The Impact of E-Commerce Live Streaming on Purchase Intention for Sustainable Green Agricultural Products: A Study in the Context of Agricultural Tourism Integration

Abstract

Growing awareness of sustainable development and green consumer concerns is driving the market expansion for green agriculture products. E-commerce live streaming gives rural enterprises a new channel through scenario-building and interaction, while agro-tourism integration combines resources to generate a variety of promotion scenarios. This study examines the effects of external stimuli, including social networks, resource endowment, infrastructure, and the characteristics of e-commerce streamers, on the perception, trust, perceived value, and purchase intention of green consumption. It is based on the SOR (Stimulus–Organism–Response) theoretical model and focuses on e-commerce live streaming in the agriculture-tourism integration scenario. According to a structural equation modeling (SEM) analysis of 350 consumer questionnaires, these external stimuli primarily influence purchase intention through perceived value, trust, and green consumption cognition, with resource endowment having the most significant impact. The effects of infrastructure on perceived value and streamer attractiveness on green consumption cognition are not statistically significant. This research not only broadens the use of the SOR model in the emerging field of agritourism integration but also offers rural businesses theoretical backing and useful guidance to maximize e-commerce live marketing and enhance agritourism integration.

1. Introduction

As a strategic decision throughout human history, sustainable development has created a methodical development paradigm and a workable path. It is also a necessity and an unavoidable decision for the advancement of humanity in the future. There is agreement in the global governance agenda regarding the strategic importance of food safety as a vital component of ensuring human populations’ existence. A strong basis for advancing sustainable development and the synchronization of human society, economics, and nature is established by the consistent development of green agricultural goods [1]. The “digital business to promote agriculture” concept was proposed in 2022 Central Document No. 1. When paired with the expansion of the e-commerce sector into rural regions, agricultural live streaming of products has emerged as a significant means of advancing the sustainable development of rural areas [2].

Green agricultural e-commerce is transformed by e-commerce live streaming, which also speeds up the digital transformation of agricultural marketing and helps the industry move into a new phase of development. Based on data issued by the China Internet Network Information Center (CNNIC) in 2022 [3,4], the total number of people who noticed live stream advertisements surpassed 703 million, accounting for 68.2% of all Internet users. Among them, the scale of e-commerce livestreaming users was 464 million, a year-on-year increase of 75.79 million or 19.5%. [5]. Live e-commerce is a new interactive direct marketing platform that SMEs are aggressively embracing globally to market a variety of items [6].

Environmentally friendly products, such as green certified agricultural products of the European Union EC 834/2007 standard [7], are collectively referred to as “green agricultural products” in the field of agricultural economics. The National Development and Reform Commission released the “Implementation Plan for Promoting Green Consumption” in 2022. It specifically calls for raising the market share of low-carbon and green products and green food to 22% by 2025 [8]. High-quality green agricultural product development is strategically important for increasing consumer purchasing intent, boosting green agricultural products’ competitiveness in the market, and meeting consumers’ dual demands for ecological responsibility and food safety [9,10].

Nevertheless, the current research still has certain flaws. First, the majority of research focuses on either a single agritourism scenario or traditional e-commerce live stream, leaving out a systematic analysis of e-commerce live stream in the context of agritourism integration and cross-composite scenarios of the two. Second, the mechanism by which external stimuli (such as social networks, resource endowment, infrastructure, etc.) influence purchase intention through green consumption cognition, trust, and perceived value is not thoroughly examined, particularly when considering the synergistic effect of several variables in particular scenes.

This work makes several contributions, aiming to close the aforementioned research gaps. It creatively applies to the developing composite scenario of e-commerce live streaming in the context of agro-tourism integration, and it is based on SOR theory. Combining a variety of outside stimuli and mediating factors that influence the intention to buy green agricultural products. In addition, it gives rural businesses and e-commerce platforms the scientific foundation they need to maximize the live streaming of green agricultural products and promote rural revival and the idea of sustainable, green consumption.

Building and validating the SOR theoretical model of how e-commerce live streaming affects consumers’ propensity to buy green agricultural products in the context of agricultural tourism integration is the aim of this work. The main external stimuli (social network, infrastructure, resource endowment, e-commerce streamer characteristics) and the internal cognitive state of consumers (green consumption cognition, trust, perceived value) are systematically examined. The precise mechanism of these intermediary variables on the willingness to buy is revealed, along with theoretical guidance for related practices.

2. Literature Review and Hypothesis Development

2.1. Agritourism Integration

The purpose of integrating agriculture and tourism is to generate cross-industry value by rearranging components. This synergistic process is fueled by the matching of market supply and demand in addition to the penetration of digital technologies [11]. In both developed nations, like the United States, Germany, and Japan, and developing nations, like Malaysia and Indonesia, the model—which had its roots in the European practice of rural tourism—has emerged as the predominant form of “experiential agriculture” in East Asia [12]. It has also been shown to be a successful means of raising farmers’ incomes and agricultural productivity [13]. Reconfiguring the rural value system, encouraging industrial economic diversification, preserving farming culture, and offering an immersive platform for the promotion of green agricultural products are some of its benefits. These aim to achieve the synergistic development of ecological preservation, cultural heritage, and economic efficiency.

This study’s primary composite scenario background is provided by agritourism integration. Through a unique chance for customers to interact with the source of green agricultural production, in-depth contact, and experience, it surpasses the restrictions of traditional e-commerce or single tourism research, allowing farmers to receive sustainable green premium returns [11]. From the perspective of psychology, humans need to be close to nature, and nature also provides cultural and social values to humans [14]. By giving visitors an instinctive sense of the standards and quality of green agriculture, their farming experiences increase trust. Travel-related sensory memories also last after consumption. Research indicates that roughly 63% of visitors will make additional online purchases following their experience [15]. By highlighting the need to look at agricultural e-commerce live streaming within the framework of agritourism integration, this study fills a research gap in this intersecting topic.

2.2. E-Commerce Live Streaming

E-commerce live streaming is a new business model that uses network media live streaming technology to create a three-dimensional and multi-dimensional commodity display and information interaction system that gives customers useful shopping advice. The Internet and retail sectors in China are witnessing the emergence of a new mode of consumption [16]. In 2025, the market will be worth over CNY 26 billion, with over 515 million users of live stream e-commerce in China by the end of 2022. The GMV of agricultural products live stream increased 127% year over year, demonstrating the market’s suitability [17]. In contrast to traditional e-commerce, e-commerce live streaming is the mainstay. It breaks through the limitations of traditional e-commerce planarization displays with real-time images and instant Q&A, allowing consumers to become aware of the efficiency of 58%–63%, reshaping the logic of the “people-goods-field” interaction, and realizing the integration of shopping and experience. In general, e-commerce live streams offer a more realistic and vivid online shopping experience, allowing customers to gain a better understanding of the items and aiding in their decision to purchase green agricultural products [18].

This paper effectively communicates the “sense of reality” and “authenticity” of the agritourism experience to online consumers through real-time interaction and scenario-based display in the context of agritourism integration, whereas prior research has primarily examined stream characteristics or platform technology alone. In order to bridge the gap in scene-based cross-study, this paper offers technical assistance and an interactive basis. It also enables the investigation of how external cues, such as stream features, impact behavior through internal cognition.

2.3. Theoretical Framework: Stimulus–Organism–Response (SOR) Model

The Stimulus–Organism–Response (SOR) model is a popular instrument for examining how external factors affect consumer behavior [19]. It examines the impact of environmental cues on human behavior. Donovan and Rossiter subsequently used the SOR model to study consumer purchasing patterns in the retail industry [20]. Chen also looked into how customers’ impulsive purchasing habits and standards of evaluation are affected by the environment of websites and promotional activities [21]. Building on these foundations, this study provides an extensive and methodical theoretical framework by analyzing the relationship between e-commerce live-streaming and the purchasing of green agricultural products using the SOR model.

In the context of green agriculture e-commerce livestreams using the SOR framework, presenting origin resources, host attributes, and real-time interaction mechanisms are stimuli (S). Organism (O): By learning about the production processes, consumers formulate concrete opinions about quality. Purchase behavior, turning dormant demand into actual consumption, is the response (R). Through experiential engagement, this mechanism propels the sales of green agricultural products. By demonstrating the beneficial moderating effect of host credibility on the stimulus-cognition pathway, creating a social did integration effect, and developing a useful theoretical framework for cognitive mediators like value perception, trust, and green consumption cognition, this study creatively adapts the model [22].

2.4. External Stimuli Influencing Factors in E-Commerce Livestreaming

Social networks and resource endowment encourage local product sales, multi-stakeholder synergy, sustainable green growth in agriculture, and agricultural tourism integration. This study reveals the mutually beneficial relationship between green agricultural sales and rural revitalization by using a structural equation model based on the SOR to analyze how social networks, infrastructure, and resource endowment affect consumer purchase intention in agritourism-integrated e-commerce livestreams.

2.4.1. Social Networks

British anthropologist A.R. Radcliffe-Brown coined the term “social network,” which has as its main goals fostering user interaction and interpersonal relationships, as well as analyzing the flow of trending information in the social relationship network. Nowadays, social media platforms like Facebook and Jitterbug serve as digital social capital carriers, and farmers communicate with one another in rural society, creating a complex structure that is indicative of small-world network characteristics [23].

Customers can learn more about green agricultural products through social networks, which increases their degree of trust. The quality of the product relationship and the network location advantage both have an impact on the decision to buy. According to the social proximity hypothesis, live streaming’s interactive, real-time environment (such as viewership, pop-up screen activity, and incentive structure) can have a big impact on how customers perceive and behave [24,25]. Internet word-of-mouth is a significant determinant of consumer purchasing behavior, as Dellarocas empirically demonstrated through the moderating effect model. The threshold of the percentage of negative comments varies significantly, and when the threshold is crossed, the degree of consumer trust becomes inversely proportional [26]. Social media is a crucial tool for increasing the agritourism experience’s green value and word-of-mouth in the context of live-streaming agritourism integration. Social networks and consumers’ perceived group identities can both improve green consumption cognition and product value while successfully lowering consumers’ perceived risk from information asymmetry and green features. But the unique impact on green consumption cognition and trust in the live streaming situation of agritourism integration has been largely overlooked in the majority of prior studies. Thus, the following theories are considered in this paper:

H1a: 

Social networks positively influence green consumption cognition.

H1b: 

Social networks positively influence perceived value.

H1c: 

Social networks positively influence trust.

2.4.2. Resource Endowments

The resource endowment theory, first put forth by Swedish economist Ohlin in 1933, includes components including technological penetration rate, ecological capital, and human skill endowment [3]. Farmland is the most important resource in agricultural communities, and ecological capital provides the material basis for economic activity [27]. Variations in resource endowment differentiate economic actors and promote agricultural economic development and unique geographical characteristics. Planting structures are optimized, buying intentions are stimulated, and expanding agricultural demand is met by long-term e-commerce livestreaming market feedback systems [8]. Resource-rich areas have more potential for green economic growth when they are in line with sustainable development and rural regeneration. A tripartite digital mechanism known as “production traceability, scene experience, instant transaction” is established via live e-commerce, which increases the value of regional agriculture, improves transformation efficiency, and promotes the expansion of tourism.

Transparency of product information and ecological value are the two main components of the perceived value of green agricultural products. The government-built certification system has significantly raised consumer confidence in product quality, which is now a key component in encouraging green consumption and rural revitalization [28]. Although agricultural productivity and resource endowment are related, little is known about how these factors affect purchase intention through perceived value and trust in live agritourism. Resource endowment is the cornerstone of agritourism integration and a crucial distinction for eco-friendly products, influencing consumer perceptions. Improved perceived ecological worth and trust are increased when e-commerce streams prominently display superior origin ecology and cutting-edge production technology. By facilitating natural customer involvement in manufacturing processes, the “direct picking + cloud supervision” approach immediately enhances perceptions of origin and production, trust, and readiness to pay. Based on the above theoretical analysis, this study proposes the following hypotheses:

H2a: 

Resource endowment positively influences green consumption cognition.

H2b: 

Resource endowment positively influences perceived value.

H2c: 

Resource endowment positively influences trust.

2.4.3. Infrastructure

With the increase in environmental awareness, people’s infrastructure is receiving increasing attention. Physical constructions that support public services for social production and citizens’ livelihoods are referred to as infrastructure [20,29]. By using the producing infrastructure, the transportation system in the agricultural tourism integration model of development makes sensible use of land resources. Convenient transportation makes it possible for consumers to engage in sustainable tourism and green agriculture in depth, which improves their comprehension of green consumption cognitions and the production processes, environmental characteristics, and health benefits of green agricultural products. This, in turn, raises the perceived value of the products. Green consumption cognition, as defined by Lien Ching-Yu et al. [30], is the perception and comprehension of environmentally friendly concepts by customers, as evidenced by the environment, facilities, and services. It is hypothesized that consumers’ green consumption cognition influences their behavioral intentions [4].

The realization of the promise of green agri-food and a smooth agritourism integration experience depends heavily on infrastructure. Customers are very conscious of and worried about the logistical timeline, freshness conditions, and quality assurance while buying fresh green agricultural items online. A flawless and dependable infrastructure (such as “official certification and traceability” and “24-h cold chain direct delivery”) can lower consumers’ perceived risk, increase their confidence in transaction security and product quality, and raise their value of convenience and quality control. Perception: By confirming that the consumer experience and quality assurance impact trust in the live agricultural tourism scenario, this paper differs from earlier studies on the effectiveness of infrastructure logistics. It demonstrates that consistently offering products with high perceived value and low perceived risk can improve consumers’ green consumption cognition and brand trust [31]. Accordingly, the following hypotheses are proposed:

H3a: 

Infrastructure positively influences green consumption cognition.

H3b: 

Infrastructure positively influences perceived value-

H3c: 

Infrastructure positively influences trust.

2.5. Influencing Factors of E-Commerce Live Streamer Characteristics

A thorough analysis of the literature on livestreaming for agricultural product e-commerce shows that its primary components are platform technology (context), product qualities (goods), and live streamer characteristics (person). These factors work together to affect consumers’ purchasing intentions through a multifaceted mechanism, showing special significance in the areas of green agricultural products and the integration of agriculture and tourism.

When relevant streamers, farmers, and others use online platforms to communicate with viewers in real time, they are engaging in agricultural product e-commerce live-streaming, which aims to promote and sell agricultural products more professionally [32]. In addition to selling and promoting goods, farmers and other professionals who work as streamers also gain the trust of their audience. While live-streaming purchasing incorporates multi-dimensional interactivity and real-time video display, traditional e-commerce shopping primarily depends on product photos. E-commerce live-streaming optimizes the shopping experience for customers and re-constructs their buying intentions by reshaping the model of interaction between consumers, items, and scenarios [33,34]. Through the development of rural tourism and the selling of green agricultural products, it exhibits both substantial commercial value and social benefits. According to a thorough analysis of pertinent research, the main traits of e-commerce streamers are professionalism, interaction, and beauty, all of which have a big impact on customers’ inclinations to buy [35].

2.5.1. Professionalism

Streamers’ professionalism is defined as their ability to effectively communicate relevant information, experience, and abilities to live viewers [36,37]. Customers are more likely to trust the information source when they perceive this inherent potential [38]. Streamers utilize live-streaming as part of the agritourism integration concept to inform viewers on the qualities and cultivation practices of the product, encourage travel and dining experiences that are unique to the area, and offer technical and informational assistance. By raising knowledge of the allure of rural travel and strengthening the concept of green consumption, this tactic increases consumer trust. Highly skilled streamers can boost consumers’ purchase intentions by 28.4%, according to research, with perceived trust mediating 68.3% of this effect. This increases consumer trust and improves product transparency [39].

Researchers have observed that consumers’ propensity to buy green agricultural products is influenced by their level of green consumption cognition, which is positively connected with their awareness of the green worth of goods and services [26,40]. The knowledgeable descriptions of agricultural techniques, such as Xinjiang apple dwarf dense planting, and quality criteria provided by streamers improve viewers’ awareness of environmental issues. Local customs, landscapes, and cultural significance are highlighted in drone footage during Xinjiang Apple livestreams, which enhances regional identity and gives items emotional meaning. Consumers’ Green Payment Premium (GPP) can increase by one to three times the levels of conventional products when the Cultural Emotion Empowerment Intensity (CEI) surpasses 0.68, according to empirical data. Customers rely heavily on expert information to evaluate the quality and authenticity of products, and the stream’s experience serves as a crucial link in interpreting agritourism resource endowment and green production processes. This article explores the impact and strengthens the rationality of the hypotheses through the elaboration and cultural empowerment of the agricultural product story by professional streams, in contrast to earlier single studies on stream expertise. Based on the above theoretical analysis, the following hypotheses are proposed:

H4a: 

The professionalism of e-commerce streamers positively influences green consumption cognition.

H4b: 

The professionalism of e-commerce streamers positively influences perceived value.

H4c: 

The professionalism of e-commerce streamers positively influences trust.

2.5.2. Interactivity

In general, interactivity, also known as real-time interaction, indicates the dynamics of relationships between people. Interactivity in e-commerce livestreaming refers to the dynamic network created by the host and viewers exchanging information in both directions at the same time, its core includes the interaction between users and products as well as the contextualized introduction mode of brand promotion [41,42,43], the frequency of efficient host responses to questions, and the contextualized presentation of product information and brand promotion. These actions help customers become more cognizant of green consumption. Real-time engagement lowers the psychological gap between hosts and customers; research shows that when hosts answer queries about the product in less than three seconds, consumers stay on the site for 18.6 min, which increases the impression of the host’s excitement and presence [44]. Consequently, this fosters confidence in the genuineness of the agricultural goods the host displays, impacting consumer choices.

Consumers’ feelings of enjoyment are positively impacted by interactions with hosts. The consumer’s sense of emotional fulfillment increases by 57%, and their cognitive pleasure index increases by 0.38 standard deviations when the frequency of interactions reaches or surpasses three times per minute. By satisfying customers’ demands, cognitive trust is created, which encourages proactive use of streaming products and strengthens perceptions of value and green consumption cognition [45]. Moreover, presenters improve the shopping experience by presenting panoramic views of green tourist amenities and ecological agritourism landscapes while engaging with viewers in real time. This “emotional resonance + scene experience” communication, which is distinct from the prior communication and engagement, improves the perceived value of the product in the live process by significantly reducing the information asymmetry caused by the perceived risk. Studies show that when there is a lot of contact, consumers spend a lot more time looking at green items. Based on the above theoretical analysis, the following hypotheses are proposed:

H5a: 

E-commerce streamers’ interactivity has a positive effect on green consumption cognition.

H5b: 

E-commerce streamers’ interactivity has a positive effect on perceived value-

H5c: 

E-commerce streamers’ interactivity has a positive effect on trust.

2.5.3. Attractiveness

Consumers can browse and watch streams conveniently at any time and from any location using e-commerce livestreaming, a growing commercial paradigm [46,47]. Positive consumer perceptions are largely influenced by the host’s and product’s attractiveness, which are the two main components of livestreaming. Customers can choose to continue watching a particular livestream or move on to another one based on its attractiveness [2,48]. Research indicates that consumers have a significant trust bias for information presented by visually appealing hosts, viewing such material as more convincing and reliable. The value positioning of green agricultural products is enhanced by hosts who highlight the benefits and features of these items through immersive presentations and live demonstrations. This deepens consumers’ interest and comprehension of the products. Positive emotional appeals, in particular, increase consumer satisfaction, bolstering emotional brand recognition and acceptance of advertising content, and raising consumers’ perceptions of green value [49].

According to social psychology studies, interpersonal attraction is determined by personality factors, including honesty, extroversion, and looks. These traits can boost customer satisfaction and loyalty to green agricultural products. Using consistent and regular interactive actions and positive feedback systems, hosts gradually strengthen customer trust, creating a strong bond of trust between the two sides. Research shows that through the emotional trust pathway, beauty positively affects purchase intention [50]. The invaluable, crucial significance of different host features in grabbing customers’ attention and influencing their purchasing decisions is confirmed by trial data. While its direct impact on green consumption cognition may be less than that of professionalism, it is essential in generating initial attention and favorable attitudes. Its path of influence centers on establishing initial trust through emotional resonance and consumer likability, as well as influencing purchase intentions by enhancing the emotional value of the viewing experience. Based on this theoretical analysis, the following hypotheses are proposed:

H6a: 

E-commerce streamer attractiveness positively influences green consumption cognition.

H6b: 

E-commerce streamer attractiveness positively affects perceived value.

H6c: 

E-commerce streamer attractiveness positively impacts.

2.6. Intra-Consumer Cognitive Mediation

2.6.1. Green Consumption Cognition

Based on the integrated framework of the Theory of Planned Behavior (TPB) and Value-Belief-Norm (VBN) theory, with a focus on green sustainability principles, key factors influencing consumers’ purchase of green agricultural products include psychological, economic, and sustainable environmental factors. Among these, Green Consumption Cognition—as a critical individual-level psychological factor—plays a multifaceted and complex role [51]. Amidst global ecological challenges and heightened sustainable environmental awareness, green consumption demand has emerged. By deepening their green consumption cognition, consumers can more accurately comprehend products’ environmental and health values, thereby enhancing perceived value and purchase intention.

Notably, Green Consumption Cognition directly contributes to the formation of Green Trust, characterized by confidence in organic certifications, low-carbon production, supply chain sustainability, and corporate environmental responsibility. This fundamentally differs from general online trust (focused on merchant reliability/logistics efficiency): the core of Green Trust lies in the conviction regarding the authenticity of ‘green’ claims [52]. Research indicates that green agricultural product purchasing decisions are predominantly made by elderly consumers or homemakers, whose primary information sources are television and online channels [53]. Given the inherent information asymmetry concerning agricultural products’ green attributes, consumer trust may be compromised, consequently diminishing purchase motivation. Aligned with the Theory of Planned Behavior (TPB), detailed presentations of green attributes through e-commerce livestreaming can positively enhance purchase intention [51,54]. Within the agritourism livestreaming context examined in this study, the aforementioned stimulus factors—through comprehensive product information, production process demonstrations, and value proposition communication—effectively shape and elevate consumers’ Green Consumption Cognition. The following theory is considered in light of the theoretical examination above:

H7: 

Green consumption cognition has a significant positive impact on consumers’ willingness to purchase green agricultural products.

2.6.2. Perceived Value

Customers’ subjective assessment of a product or service’s total usefulness, which is determined by balancing the costs and advantages, is known as perceived value. Value is acknowledged as a key behavioral predictor that influences consumer decisions to acquire company goods [55]. Within the whole consumer value perception system, perceived value is a one-dimensional but comprehensive structure [56]. Changes in the macroenvironment and increasing household incomes have enhanced the quality of life for consumers. As a result, consumers view green agricultural products as safer and healthier compared to conventional ones. Their readiness to pay a higher price arises from their perception of green value and awareness of eco-attributes. Due to increased awareness of health and environmental issues, green products are favored for their health and environmental advantages, which impact purchase intention [57].

Green consumption habits and perceived value efficacy frequently evoke pleasant sentiments on a human level, such as addressing environmental contribution and health problems. As a result, within the emotional value component, purchase intention is highly correlated with perceived effectiveness in environmental contribution. These elements—social, emotional, and functional value pathways—have a big impact on consumers’ choices when it comes to buying green agricultural products. This paper fully examines the motivating factors in the live streaming of agricultural tourism in the survey research, in contrast to earlier studies that concentrated on a single attribute in the analysis of perceived value: the host’s professional explanation (functional value), the enjoyment of the interaction (emotional value), and the ecological superiority brought about by the demonstration of the resource endowment (social/emotional value), which together shape the positive impact of the consumer’s perceived value of green agricultural products. The following hypothesis is considered based on the theoretical study above:

H8: 

Perceived value has a significant positive impact on consumers’ willingness to purchase green agricultural products.

2.6.3. Trust

As the basic mechanism in social exchange interactions, trust is a complex psychological phenomenon that can be defined as the expectation placed on individuals or organizations in uncertain situations. Consumer trust, according to institutional trust theory [58], is made up of competence trust, benevolence trust, and institutional assurance trust. It reflects the expectations of consumers that sellers would keep their commitments and behave appropriately. Establishing and upholding consumer product trust is essential to promoting customer involvement and purchasing patterns [59]. Consumer expectations and perceptions based on the inherent characteristics of green agricultural products, the competence and dependability of the producers, and the consequent desire to trust the business and its goods are referred to as “green trust”. Businesses with an excellent track record and image are more likely to succeed in winning over customers to their green agriculture products.

The COVID-19-related digital trust mechanism suggests that the rise in online commerce presents difficulties for customers worldwide. Consumer attitudes change from neutrality or hesitancy to active support when they have faith in green agricultural products. Attitude acts as a mediator between purchase intention and trust. With the ease with which green agricultural products may be chosen owing to e-commerce livestream platforms, boosting consumer trust and promoting online transactions continue to be top academic and industrial concerns. The distinctiveness of trust is elucidated by this study, which highlights the ecological qualities of green agricultural products in addition to transactional dependability. Customers’ green trust and green perception of the production process are built and strengthened in multiple dimensions through the integration of live streaming into agritourism. This is because it stimulates the display of the actual environment of origin, production process, and reliable infrastructure guarantee, a feature that makes its influence on the intention to purchase green agricultural products more crucial. Based on the above analysis, this study proposes the following hypothesis:

H9: 

Trust has a significant positive impact on consumers’ willingness to purchase green agricultural products.

2.7. Conceptual Model

Based on the Stimulus–Organism–Response (SOR) theoretical framework, this study constructs a transmission mechanism model of green consumption behavior in the live-streaming e-commerce context. As shown in Figure 1. Path coefficients are employed when combined with a second-order confirmatory factor analysis (CFA) framework constructed with Smart PLS 3 to reveal the sequence of “digital stimulus → cognitive processing → behavioral intention”.

• Stimulus layer (S): external stimulus variables include social networks, resource endowments, infrastructure, and e-commerce streamer characteristics (professionalism, interactivity, attractiveness);
• Organism layer (O): mediating variables include perceived value, green consumption cognition, and trust;
• Response layer (R): consumer willingness to purchase green agricultural products is selected as the response variable in the research model.

3. Research Methodology

This study constructs a mechanism model that demonstrates how e-commerce live streaming elements influence the willingness of customers to buy green agricultural products. It is based on a number of theoretical viewpoints. A structural equation model is created to investigate the connections between latent variables (green consumption cognition, trust, perceived value, and other purchase intention) and independent variables (social networks, infrastructure, resource endowments, and e-commerce streamer features). The Bootstrap technique is used for confirming the importance of mediating effects in data analysis, which is performed using IBM SPSS 27 and Smart PLS 3.2.9.

3.1. Quantitative Research

3.1.1. Experimental Materials and Stimulus Design

The experimental materials were selected from widely recognized Douyin live streaming e-commerce content. Following screening, editing, and consulting with industry research studies and the “White Paper on the Development of Agricultural Product E-commerce Live Streaming in China,” the final experimental material is referred to as the “green agricultural product e-commerce live streaming scenario”. Evaluation and analysis of the live streaming visual presentation’s professionalism, communication, and consumer appeal are all contained within the research material.

High-frame-rate streaming technology and multimodal capture systems, including intelligent HD cameras, are employed by short-video live streaming sites like Douyin and Kuaishou to precisely capture streamers’ body language and facial expressions. Streamers employ advanced image design that is compatible with popular aesthetic tastes, and their language and clothes are adapted to the traits of agricultural product categories. Relevant research shows that moderately lengthy live streaming rooms with modular content layout techniques boost the average watch time by 37.6% and increase the consumption conversion rate by 1.8 times the industry average.

These experimental materials were standardized employing expert editing instruments after being taken from four well-known live streaming components on the Douyin network. To ensure optimal video quality and viewing experience, video source files were captured using professional-grade equipment at 1920 × 1080 resolution and 30 frames per second. They were then edited into standardized experimental samples with a duration of 600 ± 30 s. The experimental movies show streamers describing products, answering questions from prospective customers in real time, and assisting viewers with the purchase process. This video content offers strong support for further experimental analysis and study by thoroughly showcasing the professionalism, engagement, and allure of Douyin’s live e-commerce streaming.

3.1.2. Overall and Sampling Methodology

The questionnaire survey approach was utilized in this study to obtain information from consumers about their live purchasing experiences with green agricultural products. A total of 350 valid questionnaires were received. According to Nielsen’s “China Live E-commerce Users White Paper” regional distribution statistics, the questionnaire uses the stratified quota sampling method, and the samples primarily come from Jiangsu, Anhui, Zhejiang, and Guangdong, which are typical coastal provinces in southeast China. The agriculture–tourism integration model has been studied previously and is relatively mature in these regions, which have the most developed e-commerce economies in China. They also have a large number of live streaming users and a plentiful supply of green agricultural products, which contributes to the high degree of credibility of the sample data.

Key factors including gender, age (18–50), and monthly income are included in the sample quota. Although quota sampling was used to adjust for some demographic characteristics, the sample was concentrated in highly digitalized and economically developed coastal areas. This study’s conclusions may therefore not be as applicable in frontier regions, traditional agricultural areas, or less developed inland locations. The subjects had to use situational backtracking (SBT) to relive their most recent live shopping experience to complete the questionnaire, and the data results successfully satisfied the PLS-SEM analysis requirements.

3.1.3. Research Instruments and Scales

The scale development accounted for the unique characteristics of e-commerce live streaming, following and adapting Hinkin’s scale development paradigm. All variables were measured using a 5-point Likert scale (“1” = strongly disagree, “2” = disagree, “3” = neutral, “4” = agree, “5” = strongly agree). The scale mainly draws on or is adapted from mature research: The social networks, infrastructure, and resource endowment items were taken from Moga [60], Nakasone [8], and Chevalier and Mayzlin [61]. Michell [62], Bansal [63], Ridings et al. [64], Catherine and Meng Fei [65] were the sources of the streamer characteristic items. Shi Zheng was cited in the green consumption cognition items, Sweeney [66] and Petrick [67] were the basis for the perceived value items, Liu Xiaoli was cited in the trust measurement items; and Ajzen et al. [68], Zeithaml et al., and Sukiwere’s works were adapted for the green agricultural product purchase intention items.

3.1.4. Data Collection

The data was collected using “Wenjuanxing” (https://www.wjx.cn/), a popular online survey tool in China. To ensure study compliance, the questionnaire ensured all participants were over the age of eighteen and were willing to take part by clearly stating anonymous principles, academic data utilization, and confidentiality pledges. This indicated thatdata collection was sufficient and representative, contributing to the reliability of this study’s findings [69,70]. Using a non-probability sampling technique, data collection began in November 2024 with an emphasis on typical coastal provinces in the southeast, such as Jiangsu, Anhui, Zhejiang, and Guangdong Province, to improve sample representativeness and the relevance of the study findings.

3.2. Sample Feature Description

According to the quantitative examination of live streaming consumption behavior, this study conducted a detailed analysis of the survey questionnaire as presented in

Table 1. Characteristics of the study sample.

Variable	Category	Frequency	Percentage
Gender	Male	149	42.6%
	Female	201	57.4%
Age	18–30	149	42.6%
	31–50	129	36.8%
	51–60	72	20.6%
Occupation	Students	105	30%
	Corporate staff	49	14%
	Civil servants and public institutions	168	48%
	Other	28	8%
Monthly income	<3000	106	30.3%
	3000–5000	217	62.0%
	>5000	27	7.7%
Whether to watch and purchase	Watched and purchased	219	67.38%
	Watched without purchase	78	24%
The main reasons for purchasing in the live stream	The price discount is substantial	56	16%
	Agricultural products are quite attractive	46	13.1%
	Support local farmers	45	12.9%
	A personal fondness for the live-streamer	59	16.9%
	Detailed and professional explanations	56	16%
	Agricultural product brand	31	8.9%
	Popularity and atmosphere	57	16.3%

. Women made up 57.4% of the sample, while men made up 42.6%. The youth segment (18–30 years old) dominated at 42.6%, suggesting a youthful tendency. Students (30%), civil servants, and public institutions (31.7%) made up the majority of the workforce, with corporate staff (14%) rounding out the mix. The majority of the population had undergraduate degrees or more, and the distribution of income followed a steady trend that was typical of metropolitan Chinese income structures. The data gathered from the sample confirms its representativeness and shows no signs of sampling bias.

4. Analysis and Research Findings

4.1. Reliability and Validity Analysis

To methodically evaluate the rigor of the measurement model, this study used the dual reliability and validity assessment framework put forward by Fornell and Larcker. All latent variables demonstrated convergent validity indices that were noticeably higher than academic standards, as shown in

Table 2. Validity and reliability results.

Variables	Cronbach’s Alpha	CR	AVE
Social Networks	0.838	0.903	0.756
Resource Endowments	0.837	0.902	0.755
Infrastructure	0.854	0.911	0.774
Professionalism	0.866	0.909	0.714
Interactivity	0.882	0.919	0.740
Attractiveness	0.885	0.921	0.745
Green Consumption Cognition	0.789	0.874	0.699
Perceived Value	0.858	0.903	0.699
Trust	0.798	0.879	0.709
Purchase intention	0.784	0.871	0.696

. Excellent measure stability in observed variables for latent constructs has been demonstrated by convergent reliability (CR) values that either reached or surpassed the 0.70 threshold. Values for Average Variance Extracted (AVE) exceeded the 0.50 threshold. In particular, the range of values for AVE was 0.696 to 0.774, while the range of CR values was 0.871 to 0.921. These findings validate the concept of convergent validity of the suggested measurement framework by demonstrating that all factor loadings, CR, and AVE values meet the suggested criteria.

In addition to assessing the discriminant validity, we used both the Fornell–Larcker and “Heterotrait–Monotrait Ratio” (HTMT) criteria. The results shown in

Table 3. Discriminant validity: Fornell–Larcker criterion.

	ATT	CGA	IF	INT	PI	PV	RE	SN	SPE	TR
ATT	0.816
CGA	0.396	0.760
IF	0.030	0.083	0.816
INT	0.158	0.503	0.030	0.807
PI	0.276	0.759	0.288	0.384	0.763
PV	0.101	0.611	0.446	0.354	0.830	0.773
RE	0.018	0.495	0.053	0.020	0.401	0.461	0.795
SN	0.035	0.395	0.039	0.021	0.358	0.477	0.076	0.800
SPE	0.037	0.587	0.060	0.192	0.528	0.435	0.068	0.123	0.786
TR	0.359	0.752	0.404	0.501	0.810	0.697	0.410	0.396	0.568	0.7722

show that the diagonal AVE scores (“the square root of AVE”) for each variable were greater than the variable’s correlations with any other variable, as suggested by Fornell and Larcker. Likewise, the results shown in

Table 4. Discriminant validity: Heterotrait–Monotrait ratio (HTMT).

	ATT	CGA	IF	INT	PI	PV	RE	SN	SPE	TR
ATT
CGA	0.390
IF	0.034	0.095
INT	0.156	0.499	0.040
PI	0.283	0.758	0.291	0.387
PV	0.100	0.580	0.440	0.340	0.836
RE	0.044	0.494	0.070	0.045	0.387	0.445
SN	0.036	0.379	0.045	0.080	0.354	0.470	0.088
SPE	0.050	0.584	0.069	0.192	0.521	0.418	0.073	0.122
TR	0.340	0.736	0.399	0.499	0.814	0.665	0.412	0.395	0.566

show that the HTMT values were below 0.90, as proposed by Henseler et al., indicating satisfactory discriminant validity.

4.2. Confirmatory Factor Analysis

Confirmatory factor analysis (CFA) is a statistical method that tests the structural links among observed variables and theoretically proposed underlying elements to evaluate the validity of measurement models. Confirming the measurement of the model’s integrity, the CFA findings showed substantial factor loadings of each observable variable on its corresponding latent components. Structural equation modeling (SEM) analysis is firmly based on the strong measurement validity that ensures latent variables are accurately represented by their observed indicators.

According to the model analysis, our structural model showed a good model fit with the levels suggested by Hair et al. [71], as follows: X² = 1859.232, and the Q² values for CGA (0.657), PI (0.485), PV (0.568) and TR (0.672) are all greater than zero, indicating that the model possesses predictive relevance. Finally, we validated the model using the standardized root mean square residual (SRMR), based on the recommendations of Henseler et al. An SRMR value of 0.09 or lower is recommended, indicating a good model fit [71]. Our SRMR value of 0.066 suggests that the model achieves a strong fit to the data, meeting the recommended threshold.

4.3. Empirically Validated Research Findings

To analyze the comprehensive influence relationships among these variables, a structural equation model (SEM) was developed based on the results of the reliability and validity tests. As shown in Figure 2. The SEM showed that streamer characteristics and networks, resources, and facilities were independent variables, while green consumption cognition, trust, and perceived value were mediating variables. The dependent variable was purchase intention. The impact of each variable on consumers’ propensity to buy green agricultural products was evaluated using SEM analysis.

As illustrated in Figure 2, the confirmatory factor analysis (CFA) model demonstrates adequate fit to the data according to established goodness-of-fit criteria. The standardized factor loadings from the path analysis are presented in

Table 5. Path analysis regression weights.

Path	Initial Sample	Sample Mean	SD	T-Statistics	p-Value	Interpretation
SN→CGA	0.255	0.254	0.029	8.761	***	Significance
SN→PV	0.345	0.344	0.037	9.416	***	Significance
SN→TR	0.258	0.260	0.029	8.884	***	Significance
RE→CGA	0.368	0.368	0.036	10.261	***	Significance
RE→PV	0.337	0.338	0.034	9.982	***	Significance
RE→TR	0.287	0.288	0.031	9.982	***	Significance
IF→CGA	0.010	0.008	0.036	0.284	0.777	Not significant
IF→PV	0.339	0.337	0.033	10.141	***	Significance
IF→TR	0.292	0.289	0.030	9.737	***	Significance
SPE→CGA	0.392	0.390	0.035	11.115	***	Significance
SPE→PV	0.264	0.264	0.034	7.650	***	Significance
SPE→TR	0.368	0.364	0.033	11.196	***	Significance
INT→CGA	0.319	0.316	0.034	9.415	***	Significance
INT→PV	0.242	0.239	0.036	6.781	***	Significance
INT→TR	0.323	0.322	0.035	9.318	***	Significance
ATT→CGA	0.274	0.277	0.031	8.831	***	Significance
ATT→PV	0.022	0.023	0.040	0.543	0.588	Not significant
ATT→TR	0.237	0.237	0.031	7.594	***	Significance
CGA→PI	0.244	0.246	0.045	9.774	***	Significance
PV→PI	0.415	0.416	0.043	5.375	***	Significance
TR→PI	0.285	0.283	0.046	6.204	***	Significance

*** p < 0.01.

.

The computational results demonstrate that social networks exert significant positive regression effects on green consumption cognition, perceived value, and trust, with significance tests yielding p-values < 0.001. These findings substantiate hypotheses H1a–H1c. Resource endowment exhibits the strongest positive influence across all three mediating variables, with highly significant effects (p < 0.001), thereby supporting hypotheses H2a–H2c. Infrastructure significantly impacts perceived value and trust (supporting H3b and H3c) but shows no significant effect on green consumption cognition (p = 0.777). Consequently, hypothesis H3a is not supported.

E-commerce streamers’ professionalism demonstrates the most pronounced effect on green consumption cognition while also significantly enhancing perceived value and trust (p < 0.001), validating hypotheses H4a–H4c. Streamer interactivity positively influences all three mediating variables, confirming hypotheses H5a–H5c. Streamer attractiveness significantly affects green consumption cognition and trust (p < 0.001, supporting H6a and H6c), but demonstrates no significant effect on perceived value, leading to the rejection of hypothesis H6b.

Finally, green consumption cognition, perceived value, and trust all exhibit significant positive regression effects on consumers’ purchase intention (p < 0.001), providing support for hypotheses H7–H9.

5. Discussion

Based on the SOR theoretical framework, this study examines how the external stimulation of live streaming e-commerce affects consumers’ intentions to buy green agricultural products in the context of agritourism integration. Echoing both earlier research and new developments, the presentation of the findings of this study is predicated on the queries and theories of earlier investigations.

The impact of resource endowment is the most substantial among them, with path coefficients ranging from 0.368 to 0.337. It is in line with the conclusions of Jiang et al. [24] and Yang et al. [57] regarding how resource endowment affects the degree of farm machinery and green agriculture productivity. Additionally, it confirms Sweeney’s [66] theory of perceived value, which considers that consumers’ primary determinant of decision-making is their assessment of the cost-effectiveness of green agricultural products. This study emphasizes the distinct value transfer path of resource endowment in contrast to Dong et al.’s [1] study, which just concentrates on conventional e-commerce or a specific agritourism scenario.

Through an empirical study, Delarocas [26] demonstrated the significance of social networks in influencing consumer purchase behavior, including online word-of-mouth and live streaming environments. The interaction between social networks is mostly demonstrated by the fact that viewers of live streams are frequently swayed by the number of watchers, which, in turn, affects their decision to remain on the page and make a purchase. This supports Jiang et al.’s [24] research on the influence aspect of social network opinion leaders, emphasizing the importance of creating a positive interactive atmosphere and word-of-mouth risks for agricultural products in live agritourism streams.

This study found that infrastructure exerted no significant influence on green consumption cognition. This finding contrasts with research by Nakasone [8], which emphasized the fundamental role of logistics and support systems in shaping green perceptions. This discrepancy may arise because consumer cognition of “intrinsic green attributes” primarily focuses on product characteristics (e.g., organic certification), whereas logistical facilities mainly affect “purchasing feasibility”. Furthermore, the generally well-developed infrastructure in the coastal regions of China, where the sample was drawn, might have diminished its observable impact on cognition. Consequently, the generalizability of this conclusion requires validation in future research.

The professionalism and interactivity of livestream hosts were demonstrated to be crucial external stimuli. Meng, L.M. [72], conceptualizes e-commerce livestream hosts as influencers who affect consumers’ purchase intentions by leveraging their recommendations. This aligns with Li et al.’s [36] findings regarding host professionalism and impulse buying, as well as Sun et al.’s [44] research on how interactivity influences consumer psychology and behavior. The results underscore the core value provided by hosts, which lies in offering professional information, fostering emotional connections, and promptly addressing consumer needs.

However, contrasting with studies that emphasize the role of host physical attractiveness, such as those of Peng et al. [38] and Zhao, J. [50] this study revealed that host attractiveness exerted a significant effect solely on green consumption cognition and trust, but not on perceived value. This indicates that within the agritourism-integrated livestreaming context, the host’s expertise and interactive competence are more critical than mere physical attractiveness in shaping value perception. Attractiveness likely functions primarily in capturing initial attention and generating liking, yet its direct contribution to perceived value appears limited.

Concurrently, this study confirmed that perceived value and trust act as the strongest mediating variables influencing purchase intention. This aligns with Mouloudj, K.’s [51] research on trust’s impact on attitudes towards green agricultural products, as well as the findings of Yang et al. [22] and Xu et al. [18] concerning perceived value. Furthermore, it refines the dual-influence pathway proposed by Wang X. et al. [73]. Notably, the present research specifically clarifies the nature of green trust, emphasizing the authenticity conviction regarding the product’s genuine “green” essence. This conviction is demonstrably reinforced within agritourism-integrated livestreams through authentic settings showcasing production processes.

6. Conclusions

Grounded in the Stimulus–Organism–Response (SOR) framework, this study employed structural equation modeling (SEM) to analyze 350 valid consumer questionnaires. The results demonstrate that within the agritourism-integrated context, resource endowment serves as the primary stimulus triggering purchase intention. Social networks, host professionalism, and interactivity function as key stimuli. These stimuli effectively enhance consumers’ green consumption cognition, perceived value, and trust. Furthermore, all of these mediating effects exhibit significant positive influences on purchase intention.

6.1. Contributions

The contributions of this study are manifested in three primary dimensions: theoretical, practical, and societal. Firstly, regarding theoretical contribution, this research innovatively applies the SOR theory to the complex context of “agritourism integration + e-commerce livestreaming”. It elucidates the pathway through which multiple cognitive mediators influence purchase intention. This approach addresses a gap in traditional models applied to intersectional domain research, deepens the understanding of consumer decision-making mechanisms within “online-offline” dual-integration scenarios, and provides a novel perspective for applying this theory to new digital-era consumption contexts.

In terms of practical and societal contributions, the “agritourism integration + e-commerce livestreaming” model offers effective strategies for promoting green consumption, facilitating rural digital transformation, and informing policy formulation. Furthermore, leveraging e-commerce livestreaming can enhance public awareness and trust in green agricultural products. It effectively transforms green agricultural products and agritourism resources into consumption drivers. This dual approach simultaneously addresses consumer demand for a green lifestyle while propelling the digital upgrading of rural industries. Consequently, it injects sustained momentum into achieving the Sustainable Development Goals (SDGs) and plays a proactive role in fostering a resource-efficient and environmentally friendly society.

6.2. Policy Recommendations

Based on the results of this study, this paper puts forward the following suggestions for e-commerce platforms, streamers, and rural policy makers. First, e-commerce platforms should focus on the original resources of green agricultural products, utilize immersive technologies such as drone aerial photography to enhance the realism of the scene, and extend the live stream content to show the reasonable ecological environment of farms, highlighting the beneficial role of resource endowment. These should also promote the branding of streamers, discover streamers with strong expression ability, show the advantages of the platform, and establish a reliable image [74]. Secondly, the platforms need to strengthen the training of streamers in agricultural knowledge, green ecological value explanation, etc., so that they can professionally convey the core value of products. These also need to improve the responsiveness of the streamer team to shorten the psychological distance and enhance trust through efficient real-time Q&A. Utilizing social network relationships, users are encouraged to share their edible experiences, green consumption tips, etc., and use their social influence to attract new users to place orders [75,76].

For rural policymakers, first, priority should be given to improving the cold chain logistics system, solving infrastructure problems, reducing losses to reduce losses and consumption risks. Second, agricultural live streaming should be encouraged and the integration of live e-commerce and agricultural tourism resources promoted. Third, the green agricultural product standard system should be improved, the certification supervision strengthened, and the public refined. Finally, training and resource support for farmers or enterprises in green production technology and the operation of live e-commerce platforms should be provided to promote the development of green live e-commerce.

6.3. Limitations and Future Research

This study acknowledges several limitations. The sample was predominantly drawn from economically developed and digitally advanced coastal regions in Southeast China (Jiangsu, Anhui, Zhejiang, Guangdong), with a relatively high proportion of young female participants. Consequently, the generalizability of the findings to inland less-developed regions, traditional agricultural areas, frontier regions, or diverse demographic groups (e.g., different ages, genders) requires further validation. Future research should broaden the sampling scope, employing a stratified sampling frame and implementing quota sampling based on criteria such as region, socioeconomic status, and cultural background, potentially guided by sources like the China Regional Economic Statistical Yearbook, to enhance the external validity and applicability of the results.

The primary reliance on self-reported questionnaires presents limitations due to finite sample size and potential subjectivity and response biases. Future studies could collaborate with e-commerce platforms and integrate experimental methodologies, such as eye-tracking and Mixed Reality (MR) technology, to objectively capture consumers’ attention, focus, and cognitive transformation processes in real-time [77,78]. Additionally, constructing panel data would allow for a deeper investigation into the dynamic impact of nuanced host characteristics like linguistic style and emotional expression.

While the applied SOR model effectively explains the linkages between external stimuli and internal organismic states, real-world consumption contexts are inherently complex. Purchase intentions are also influenced by situational variables such as socio-cultural norms, consumption trends, and public events. Future research should incorporate additional moderating variables, establish cross-cultural and cross-regional theoretical frameworks, or develop more comprehensive models (e.g., integrating Theory of Planned Behavior (TPB) or Value–Belief–Norm (VBN) theory). This would facilitate comparative analyses of how the agritourism-integrated livestreaming model differentially impacts green agricultural product purchase intentions across diverse socioeconomic and cultural contexts.