the shopping experience [67]. In addition, effective demonstrations present product features clearly, reducing information-processing difficulty. According to consumer psychology, this promotes cognitive ease, enabling consumers to process information with less effort, reduces uncertainty, and makes information appear more credible [68].

Unlike traditional online shopping, live commerce conveys stronger product authenticity because content is unedited and delivered in real time. However, its instantaneous nature can also trigger concerns about missing key information [69]. Nevertheless, streamers mitigate this risk by visualizing product functions and benefits, thereby enhancing perceived functional value and motivating purchases [70].

Furthermore, product authenticity, as demonstrated through the product demonstration process, also affects consumers’ trust in products [71]. According to Ang et al. [28], streamers’ real-time and efficient demonstration of products can enhance consumers’ perceived trust in products. Similarly, Zhong et al. [9] argued that the trial of a product by a streamer enhances consumer trust. In addition, Li et al. [72] pointed out that high-definition displays and trial results of products increase consumer confidence in products. Based on these discussions, we propose the following hypotheses:

H2a. 

The demonstration skills of live commerce streamers positively affect consumers’ perceived functional value of products.

H2b. 

The demonstration skills of live commerce streamers positively affect consumers’ perceived trust in streamers’ product recommendations.

Consumers evaluate their perception of service quality based on their subjective expectations, which are influenced by both the verbal and nonverbal behaviors of salespeople [73]. As highlighted by Fang [74], one of the pivotal features of live commerce is the interaction between consumers and streamers. Sun et al. [16] emphasized that streamers facilitate consumers’ understanding of product functional features by answering consumer questions in a timely manner. Such interactive behaviors promote consumers’ perceived functional value of products. Moreover, Zhou and Huang [75] found that heightened streamer–consumer interactions enhance consumers’ perception of a product’s functional and value, ultimately driving stronger purchase intentions. Simultaneously, the interactive ability of streamers emerges as a significant factor positively impacting consumers’ trust [64]. In this context, the interactive ability of streamers is recognized as a critical factor that positively influences consumer trust [64]. Chen et al. [76] also argued that streamers’ frequent and efficient interactions with consumers significantly bolster viewers’ trust in streamers’ recommendations. Based on these discussions, we develop the following hypotheses:

H3a. 

The interactive ability of live commerce streamers positively affects consumers’ perceived functional value.

H3b. 

The interactive ability of live commerce streamers positively affects consumers’ perceived trust in streamers’ product recommendations.

3.3. Consumers’ Internal States and Consumers’ Purchase Intention

Perceived functional value is a primary factor influencing consumers’ purchase intentions [77]. Parasuraman [78] found that consumers’ perceived value can influence their consumption intentions. Additionally, O’Neal [79] proposed that the perceived functional value of products continues to play a pivotal role in consumers’ decision-making for subsequent purchases. According to Chattalas and Shukla [80], the perception of a product’s functional value enhances consumers’ luxury purchase intentions among British and American consumers. Lakhan et al. [81] also argued that consumers’ perceived functional value on live streaming platforms positively influences their purchase intention. Based on these discussions, we propose the following hypothesis:

H4. 

Consumers’ perceived functional value of products positively affects consumers’ purchase intention.

Live commerce streamers play a pivotal role in fostering consumer trust regarding the products or services they promote. Researchers such as Gefen et al. [19] demonstrated a direct link between trust in online product recommendations and consumers’ willingness to purchase. In line with this, Hsiao et al. [82] found a significant impact of consumers’ trust in product recommenders on their purchase intentions. Additionally, Zafeiropoulou [83] argued that trust, as a social relationship, positively influences consumers’ purchasing behavior. In the live commerce context, Sun et al. [16] and Zhong et al. [9] also found that consumers’ trust in products significantly promotes their purchase intention. Based on these discussions, we propose the following hypothesis:

H5. 

Consumers’ perceived trust in streamers’ product recommendations positively affects consumers’ purchase intention.

3.4. Role of Situational Factors: Moderating Effects

Physical factors such as shop design, background music, and layout positively influence purchase intentions [48,51]. In live commerce, a well-designed live room fosters a comfortable atmosphere, enhancing consumers’ immersive and dwell time [72]. Longer dwell time enables consumers to acquire detailed product information, enhancing perceived functional value and purchase intention [13,48]. From the perspective of environmental psychology, immersion directs consumers’ attention to product cues and keeps them engaged in the experience [84], while a comfortable environment reduces stress and makes product information easier to process [72]. When consumers feel comfortable and at ease, they are more likely to perceive product information and to develop trust in the seller’s recommendations [85]. Additionally, a comfortable and well-designed shopping scenario can enhance the consumer’s perceived aesthetics [86]. Consumers tend to enter a comfortable and well-designed shopping environment, which leads consumers to perceive that sellers value their shopping experience. Based on these discussions, we propose the following hypotheses:

H6a. 

Physical Surroundings positively moderate the relationship between consumers’ perceived functional value of products and consumers’ purchase intention.

H6b. 

Physical Surroundings positively moderate the relationship between consumers’ perceived trust in streamers’ product recommendations and consumers’ purchase intention.

The social surroundings formed through interactions with others can affect consumers’ evaluation of products, which can ultimately have a significant impact on consumers’ purchase intentions [48]. Specifically, word-of-mouth information from friends reduces consumers’ uncertainty about products and shapes their expectations of products [87]. From a social influence perspective, consumers often follow friends’ opinions through conformity, which helps them feel more confident in their choices and reduces decision risk [88]. In addition, social identity theory suggests that when consumers see themselves as part of a group, purchasing the same products strengthens their sense of emotional connection with the group, which in turn enhances their purchase decisions [89]. When recommended products align with consumers’ needs, this alignment enhances their focus on the products and engagement with the information [89]. Based on these discussions, we propose the following hypotheses:

H7a. 

Social Surroundings positively moderate the relationship between consumers’ perceived functional value of products and consumers’ purchase intention.

H7b. 

Social Surroundings positively moderate the relationship between consumers’ perceived trust in streamers’ product recommendations and consumers’ purchase intention.

4. Methods

4.1. Measures

The measurement items for the components used in this study are shown in Table 2. Among the constructs, the measurement items for expertise, interactive ability, functional value of products, trust in product recommendations, and purchase intention were employed from previously validated scales in existing studies. Since no established measurement items exist in prior research for the constructs of demonstration skills, physical surroundings, and social surroundings, this study developed the measurement items directly based on relevant theories.

The measurement items for demonstration skills were developed in this study based on the concept of “pure (literal) authenticity” introduced in Section 2.2. This concept emphasizes the value of providing consumers with an unedited view that presents a product’s attributes as they are, without exaggeration [45]. For example, to convey pure authenticity in beer advertising, Beverland et al. [45] suggested that it is effective to directly show how brewers apply traditional methods to beer production. This can be done through photos of using traditional equipment to produce beer, storing beer in cellars, and serving beer with traditional costumes. Such visual elements reinforce the true nature of the product and provide a factual connection to its origins and intended use.

In the traditional e-commerce context, pure authenticity is often undermined by decoration and image manipulation, as heavily edited visuals on shopping websites can create unrealistic images. As Mavlanova and Benbunan-Fich [90] warned, this mismatch between online images and reality can erode consumer trust. However, in the context of live commerce, streamers can transparently convey the genuine aspects of a product (i.e., its pure authenticity) to viewers through their demonstration skills. Such demonstration skills that effectively deliver transparency enable viewers to assess the product’s actual performance, materials, and functions without being misled by exaggeration or distortion. Based on this perspective, this study developed four measurement items. These items assessed how effectively a streamer conveys a product’s pure authenticity in a clear, persuasive, and impressive manner, thereby enabling consumers to confidently judge its functions and intrinsic value.

For measuring physical and social surroundings, prior studies such as Tong et al. [51] provided ready-to-use scales in offline contexts. Tong et al. [51], in the context of women’s shoe stores, measured physical surroundings with items such as impressive exterior and interior design, a large store, a pleasant odor, music, and a bright environment, and measured social surroundings with a single item, “shopping with friends.” However, some of the existing items from an offline environment were not directly transferable to a live streaming room, as certain metrics, such as nice odors, cannot be perceived by online consumers. Accordingly, in this study, the measurement items for the physical environment were adapted from the relevant items of Tong et al. [51] to better fit the live streaming shopping context. These adaptations focused on aspects of the live streaming room that consumers can directly perceive, such as the sophistication and luxury of the interior design and the quality of the music. In contrast, for the social environment, since Tong et al. [51] measured this construct using a single item, all measurement items were newly developed in this study to assess the extent to which friends share and discuss their experiences or opinions about live commerce shopping. All items in this study were measured on a seven-point Likert scale ranging from “strongly disagree” to “strongly agree”.

As explained earlier, this study independently developed measurement items for the variables of demonstration skills and social surroundings. When new measurement items are developed, it is essential to verify their appropriateness through a pilot test [91]. In this study, the pilot test was conducted in two stages. In the first stage, the study collected responses from 150 participants with live commerce shopping experience through an offline survey and conducted exploratory factor analysis (EFA) on all measurement items using SPSS 25. The results showed that the Kaiser–Meyer–Olkin (KMO) value was 0.929 and Bartlett’s test of sphericity yielded a significant level (p < 0.05), indicating that the samples were suitable for factor analysis [92]. Next, to extract the dimensions of scale, we used principal components analysis with varimax orthogonal rotation with eigenvalues greater than 1.0 [93]. As shown in Table 2, after EFA, a scale comprising eight unidimensional factors with a total of 28 measurement items emerged, each with coefficients greater than 0.40 and not double-loaded onto multiple factors [93]. The results of EFA showed that 8 factors extracted can explain 62.786% of the total variance, which exceeded 60% [94].

In the second stage, to validate the factor structure of latent constructs from the EFA procedures, we conducted confirmatory factor analysis (CFA) by using another sample of 214 respondents in SmartPLS 4.0. We examined reliability, convergent validity, and discriminant validity of the measurement scale. The Cronbach’s alpha values of all variables ranged from 0.763 to 0.953, and the composite reliability ranged from 0.814 to 0.910, exceeding the cut-off value of 0.7 and indicating a secure reliability [95]. Convergent validity was assessed by examining the factor loading and average variance extracted (AVE) for each dimension. All factor loadings were higher than 0.7 [95], and the values of AVE for each dimension were higher than 0.5 [96], indicating that the measurement model has substantial convergent validity. Discriminant validity was confirmed by the Fornell–Larcker criterion [97]. The result showed that the correlation coefficients between constructs were lower than the square root of the AVE, indicating substantial discriminant validity [97]. The model fit also showed satisfactory results: χ² = 854.23, d.f. = 324, χ²/d.f. = 2.64, RMSEA = 0.049, GFI = 0.975, NFI = 0.954, CFI = 0.976 [94].

Table 2. Measurement Items for Constructs.

Constructs	Measurement Items	Sources
Expertise (PE)	PE1: Streamers are very knowledgeable about products.	Chen et al. [34]
	PE2: Streamers are experts in products.
	PE3: Streamers have a high understanding of the product.
Demonstration Skills (DSs)	DSs1: Streamers persuasively communicate the necessity and usefulness of the product.	Self-development
	DSs2: Streamers impressively explain the necessity and usefulness of the product.
	DSs3: Streamers clearly explain the necessity and usefulness of the product.
	DSs4: Streamers effectively explain the necessity and usefulness of the product.
Interactive Ability (IA)	IA1: Streamers listen to customers attentively to get a proper understanding of their specific needs.	Homburg et al. [39]
	IA2: Streamers are very committed to resolving disagreements with customers.
	IA3: Streamers adapt their sales pitch very much to customers’ interests.
Functional Value of Products (PV)	PV1: Products are generally of good quality.	Sweeny and Soutar [98]
	PV2: Products are generally well-made.
	PV3: Products are reasonably priced.
	PV4: Products offer value for money.
Trust in Product Recommendations (PT)	PT1: I think that streamers’ product recommendations are credible.	Hsiao et al. [82]
	PT2: I trust streamers’ product recommendations.
	PT3: I believe that streamers’ product recommendations are trustworthy.
Purchase Intention (PI)	PI1: I would like to purchase products through Live commerce.	Chen et al. [34]
	PI2: I would like to recommend that my friends and family purchase products through live commerce.
	PI3: If I want to purchase a product, I would prefer to do so through live commerce.
	PI4: I intend to purchase products through live commerce.
	PI5: I expect that I will purchase products through live commerce.
Physical Surroundings (PS)	PS1: When the interior design of the live streaming room is refined, I want to purchase the product even more.	Tong et al. [51]
	PS2: When the interior design of a live streaming room looks luxurious, I want to purchase the product even more.
	PS3: When the music in the live streaming room is good, I want to purchase the product even more.
Social Surroundings (SS)	SS1: My friends enjoy sharing their experiences or opinions about live commerce shopping with each other.	Self-development
	SS2: My friends enjoy discussing their experiences and opinions about live commerce shopping with each other.
	SS3: My friends are interested in each other’s experiences or opinions about live commerce shopping.

Notes: All constructs were measured on a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree).

Table 2. Measurement Items for Constructs.

Constructs	Measurement Items	Sources
Expertise (PE)	PE1: Streamers are very knowledgeable about products.	Chen et al. [34]
	PE2: Streamers are experts in products.
	PE3: Streamers have a high understanding of the product.
Demonstration Skills (DSs)	DSs1: Streamers persuasively communicate the necessity and usefulness of the product.	Self-development
	DSs2: Streamers impressively explain the necessity and usefulness of the product.
	DSs3: Streamers clearly explain the necessity and usefulness of the product.
	DSs4: Streamers effectively explain the necessity and usefulness of the product.
Interactive Ability (IA)	IA1: Streamers listen to customers attentively to get a proper understanding of their specific needs.	Homburg et al. [39]
	IA2: Streamers are very committed to resolving disagreements with customers.
	IA3: Streamers adapt their sales pitch very much to customers’ interests.
Functional Value of Products (PV)	PV1: Products are generally of good quality.	Sweeny and Soutar [98]
	PV2: Products are generally well-made.
	PV3: Products are reasonably priced.
	PV4: Products offer value for money.
Trust in Product Recommendations (PT)	PT1: I think that streamers’ product recommendations are credible.	Hsiao et al. [82]
	PT2: I trust streamers’ product recommendations.
	PT3: I believe that streamers’ product recommendations are trustworthy.
Purchase Intention (PI)	PI1: I would like to purchase products through Live commerce.	Chen et al. [34]
	PI2: I would like to recommend that my friends and family purchase products through live commerce.
	PI3: If I want to purchase a product, I would prefer to do so through live commerce.
	PI4: I intend to purchase products through live commerce.
	PI5: I expect that I will purchase products through live commerce.
Physical Surroundings (PS)	PS1: When the interior design of the live streaming room is refined, I want to purchase the product even more.	Tong et al. [51]
	PS2: When the interior design of a live streaming room looks luxurious, I want to purchase the product even more.
	PS3: When the music in the live streaming room is good, I want to purchase the product even more.
Social Surroundings (SS)	SS1: My friends enjoy sharing their experiences or opinions about live commerce shopping with each other.	Self-development
	SS2: My friends enjoy discussing their experiences and opinions about live commerce shopping with each other.
	SS3: My friends are interested in each other’s experiences or opinions about live commerce shopping.

Notes: All constructs were measured on a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree).

4.2. Mian Survey and Samples

We conducted the main survey by using one of the largest professional survey data collection websites in China, the Wenjuanxing website (https://www.wjx.cn/). We distributed the questionnaire to friends and WeChat groups through QR codes and links. To ensure the suitability of potential respondents, a pre-screening question was added to determine whether respondents had a live commerce shopping experience. As a result, we received responses from 432 respondents who had purchase experience through live commerce. After excluding the questionnaires that were judged to be insincere, 390 questionnaires were finally used in the analysis of this study.

The demographic characteristics of participants are shown in

Table 3. Sample characteristics.

Characteristics	Options	No. (n = 390)	Percentage
Gender	Male	149	38.21%
	Female	241	61.79%
Age	<19	6	1.54%
	19–24	154	39.49%
	25–29	132	33.85%
	30–40	74	18.97%
	>40	24	6.15%
Educational Level	Junior high school and below	11	2.82%
	High school	52	13.33%
	Undergraduate	291	74.62%
	Graduate and above	36	9.23%
Occupation	Students	212	54.36%
	Company employee	96	24.62%
	Profession (Teachers, Researchers, Doctors, and so on)	69	17.69%
	Self-employed person	8	2.05%
	Others	5	1.28%
Monthly Income (RMB)	<3000	192	49.23%
	3000–5999	104	26.67%
	6000–9000	78	20.00%
	>9000	16	4.10%
Live Streaming Watching Experience	Less than 6 months	10	2.56%
	6 months–1 year	32	8.21%
	1–1.5 years	86	22.05%
	1.5–2 years	119	30.51%
	More than 2 years	143	36.67%
Purchase Frequency (Per Month)	<3	42	10.77%
	3–6	172	44.10%
	7–10	142	36.41%
	>9	34	8.72%

. The gender distribution was predominantly female (n = 241, 61.79%). Most of the respondents were aged either 19–24 (n = 154, 39.49%) or 25–29 (n = 132, 33.85%), and most had a bachelor’s degree (n = 291, 74.62%). The highest proportion of respondents were students (n = 212, 53.36%) and had a monthly income of <3000 RMB (n = 192, 49.23%). This reflects a growing interest in value for money in live streaming products, as well as the availability of affordable options aimed at lower-income buyers. In terms of user experience, most respondents had more than two years of live streaming experience (n = 143, 36.67%) and made purchases 3–6 times per month (n = 172, 44.10%). These figures show that live commerce users are highly active, suggesting that it is a subject worth studying.

5. Data Analysis and Results

5.1. Measurement Model Assessment

This study employed the partial least squares structural equation modeling (PLS-SEM) approach for statistical data analysis, utilizing SmartPLS 4.0 as the analysis software. This study evaluated the reliability and validity of the measurement model. First, reliability was evaluated based on whether Cronbach’s alpha and composite reliability were 0.7 or higher, which indicates internal consistency [95]. As shown in

Table 4. Results of reliability and convergent validity assessment.

Constructs	Items	Factor Loadings	Cronbach’s Alpha	Composite Reliability	AVE
Expertise (PE)	PE1	0.834	0.809	0.887	0.724
	PE2	0.841
	PE3	0.876
Demonstration Skills (DSs)	DS1	0.825	0.790	0.864	0.613
	DS2	0.794
	DS3	0.727
	DS4	0.784
Interactive Ability (IA)	IA1	0.839	0.797	0.880	0.711
	IA2	0.849
	IA3	0.841
Functional Value of Products (PV)	PV1	0.877	0.859	0.914	0.780
	PV2	0.843
	PV3	0.747
	PV4	0.839
Trust in Product Recommendations (PT)	PT1	0.889	0.846	0.897	0.686
	PT2	0.876
	PT3	0.885
Purchase Intention (PI)	PI1	0.839	0.898	0.925	0.710
	PI2	0.833
	PI3	0.860
	PI4	0.854
	PI5	0.827
Physical Surroundings (PS)	PS1	0.866	0.793	0.879	0.708
	PS2	0.847
	PS3	0.810
Social Surroundings (SS)	SS1	0.892	0.864	0.917	0.785
	SS2	0.868
	SS3	0.898

, the Cronbach’s alpha values of all variables were 0.790 or higher, and the minimum value for composite reliability was 0.864, indicating that the measurement tool used in this study was evaluated to have secured reliability for all constructs.

Next, validity was evaluated from two perspectives: convergent validity and discriminant validity [95]. Convergent validity was evaluated through factor loading and average variance extracted (AVE). As shown in Table 4, the factor loading values were all 0.7 or higher [95] and the AVE values were all 0.5 or higher [96], indicating that convergent validity was secured. Discriminant validity was evaluated using the Fornell–Larcker criterion and the heterotrait–monotrait (HTMT) ratio criterion. The results of the analysis by Fornell–Larcker criterion are shown in

Table 5. Results of Fornell–Larcker criterion analysis for discriminant validity assessment.

	PE	DSs	IA	PV	PT	PI	PS	SS
Expertise (PE)	0.851
Demonstration Skills (DSs)	0.684	0.783
Interactive Ability (IA)	0.561	0.535	0.843
Functional Value of Products (PV)	0.462	0.438	0.452	0.828
Trust in Product Recommendations (PT)	0.478	0.481	0.432	0.730	0.883
Purchase Intention (PI)	0.437	0.484	0.399	0.626	0.674	0.843
Physical Surroundings (PS)	0.333	0.404	0.335	0.475	0.528	0.616	0.841
Social Surroundings (SS)	0.372	0.445	0.358	0.514	0.539	0.625	0.608	0.886

Notes: The values bolded on the diagonal are the square root of AVE values.

. All the square roots of the average variance extracted were higher than the correlation coefficients with other variables, so it was determined that the discriminant validity was secured [97]. Additionally, the HTMT ratio criterion results are presented in

Table 6. Results of the Heterotrait–Monotrait (HTMT) ratio criterion analysis for assessing discriminant validity.

	PE	DSs	IA	PV	PT	PI	PS	SS
Expertise (PE)
Demonstration Skills (DSs)	0.854
Interactive Ability (IA)	0.696	0.670
Functional Value of Products (PV)	0.557	0.531	0.547
Trust in Product Recommendations (PT)	0.573	0.577	0.518	0.854
Purchase Intention (PI)	0.511	0.571	0.467	0.716	0.767
Physical Surroundings (PS)	0.415	0.513	0.415	0.576	0.640	0.730
Social Surroundings (SS)	0.444	0.537	0.427	0.599	0.623	0.707	0.731

. All the HTMT values were below the critical value of 0.9, confirming secured discriminant validity [99].

Additionally, Harman’s single-factor test was employed to assess the risk of common method bias [100]. The results showed that the first unrotated factor explained 39.564% of the total variance, which is below the 50% threshold. This suggests that common method bias will not affect the results of this study.

Finally, the risk of potential endogeneity was further evaluated using the Gaussian copula approach [100]. We first conducted Kolmogorov–Smirnov tests with Lilliefors correction on the latent variable scores used as independent variables in the partial regression of the PLS path model [101]. The results indicated that all the constructs followed a non-normal distribution, which allowed us to apply the Gaussian copula approach [100]. The results showed that all Gaussian copulas, as well as all other combinations of Gaussian copulas included in the model, were non-significant (p > 0.05). Therefore, we concluded that endogeneity is not a concern in this study [102].

5.2. Structural Model Assessment

To evaluate the structural model, this study first conducted a collinearity analysis. The results showed that the VIF values of all constructs were in the range of 1.460 to 2.768, which were lower than the acceptable standard of 5.0 [95], confirming that there was no collinearity problem. Next, this study conducted hypothesis testing using the bootstrapping method using 5000 resamples. As shown in Figure 2 and

Table 7. Hypothesis testing results of structural model path coefficients.

Hypotheses	Paths	Path Coefficients (β)	T Values	p Values	95% Confidence Intervals	Supported?
H1a	Expertise (PE) -> Functional Value of Products (PV)	0.352	11.153	0.000	[0.288, 0.409]	Yes (***)
H1b	Expertise (PE) -> Trust in Product Recommendations (PT)	0.376	12.064	0.000	[0.312, 0.434]	Yes (***)
H2a	Demonstration Skills (DSs) -> Functional Value of Products (PV)	0.238	7.735	0.000	[0.176, 0.299]	Yes (***)
H2b	Demonstration Skills (DSs) -> Trust in Product Recommendations (PT)	0.314	7.686	0.000	[0.229, 0.390]	Yes (***)
H3a	Interactive Ability (IA) -> Functional Value of Products (PV)	0.357	8.646	0.000	[0.273, 0.435]	Yes (***)
H3b	Interactive Ability (IA) -> Trust in Product Recommendations (PT)	0.271	5.959	0.000	[0.180, 0.355]	Yes (***)
H4	Functional Value of Products (PV) -> Purchase Intention (PI)	0.232	3.598	0.000	[0.112, 0.362]	Yes (***)
H5	Trust in Product Recommendations (PT) -> Purchase Intention (PI)	0.225	3.571	0.000	[0.098, 0.342]	Yes (***)
H6a	Functional Value of Products (PV) × Physical Surroundings (PS) -> Purchase Intention (PI)	0.112	2.413	0.016	[0.001, 0.174]	Yes (*)
H6b	Trust in Product Recommendations (PT) × Physical Surroundings (PS) -> Purchase Intention (PI)	0.113	2.380	0.017	[0.032, 0.219]	Yes (*)
H7a	Functional Value of Products (PV) × Social Surroundings (SS) -> Purchase Intention (PI)	0.188	3.072	0.002	[0.049, 0.290]	Yes (**)
H7b	Trust in Product Recommendations (PT) × Social Surroundings (SS) -> Purchase Intention (PI)	0.191	2.779	0.005	[0.068, 0.341]	Yes (**)

Notes: *: p < 0.05; **: p < 0.01; ***: p < 0.001.

, the analysis results indicated that all hypotheses are supported. These results implied that streamers’ expertise, demonstration skills, and interactive ability positively affect consumers’ perceived functional value of products and their trust in streamers’ product recommendations. Furthermore, the results confirmed that both the consumers’ perceived functional value of the product and their trust in streamers’ product recommendations have a positive effect on consumers’ purchase intentions.

This study also analyzed the explanatory power of the research variables. As shown in Figure 2, the R² value for purchase intention was 0.710, which exceeds the threshold of 0.60. This indicates that consumers’ perceived functional value and trust in streamers’ product recommendations strongly explain their purchase intentions [95]. In addition, the R² values for perceived functional value and trust in product recommendations were 0.538 and 0.550, respectively, demonstrating a moderate level of explanatory power.

To further examine the model’s generalizability across diverse respondent groups, this study incorporated control variables including age, gender, and live streaming watch experience. The results showed that these variables had no significant impact on purchase intention (age: p = 0.036; gender: p = 0.524; live streaming watch experience: p = 0.653). These findings indicate that the model’s structural relationships were not sensitive to these sample characteristics, suggesting the model’s robustness.

In addition, to examine the incremental value of demonstration skills, we compared the predictive relevance obtained from the PLS predict procedure. When only expertise and interactive ability were included, the Q² values for perceived functional value, trust in product recommendations, and purchase intention were 0.486, 0.466, and 0.423, respectively. After adding demonstration skills, the Q² values increased to 0.529, 0.541, and 0.439. As all values were well above the threshold of 0.35, confirming stronger predictive relevance [103]. This suggests that demonstration skills provide incremental improvements, particularly in terms of perceived functional value and trust in product recommendations.

Furthermore, we conducted a multi-group analysis using the PLS-MGA approach to examine whether the structural relationships in our model differ across key respondent characteristics [95]. Specifically, we compared groups based on gender and purchase frequency. The results showed no significant differences across the PLS models for either gender or purchase frequency (all p > 0.05). This finding suggests that the proposed relationships are stable and generalizable across consumer subgroups, regardless of gender or purchasing behavior.

Regarding the tests of moderation effects, the results indicated that both the physical and social surroundings significantly moderated the relationship between consumers’ perceived functional value and their purchase intention, as well as the relationship between their trust in streamers’ product recommendations and consumers’ purchase intention. These moderating effects were further investigated through simple slope analyses at three levels (−1 SD, mean, and +1 SD), as suggested by Dawson [104] As shown in Figure 3a,b, higher levels of physical surroundings and social surroundings exhibit notably steeper interaction slopes than lower levels. This indicates that both moderators significantly moderate the relationship between the functional value of products and purchase intention. This implies that consumers who are more satisfied with the physical environment of the live streaming room and those who have friends who more actively share their live commerce shopping experiences or opinions are more likely to translate their perceptions of a product’s functional value into purchase intentions. Similarly, as shown in Figure 3c,d, the interaction slopes exhibit a sharper upward pattern at higher levels of physical and social surroundings compared to lower levels. This indicates that both moderators significantly moderate the relationship between trust in product recommendations and purchase intention. These results imply that consumers who are more satisfied with the physical environment of the live streaming room and those who have friends who more actively share their live commerce shopping experiences or opinions are more likely to translate their trust in streamers’ product recommendations into purchase intentions.

6. Discussion and Conclusions

This study aims to identify the cognitive mechanisms through which consumers’ perceptions of stimulus factors, such as streamers’ competencies, influence their purchase intention. It also seeks to explore strategies for enhancing the performance of live commerce businesses based on these findings. To achieve this, this study employed the stimulus–organism–response model as its theoretical foundation and examined whether streamers’ competencies (expertise, demonstration skills, and interactive ability) influence consumers’ internal states regarding products (functional value of products and trust in product recommendations) and how these internal states, in turn, affect consumers’ purchase intentions. Furthermore, to explore ways to strengthen the impact of consumers’ internal states on their purchase intentions, this study analyzed the moderating effects of situational factors (physical and social environments). All hypotheses were supported. The managerial implications derived from these results are discussed in detail in Section 6.1.

6.1. Managerial Implications

6.1.1. Strengthening Streamer Competencies to Enhance Consumers’ Perceived Functional Value of Products and Trust in Product Recommendations

First, this study confirmed that a streamer’s expertise has a significant positive effect on consumers’ perceptions of the product’s functional value and their trust in streamers’ product recommendations (H1a, H1b). This suggests that it is crucial for streamers to effectively strengthen their expertise in the product to increase consumers’ perception of product quality and strengthen trust in streamer recommendations. From this perspective, managers should make deliberate efforts to ensure that streamers can effectively and sufficiently acquire knowledge about the products they are scheduled to promote. To this end, managers should first review session operation practices that enable streamers to understand the product and marketing points in advance and then identify areas for improvement. At this point, it is crucial to actively involve product experts from manufacturing companies to provide accurate and in-depth knowledge during the session.

Moreover, product experts should support streamers beyond preparatory sessions. It is equally essential to establish a reliable and easily accessible communication channel, such as a hotline or a dedicated messaging platform, allowing streamers to consult experts whenever necessary. This ensures that streamers can freely ask questions and clarify uncertainties throughout their preparation process, thereby deepening their understanding and confidence in presenting the product. In particular, if streamers can communicate quickly with product experts even during the live session, this could be highly helpful in responding effectively to unexpected questions from the audience.

In addition, managers should consider providing a digital learning environment that enables streamers to study the product independently and at their own pace. This can be achieved by developing high-quality learning materials that include detailed information about product specifications, key differentiating features, usage instructions, and core marketing points. These materials should be presented in various formats, such as text, images, and videos, to support different learning preferences. Hosting such resources on a cloud-based platform would allow streamers to access the information conveniently, anytime, and anywhere. Moreover, during the learning process, streamers should be encouraged to actively use the established communication channels with product experts to clarify any uncertainties they may have. This integrated approach would help them build confidence and deepen their understanding of the product.

After the self-learning phase, it is also essential to organize briefing or rehearsal sessions where streamers present their knowledge of the product to both the product manufacturer’s experts and the live commerce planning team. Through interactive Q&A during these sessions, streamers’ knowledge can be evaluated, and any knowledge gaps can be identified and addressed. These feedback-oriented sessions provide a valuable opportunity to reinforce the streamers’ expertise before going live, ensuring more convincing and trustworthy live presentations.

Second, this study confirmed that a streamer’s demonstration skills have a significant positive effect on consumers’ perceptions of the product’s functional value and their trust in streamers’ product recommendations (H2a, H2b). This finding suggests that in order to enhance consumers’ perceptions of product quality and strengthen trust in streamers’ recommendations, streamers need to further develop their demonstration skills and explore ways to more effectively highlight the benefits of their demonstrations to consumers. To achieve this, managers should first invest significant effort in developing a variety of demonstration scenarios that can help maximize streamers’ demonstration effectiveness. Such scenarios should be provided as scripts emphasizing key product functions and benefits. In addition, streamers should then rehearse demonstrations using these scripts. To facilitate this, it is essential to design the tasks related to demonstration preparation and rehearsal as an efficient process supported by a structured management system. Within this framework, the roles of various participants who support the preparation and rehearsal stages should be clearly defined. The process should also incorporate procedures for continuous improvement of streamer competencies through broadcast monitoring and constructive feedback.

In addition, managers may consider investing in multi-camera equipment kits that allow a streamer to showcase products from multiple angles or on-screen display tools to highlight key product parameters (e.g., a product’s size, material, or battery life) or usage scenarios. To implement these visual aids effectively, managers should also provide streamers with technical training on how to utilize multi-camera systems and use display tools smoothly, thereby improving product understanding during live broadcasts. For example, a streamer may direct camera staff to capture close-ups of product specifications and switch views for usage demonstrations.

Moreover, managers should implement problem-solving strategies that help a streamer handle unexpected issues during live demonstrations, such as technical failures (e.g., camera or microphone malfunctions) or product demonstration failures. These problem-based training sessions should not be neglected to ensure smoother live broadcasts. Through training, managers can simulate real failures and equip streamers with practical responses, including switching to backup equipment, transparently addressing issues while maintaining trust, or offering alternative demonstrations.

Third, this study found that a streamer’s interactive ability has a significant positive impact on consumers’ perceptions of a product’s functional value and their trust in streamers’ product recommendations (H3a, H3b). This finding suggests that to enhance consumers’ perceptions of product quality and strengthen trust in streamers’ recommendations, it is essential for streamers to further develop their interaction skills and promote more engaging and responsive communication with viewers. From this perspective, managers should strive to enhance the platform environment where streamers interact with consumers, enabling them to capture consumer needs more efficiently and promptly. Although most live commerce platforms provide chat panels, the overwhelming volume of greetings, emojis, and repetitive or irrelevant messages often prevents streamers from quickly identifying purchase-related questions.

To address this issue, managers could assign support staff to monitor live chat in real time and highlight high-priority questions, especially those concerning product functions, usage, or features. At the same time, less critical comments, such as greetings or encouragement, could be managed by support staff using predefined responses or replying on behalf of the streamer. Such improvements in platform functionality and task delegation enable streamers to focus on product-related interactions, thereby shifting viewer engagement toward information-centric exchanges that support purchase decisions. To further enhance this effect, managers could explore AI-based tools that evaluate and prioritize viewer messages by their relevance to decision-making. Managers could also introduce a voting feature that allows consumers to indicate the product attributes of greatest interest. This would help streamers focus interactions on priority areas, fostering more targeted engagement and encouraging consumers to actively understand the product, thereby enhancing purchase intentions.

Managers should also provide streamers with opportunities for interaction training through simulations. Role-playing exercises, in which team members act as viewers, or online simulation programs can serve this purpose. These simulations should train streamers to respond quickly and accurately to product-related inquiries, manage high volumes of comments, and address negative or challenging messages—common scenarios in live commerce. Such training helps streamers develop skills in rapid decision-making, emotional regulation, and effective viewer engagement strategies, all of which are critical for successful interaction outcomes. For clarity, we summarized the link between competency-improving strategies and expected outcomes in

Table 8. Summary of competency-improving strategies and expected outcomes.

Competency	Managerial Strategies (For Manager)	Expected Outcomes (For Streamer)
Expertise (PE)	Review live session practices	Enhance preparation and product understanding
	Invite product experts to participate in live sessions	Obtain accurate and in-depth knowledge and strengthen confidence
	Establish communication channels for streamers to consult experts anytime	Clarify uncertainties and unexpected audience questions
	Provide a digital self-learning cloud platform	Learn independently at their own pace and enhance knowledge
	Organize briefing/rehearsal sessions to present knowledge to product experts and the planning team	Identify and fill knowledge gaps
Demonstration Skills (DSs)	Develop various demonstration scenarios and provide detailed scripts	Highlight key product functions and benefits in an effective way
	Provide opportunities to rehearse product demonstrations based on the scripts	Improve rehearsal quality and increase confidence
	Use multi-camera equipment kits and on-screen display tools; provide technical training	Improve product visualization demonstration skills
	Implement problem-solving and failure management training	Handle unexpected issues for live demonstration
Interactive Ability (IA)	Assign support staff to monitor live chat	Focus on product-related interactions
	Use AI-based tools to evaluate and prioritize audience messages	Reply to purchase decision messages efficiently
	Introduce a voting feature for consumers to select product features of interest	Focus on high-interest features and target engagement
	Provide interaction training through role-playing exercises or online simulation programs	Develop rapid decision-making, emotional regulation, and effective interactive strategies

.

6.1.2. Enhancing Consumers’ Perceived Functional Value of Products and Trust in Product Recommendations to Increase Consumers’ Purchase Intention

First, this study confirmed that consumers’ perception of a product’s functional value positively influences their purchase intention (H4). This finding suggests that in addition to enhancing the three key streamer competencies discussed earlier, live commerce platform managers should also explore additional strategies to strengthen consumers’ perception of functional value and promote purchasing behavior. One potential strategy is to invite testers with experience using the product to participate in live broadcasts. When these testers introduce the product’s benefits or share their opinions, it can help viewers better understand how the product meets their needs and ultimately increase the persuasive power of the broadcast.

In addition, platform managers could establish a functional value evaluation team to highlight product advantages. This team would consist of users who regularly assess new or promoted products using predefined scenarios. Their evaluations would focus on key functional attributes such as durability, usability, and real-world applicability. Standardized criteria, developed by the live commerce quality assurance team, should guide the evaluation process. The results could then be presented on a product quality panel to help consumers quickly identify a product’s functional strengths. By providing credible and detailed insights into product usefulness, this approach reinforces a function-oriented and quality-driven perception among potential buyers.

Second, the study confirmed that consumers’ trust in product recommendations positively influences their purchase intentions (H5). This finding suggests that in addition to enhancing the core competencies of streamers, live commerce platform managers should develop practical strategies to increase purchase intention by introducing greater transparency and accountability into the recommendation process. One option is to invite independent third-party verifiers to join live streams and provide factual explanations of key product features in real time. These verifiers may include technical experts from certified laboratories, consumer advocacy representatives, or recognized opinion leaders. Their role is not to promote products but to deliver objective assessments that support streamers’ claims. Before streaming, managers should prepare a verification checklist that covers product performance, compatibility, user experience, comparisons with competitors, laboratory test results, and warranty terms. During the session, consumers should also be given access to the verification report. This approach is expected to strengthen consumer trust by demonstrating accountability, while positioning the seller as a reliable and responsible brand.

Additionally, after each live streaming session, managers should conduct an internal review to assess the accuracy and clarity of all product recommendation processes. This review should track customer satisfaction, complaints, returns, and post-purchase evaluations to assess whether actual product usage aligns with the claims made during the stream. The findings should be used to improve scripts, product positioning, and FAQ content for future broadcasts. Additionally, disclosing key results of these reviews to consumers can highlight the platform’s responsibility and transparency, thereby strengthening consumer trust.

6.1.3. Optimizing Physical and Social Surroundings to Increase Consumers’ Purchase Intention

First, this study found that the physical environment can enhance the impact of consumers’ perceptions of a product’s functional value and their trust in streamers’ product recommendations on purchase intentions (H6a, H6b). These findings suggest that to increase consumers’ purchase intentions, managers should create a favorable physical environment tailored to the product’s characteristics, using lighting, background, product placement, and music to foster persuasive communication. Managers need to invest in environmental elements—such as background panels, lighting kits, movable shelves, and decorative props—that can be easily replaced or rearranged according to product type, brand campaigns, or seasonal themes. To ensure these investments are effective, it is essential to develop usage scenarios in advance. For example, different product categories—such as clothing, beauty, electronics, household goods, and sporting goods—are used in distinct contexts. Therefore, these scenarios should be designed to recreate studio settings that best showcase the product’s functions, features, or appeal. Such efforts are also important for managers who continuously stream within a single product category. Even in these cases, it is necessary to present the product in specialized contexts, such as seasonal product updates, new promotional campaigns, or comparisons between product lines (e.g., premium models vs. basic models). Strategically investing in scenario-based environmental elements is expected to enable streamers to more effectively increase consumer engagement and purchase intentions.

To build studio environments efficiently, managers should develop a digital library that shares details of previous studio setups. This library should include information on lighting, backgrounds, product placement, and music from past broadcasts, accompanied by photos or videos. It should be organized and searchable by product type and demonstration scenario. In addition, a studio resource database should store management information, such as the storage locations of props, and be linked to the digital library. This infrastructure facilitates the reuse of existing setup know-how, enhancing both the efficiency of transitioning to new broadcast environments and the overall effectiveness of broadcasts.

Second, this study confirmed that social surroundings can enhance the impact of consumers’ perceptions of the product’s functional value and their trust in streamers’ product recommendations on their purchase intention (H7a, H7b). These findings suggest that to increase consumers’ purchase intentions, managers should provide a social environment where participants can more effectively share their opinions and experiences during live broadcasts. This can be achieved through separate group spaces where viewers can communicate around specific products or themes related to the product category. For example, in a session featuring kitchen appliances, discussion groups around air fryers, coffee machines, or dishwashers can boost consumer engagement and influence purchase intentions. Group themes can be predefined by managers based on anticipated audience interest or content presented by audiences during sessions. For single-product broadcasts, groups might focus on specific use cases, such as baking bread in an oven. Features that forward group questions directly to the streamer can further enhance interaction. To encourage participation, managers could offer incentives, such as discount coupons awarded through raffles, for active group engagement.

In addition, managers could provide an environment for consumers who wish to make purchases after the live broadcast, allowing them to refer to other consumers’ opinions. Many viewers do not purchase during the stream but take time to evaluate products afterward. To support these consumers, managers can implement voting mechanisms on product preferences, such as colors or features, and share the results with undecided viewers. This approach mimics everyday social shopping behavior, where consumers seek input from peers before purchasing. Such post-stream interactions are expected to increase the likelihood that hesitant consumers will make a final purchase.

6.1.4. Extending Managerial Implications to Western Live-Commerce Platforms

While the preceding implications are based on the Chinese market, it is equally important to consider their application to Western live-commerce platforms. Implementing these implications on platforms such as TikTok Shop and Instagram Live requires adapting to an entertainment-focused user experience (UX) and stricter regulations. Streams must capture viewers’ attention within seconds, so demonstrations should be concise, and interactive elements should be rapid. Checkout flows are often less integrated, making clear calls-to-action essential. Regulations, including the Federal Trade Commission (FTC) guidelines in the U.S. and data privacy rules such as the General Data Protection Regulation (GDPR) in the EU and the California Consumer Privacy Act (CCPA) in the U.S., require transparency in sponsored content and careful handling of user data, making compliance a key part of streamers’ role.

Given these platform constraints, e-commerce companies must strategically allocate streamer competencies—expertise, demonstration skills, and interactivity—based on physical and social surroundings. For complex products, the physical setup should be minimalist to support clear demonstrations, with bright lighting and a fixed camera to help viewers focus on details. For simpler products, the physical context can enhance atmosphere through dynamic camera angles, styled backgrounds, and curated music. This allows the streamer to emphasize storytelling and interactivity, making viewers feel part of an exclusive experience. Social surroundings are equally critical. When social buzz is low, the streamer must lead by using expertise and demonstration skills to establish credibility and encourage engagement. When social buzz is high, interactivity takes priority. The streamer should manage conversation, answer questions in real time, and channel audience energy effectively. By dynamically adjusting streamer competencies to both physical and social contexts, e-commerce companies can maximize viewer engagement and drive higher sales performance.

6.2. Theoretical Implications

First, this study extends the existing research perspective by introducing a new variable: demonstration skills, which are a key competency of streamers. Previous studies have shown that streamers’ competencies, such as expertise and interactive ability, affect purchase intentions [8], but they have largely overlooked the ability to present products in a raw and factual manner during live broadcasts to meet consumers’ desire for product authenticity. This study addresses this research gap by statistically demonstrating that such demonstration skills play a significant role in consumers’ decision-making processes, comparable to expertise and interactive ability. Accordingly, this study is significant in that it adds demonstration skills as a meaningful component of streamer competency and develops measurement items for it, thereby establishing a theoretical foundation for the practical application of this concept in future research.

Second, unlike previous related research, this study statistically demonstrates that the key environmental factors in live commerce, specifically the physical and social surroundings, serve as moderators that enhance consumers’ purchase intentions. This finding is expected to serve as a theoretical basis for developing strategies that enhance consumer engagement and stimulate purchasing behavior in live commerce. In addition, the proposed research model, which integrates these environmental factors, provides a more comprehensive explanation of the psychological mechanisms shaping consumers’ purchase decisions. Moreover, the development and statistical validation of measurement items for these environmental constructs are expected to serve as a useful theoretical foundation and tool for future research in this area.

6.3. Limitations and Future Research Directions

First, the analyzed data were collected solely from the Chinese market, with a sample skewed toward female and student respondents. Due to the scope and characteristics of these data, the generalizability of the findings may be limited. Moreover, the psychological mechanisms influencing consumers’ purchase decisions in live commerce environments are likely to be shaped by cultural factors. For instance, cultural norms may determine the extent to which individuals are conscious of others’ opinions or prefer certain communication and expression styles. Therefore, future research should expand the sampling framework to reflect not only the demographic characteristics of the market but also the features of more diverse cultural contexts. On this basis, the proposed model should be validated from a comparative perspective across demographic and cultural settings.

Second, this study collected data through a self-reported questionnaire, which may be subject to potential biases from social desirability, recall errors, and subjective interpretation. Future research is encouraged to validate this model by collecting behavioral and transactional data in live sessions, such as click-through rates, add-to-cart actions, or actual sales data. This would help determine whether the results are consistent in actual live commerce settings, beyond the limitations of questionnaire-based evidence. Such an approach can also improve external validity and provide actionable insights for platform managers seeking to optimize sales performance.

Third, consistent with Mehrabian and Russell [14], this study used the S-O-R model to assume a unidirectional relationship based on the S-O-R process. While this framework is widely adopted, it does not capture potential feedback loops or bidirectional relationships in a highly interactive live commerce environment. For instance, consumers’ immediate feedback may shape the way streamers respond, such as by changing their explanations, emphasis, or pace of product demonstration. Future research could explore such feedback loops or bidirectional dynamics to provide a more comprehensive understanding of consumer behavior in interactive settings. In addition, the proposed model assumed positive and significant linear associations across all variables, which may have led to confirmation bias. Future research should explicitly theorize and test for non-significant, negative, and non-linear relationships (e.g., inverted U-shaped effects) to uncover the potential boundary conditions and enhance theoretical tension.

Fourth, this study did not consider the potential influence of individual-level factors such as consumers’ cultural differences, product knowledge, various prior experiences, and involvement level. Future studies are recommended to incorporate these variables as moderators to the S-O-R process, thereby strengthening the theoretical framework and expanding the practical implications.

Fifth, this study focused on streamers’ competencies and physical and social surrounding factors but did not sufficiently account for diverse contextual factors that may appear as potential distracting variables. Specifically, the relationship between streamer competencies and purchase intention may not hold in all contexts—for instance, it could be weaker for low-involvement products or passive audiences. Future research should consider incorporating contextual factors such as product type, stream duration, promotional intensity, streamer followers, and involvement level as control variables, moderators, or mediators. This approach will allow readers to understand how different contextual factors influence consumers’ purchase behavior.