Abstract
The concept of “digital business to revitalize agriculture” has emerged as a novel strategy to transform rural economies through the integration of digital technologies. This shift has led to the rise of rural e-commerce ecosystems, which are complex networks of interdependent actors and digital elements that drive innovation and value creation. However, the development of these ecosystems is highly complex and cannot be fully understood through traditional linear analyses. This study explores the development mechanisms of rural e-commerce ecosystems using Complex Adaptive Systems (CAS) theory and fuzzy-set Qualitative Comparative Analysis (fsQCA). Empirical data from multiple counties reveal several configurational pathways associated with high and low levels of ecosystem development. The findings indicate that no single digital element can independently drive high-level development; instead, synergistic interactions among digital production, supply chains, marketing, and financial support are crucial. Distinct configurations, such as the “Quadruple-Ring Synergistic Drive Type,”“Production-Marketing Dual-Core Drive Type,” and “Core Triple-Ring Drive Type,” highlight the importance of multi-stakeholder collaboration and adaptable strategies. This study provides a holistic framework through CAS and QCA, challenging traditional linear perspectives. It offers actionable insights for policymakers, practitioners, and stakeholders, emphasizing the need for flexible and context-specific approaches to promote sustainable rural economic growth.
Plain Language Summary
The concept of “digital business to revitalize agriculture” has emerged as a novel strategy to transform rural economies through the integration of digital technologies. This shift has led to the rise of rural ecommerce ecosystems, which are complex networks of interdependent actors and digital elements that drive innovation and value creation. However, the development of these ecosystems is highly complex and cannot be fully understood through traditional linear analyses. This study explores the development mechanisms of rural e-commerce ecosystems using Complex Adaptive Systems (CAS) theory and fuzzyset Qualitative Comparative Analysis (fsQCA). Empirical data from multiple counties reveal several configurational pathways associated with high and low levels of ecosystem development. The findings indicate that no single digital element can independently drive high-level development; instead, synergistic interactions among digital production, supply chains, marketing, and financial support are crucial. Distinct configurations, such as the “Quadruple-Ring Synergistic Drive Type,”“ProductionMarketing Dual-Core Drive Type,” and “Core Triple-Ring Drive Type,” highlight the importance of multistakeholder collaboration and adaptable strategies. This study provides a holistic framework through CAS and QCA, challenging traditional linear perspectives. It offers actionable insights for policymakers, practitioners, and stakeholders, emphasizing the need for flexible and context-specific approaches to promote sustainable rural economic growth.
Keywords
Introduction
In the modern era, the concept of “digital business to revitalize agriculture” has been proposed by the CPC Central Committee and the State Council as a novel strategy following the digital rural strategy. It represents a significant shift in agricultural development, emphasizing the integration of digital technologies into rural economies to drive growth and innovation. The digitization of agriculture is a transformative process that extends beyond technological upgrades, encompassing the use of digital tools in various aspects of rural life, such as e-commerce, online banking, e-government services, and social media (Leong et al., 2016; Y. Wang et al., 2024).
Digital tools have become increasingly important in rural areas, addressing challenges like limited access to markets, financial services, and information. They have the potential to bridge the digital divide, empower rural communities, and enhance the competitiveness of agricultural products in the global market (Y. Wang et al., 2024). The rapid evolution of digital business in agriculture is evident in phenomena such as the rise of “Taobao Villages” and the recognition of the “Top 100 Counties for the digitalization of agricultural products,” which highlight the success of rural e-commerce in promoting economic development.
Rural e-commerce ecosystems have garnered significant attention for their potential to integrate rural economies into broader markets, enhance agricultural value chains, and improve the quality of life for rural residents (Leong et al., 2016; Y. Wang et al., 2024). These ecosystems are complex networks of interdependent actors and digital elements that interact to create value and drive innovation. However, the development of rural e-commerce ecosystems is influenced by multiple factors, including digital infrastructure, governance, production capabilities, supply chains, marketing strategies, financial support, and digital lifestyles (Z. Huang et al., 2025; Leong et al., 2016; Y. Wang et al., 2024; W. Wu et al., 2020). Understanding these interactions is crucial for promoting sustainable rural economic growth.
Despite their importance, existing studies on rural e-commerce ecosystems have predominantly focused on individual factors or linear relationships, often neglecting the complex interdependencies and synergies among multiple elements. For example, some studies have examined the impact of digital infrastructure (Leong et al., 2016; Lu et al., 2024; Y. Wang et al., 2024; J. Zhou et al., 2021), while others have focused on financial support (L. Li et al., 2019; Chao et al., 2021). These traditional approaches may oversimplify the dynamics of rural e-commerce ecosystems, failing to capture the full complexity of their development processes. This gap highlights the need for a more holistic and configurational approach that considers the interactions among various digital elements and stakeholders.
To address this gap, this study adopts the Complex Adaptive Systems (CAS) theory as its theoretical foundation. CAS theory posits that ecosystems are dynamic, non-linear, and self-organizing entities characterized by interdependent agents and emergent properties (L. Huang et al., 2024; Leong et al., 2016; L. H. Sun & Shu, 2025; D. Zhang et al., 2024). In the context of rural e-commerce, this means that ecosystem development is influenced by the interactions among multiple digital elements and stakeholders, rather than by any single factor. CAS theory emphasizes understanding these interactions and configurations, as different combinations of elements can lead to diverse outcomes.
To operationalize the CAS framework, this study employs fuzzy-set Qualitative Comparative Analysis (fsQCA), a configurational method that examines multiple pathways leading to a particular outcome (Liao et al., 2024; Ordanini et al., 2014; S. N. Zhang et al., 2023). fsQCA focuses on the interactions among conditions and their configurations, rather than on individual factors in isolation (Ragin, 2010; Tekic & Tekic, 2024). By identifying different configurations of digital elements associated with high and low levels of rural e-commerce ecosystem development, fsQCA provides a nuanced understanding of the mechanisms at play. This approach highlights the adaptability and flexibility required for ecosystem development in diverse contexts.
This study aims to fill the research gap by exploring the development mechanisms of rural e-commerce ecosystems through a configurational lens. It seeks to answer the following research questions:
① How do different configurations of digital rural elements influence the development of rural e-commerce ecosystems?
② What are the unique and asymmetric configurations associated with high and low levels of rural e-commerce ecosystem development?
③ How do interactions among key stakeholders—such as new farmers, e-commerce platforms, logistics companies, governments, and financial institutions—shape these configurations and drive ecosystem development?
This study contributes to the literature by providing a comprehensive understanding of rural e-commerce ecosystem development through the lens of CAS theory and fsQCA. It offers a flexible framework for analyzing the complex dynamics of these ecosystems, emphasizing the importance of configurational approaches in capturing their adaptability and resilience. The findings have significant implications for policymakers, practitioners, and researchers seeking to promote sustainable rural economic growth through e-commerce.
The remainder of this paper is structured as follows: Section 2 provides a detailed review of the relevant literature on the concept and key components of the rural e-commerce ecosystem, its development mechanisms, the relationship between digital village development and the rural e-commerce ecosystem, and applications of Complex Adaptive Systems (CAS) theory and the fsQCA method. Section 3 outlines the research methodology, including study design, data collection, and fsQCA procedures. Section 4 presents the findings of the fsQCA analysis, identifying configurational pathways associated with high and low levels of rural e-commerce ecosystem development. Section 5 discusses the implications of these findings for theory and practice, while Section 6 concludes the paper by summarizing the key contributions and suggesting avenues for future research.
Literature Review
The Concept and Key Actors of the Rural E-commerce Ecosystem
The rural e-commerce ecosystem refers to a complex, organic system facilitated by internet information technology. Centered on the distribution of agricultural products and the supply of industrial goods to rural areas, it involves the synergistic interaction of multiple actors, efficient flow of resources and factors, and co-creation and sharing of value (Z. Wang, 2025). Its core function is to bridge the urban-rural geographical divide through digital connectivity, thereby restructuring the rural economic value chain (Ren, 2023). The composition of actors within this ecosystem is characterized by diversity and network interdependence.
Core actors include farmers, cooperatives, and agribusinesses engaged in the production, processing, and sale of agricultural products. As direct suppliers, their willingness and capacity to participate (e.g., digital skills, brand awareness) significantly influence the ecosystem’s vitality (X. Chen et al., 2024; Fan, 2023). E-commerce platforms (e.g., Taobao Villages, Pinduoduo) serve as critical hubs connecting supply and demand, providing essential infrastructure and services for transactions, payment, logistics, and marketing (J. Gao et al., 2024; Z. Wang, 2025).
Within the supporting layer, logistics and delivery companies address the “first and last mile” challenges; the coverage depth and efficiency of their service networks are paramount (Q. Liu et al., 2024; H. Zhao & Li, 2023). Financial service institutions (e.g., digital inclusive finance) provide credit, insurance, and payment solutions to farmers and SMEs, alleviating financing constraints (X. Guo et al., 2025; Z. Huang et al., 2025; W. Wang & Yan, 2025). Information technology service providers offer technical support for e-commerce operations, data analytics, and digital marketing (Ahmed et al., 2024).
Actors in the environmental layer include government agencies that foster a conducive environment through policy formulation (e.g., Digital Village Strategy, E-commerce Demonstration Projects), infrastructure development (broadband, cold chain), market regulation, and talent cultivation (J. Gao et al., 2024; P. Guo et al., 2024; Liu M.Z.Y. & Liu H., 2024). Social organizations (e.g., associations, research institutes) contribute through standard-setting, knowledge dissemination, and resource brokering (Singh et al., 2023). Consumers represent the endpoint of value realization, with their evolving demand preferences driving ecosystem evolution (Jing et al., 2025; H. Zhou et al., 2025). These actors are tightly coupled through the flows of information, goods, and capital, forming a symbiotic network (L. Li et al., 2024; Z. Wang et al., 2024).
Development Mechanisms of the Rural E-commerce Ecosystem
The rural e-commerce ecosystem functions as an open system operating far from equilibrium (Deng et al., 2024). It sustains dynamic order and achieves upgrading by continuously exchanging matter, energy, and information with its external environment (urban markets, advanced technologies, external capital, policy information)—introducing “negative entropy flow” to overcome internal resistances (e.g., fragmented smallholder farming, logistical bottlenecks; X. Li et al., 2024; D. Sun et al., 2024; Xu & Yang, 2025; Yuan et al., 2025). Ecosystem development is non-linear, characterized by actors constantly learning, adapting strategies, and responding to environmental shifts through interaction (Lu & Wu, 2025; J. Lin & Tao, 2024). Early successful models (e.g., “Taobao Villages”) can create path dependence, influencing subsequent development trajectories (J. Gao et al., 2024; Z. Wang, 2025). This evolutionary process often involves the emergence of new actors (e.g., live-streaming influencers, supply chain service providers) and the restructuring of existing configurations (Deng et al., 2024; L. Zhao, 2024).
The emergence and growth of the rural e-commerce ecosystem result from the interplay of multiple driving mechanisms, exhibiting significant complex adaptive system (CAS) characteristics. Growing urban consumer demand for high-quality, distinctive agricultural products, coupled with rural residents’ demand for convenient access to diverse industrial goods and services, constitutes the core market pull (X. Wang et al., 2025; Y. Wang & He, 2024). The proliferation and penetration of technologies like the internet, mobile payment, big data, and artificial intelligence significantly reduce transaction costs and expand market boundaries, acting as a key technological push (G. Li, 2024; Ahmed et al., 2024; Mack et al., 2024). Ecosystem development relies on effective synergy among actors. Platforms empower farmers to enhance operational capabilities (K. Gao & Qiao, 2025), governments guide resource allocation and policy support (J. Gao et al., 2024), while logistics and financial firms provide critical support services (W. Wang & Yan, 2025; H. Zhao & Li, 2023). Farmers engage in direct interaction and value co-creation with consumers through live-streaming e-commerce (X. Chen et al., 2024; L. Zhao, 2024) and community marketing (R. Zhou, 2025). Governmental top-level design and policy intervention serve as crucial catalysts. Initiatives like the “Digital Village” strategy and “Comprehensive E-commerce Demonstration in Rural Areas” directly drive infrastructure improvement, talent training, and market actor cultivation (Liu M.Z.Y. & Liu H., 2024; X. Li et al., 2024). Institutional innovations, such as exploring supply chain finance models suited to rural contexts (S. Bai & Jia, 2023) and land policies supporting returning entrepreneurship (Cao & Liang, 2025; Y. Sun & Ren, 2025), inject vitality into the ecosystem.
Synergistic Advancement of Digital Village Construction and Rural E-commerce Development
Digital village construction provides foundational support for the rural e-commerce ecosystem, while rural e-commerce represents its most dynamic application scenario and value manifestation. The proliferation of digital infrastructure—broadband networks, mobile communications, and the Internet of Things—is a prerequisite for rural e-commerce, significantly enhancing information accessibility, and connectivity (X. Liu et al., 2024; Mack et al., 2024). The application of smart agriculture technologies (e.g., IoT monitoring, precision farming) improves product quality and supply chain transparency, strengthening e-commerce competitiveness (Cheng et al., 2024; X. Li et al., 2024).
Data generated during digital village construction—encompassing agricultural production, rural governance, and farmer livelihoods—can be integrated and analyzed to precisely map market demand, optimize cultivation structures, and guide targeted marketing, thereby driving efficient ecosystem operation (Ahmed et al., 2024; X. Liu et al., 2024). Digital village initiatives emphasize enhancing farmers’ digital literacy and skills, directly boosting their capacity to engage in e-commerce (e.g., using platforms, conducting live streams; X. Chen et al., 2024; K. Gao & Qiao, 2025). Concurrently, they attract returning migrant workers and university graduates as entrepreneurs (Cao & Liang, 2025; Qi et al., 2025), injecting new ideas and dynamism into the ecosystem. Digitized rural governance (e.g., “Internet + Government Services”) improves public service efficiency and optimizes the business environment (J. Wang & Vansant, 2025). The extension of smart logistics and digital inclusive financial services directly addresses critical pain points in e-commerce development (Z. Huang et al., 2025; H. Zhao & Li, 2023). In turn, thriving rural e-commerce, by generating economic benefits and demonstrating the value of digital technologies, incentivizes farmers, businesses, and governments to invest further in digital infrastructure, fostering its continuous improvement and the deepening of application scenarios (Liu M.Z.Y. & Liu H., 2024; Xin et al., 2025). E-commerce activities also catalyze the modernization of farmer mindsets and transformations in rural social structures (L. Li & Song, 2023; X. Wu & Liu, 2025).
Application of CAS Theory and fsQCA Methodology in Related Research
Complex Adaptive System (CAS) theory and fuzzy-set Qualitative Comparative Analysis (fsQCA) methodology provide powerful analytical tools for understanding the formation, evolutionary pathways, and multiple conjunctural causations within complex systems like the rural e-commerce ecosystem. CAS theory emphasizes that systems comprise adaptive agents whose interactions and learning behaviors drive systemic evolution and emergence. This framework has been effectively applied to analyze rural e-commerce ecosystems. First, regarding agent adaptive behavior, CAS explains how farmers adjust their e-commerce participation strategies (e.g., adopting live-streaming, choosing platforms) based on environmental changes (e.g., policy, market, technology; X. Chen et al., 2024; K. Gao & Qiao, 2025). It also elucidates how platforms and local governments formulate differentiated strategies considering local resources and agent feedback (J. Gao et al., 2024; J. Lin & Tao, 2024). Second, in system evolution and path analysis, CAS helps understand how the ecosystem emerges from initial states (e.g., early adopters) through agent interactions (e.g., imitation, competition, cooperation), rule adjustments (e.g., policy iterations, platform rules), and environmental shifts (e.g., consumption upgrading, technological breakthroughs), leading to complex structures (e.g., industrial clusters, service ecosystems) and diverse development paths (e.g., live-streaming driven, supply-chain integrated; Deng et al., 2024; Lu & Wu, 2025; Z. Wang, 2025). It highlights non-linear features, such as threshold effects in the impact of digital investment on rural enterprise resilience (Lu & Wu, 2025). Third, in ecosystem resilience analysis, CAS is used to study how the ecosystem responds to external shocks (e.g., pandemics, market volatility) and internal challenges (e.g., logistics disruptions, product homogeneity), analyzing how adaptation strategies of different agents (farmers, platforms, government) collectively shape overall system resilience (J. Lin & Tao, 2024; Zhong et al., 2024).
fsQCA methodology, adept at handling small-to-medium-N cases, identifies combinations of conditions (configurational paths) leading to specific outcomes (e.g., thriving e-commerce, significant farmer income growth). Its value in rural e-commerce research lies in: ① Identifying Multiple Development Paths: Moving beyond single linear explanations, fsQCA reveals distinct combinations of conditions (configurations) leading to successful e-commerce development across regions. For instance, high e-commerce development may be driven by combinations like “strong government support + robust logistics + high farmer digital literacy” or “distinctive industrial base + strong platform intervention + effective financial support” (J. Gao et al., 2024; P. Guo et al., 2024; L. Li et al., 2024). ② Analyzing Interdependencies: fsQCA examines how different factors (e.g., policy intensity, infrastructure level, social capital, industry uniqueness) interact (complement or substitute) to jointly influence outcomes (D. Wang, 2023; Z. Zhang et al., 2025). ③ Explaining Equifinality: It demonstrates that different combinations of conditions (configurations) can lead to the same outcome (e.g., rural industrial revitalization), aligning with the CAS concept of “multiple paths to evolution” (M. Chen & Long, 2024; Xu & Yang, 2025). ④ Capturing Causal Complexity: fsQCA handles the interdependence of explanatory variables and causal complexity, offering a closer fit to the real-world context of ecosystem development (Ma et al., 2024; Schwering et al., 2023).
Research Gaps and Future Directions
Existing research provides valuable theoretical insights and empirical evidence for understanding the conceptual framework, actor interactions, driving mechanisms, and synergistic relationship with digital village construction in the rural e-commerce ecosystem. Explorations using CAS and fsQCA are also promising. However, limitations and future research directions remain:
First, theoretical integration requires deepening: While CAS theory has been introduced, systematic and deep empirical integration of its core concepts (e.g., adaptability, non-linearity, emergence, building blocks) into the analysis of the ecosystem’s evolutionary mechanisms remains insufficient (J. Lin & Tao, 2024; Lu & Wu, 2025). Future research needs more granular modeling of agent adaptation rules and tracking how micro-level behaviors lead to macro-level emergence.
Second, actor heterogeneity and interaction networks are inadequately captured: Current studies pay insufficient attention to the heterogeneous goals and behavioral patterns of different actor types (e.g., smallholders vs. cooperatives, large platforms vs. local service providers) and their varying positions, roles, and influence within complex interaction networks (K. Gao & Qiao, 2025; R. Zhou, 2025). Methods like Social Network Analysis (SNA) could more precisely map relational structures and their impact on system functions.
Third, research on dynamic evolutionary processes is weak: Most studies offer static or cross-sectional analyses. There is a lack of longitudinal tracking of how the ecosystem dynamically evolves over time—through stages of formation, growth, stability, transformation, or decline—driven by external environmental shifts (technological change, policy adjustments, market fluctuations) and internal agent adaptation (Deng et al., 2024; Z. Wang, 2025). Longitudinal studies and Agent-Based Modeling (ABM) are key avenues to address this gap.
Fourth, QCA’s potential is underutilized: Current QCA applications focus mainly on identifying static condition sets for successful outcomes, with limited use of temporal QCA to track path evolution (e.g., “reverse entrepreneurship” mentioned in Qi et al., 2025). There is also scant effort to connect QCA-derived configurations to specific CAS mechanisms (e.g., how particular adaptation rules lead to specific emergent structures) for deeper theoretical dialog and validation.
In conclusion, future research should deepen mechanism exploration within the CAS framework, enhance modeling of multi-actor heterogeneity and dynamic interactions, and innovatively apply QCA and other methods to uncover complex causality and multiple pathways. This will provide more systematic, profound, and policy-relevant scientific understanding of the complex adaptive processes within rural e-commerce ecosystems under digital business empowerment of agriculture and explore pathways for their sustainable development.
Methodology
Research Framework and Hypotheses
Construction of the Research Framework
Complex Adaptive Systems (CAS) theory emphasizes that system components are adaptive agents that actively respond to stimuli from other agents and the environment (Ahmed et al., 2024; Holland, 1992). Through interactions and learning, these agents evolve to form complex macro-level phenomena (Khouja et al., 2008). The core idea is that “adaptability creates complexity” (Holland, 1992). CAS theory is thus suitable for studying the formation and development of rural e-commerce ecosystems, which result from interactions among internal agents. L. Huang et al. (2024) used CAS theory to construct a TOE (Technological-Organizational-Environmental) framework to analyze rural e-commerce entrepreneurial ecosystems, highlighting the need for effective internal interactions and adaptability to external changes. However, they did not deeply explore agent interactions. Building on this, this study uses CAS theory to explain the configurational effects of multi-agent interactions in rural e-commerce ecosystem development under digital rural construction and constructs an analytical framework.
Figure 1 shows the basic “stimulus-response” model of Complex Adaptive Systems (CAS) theory (Holland, 1996; Malina, 2017). In this model, environmental factors generate stimuli detected by agents through sensors. These stimuli are processed and analyzed, prompting agents to respond based on internal rules or experience. These responses then affect the environment, potentially triggering new stimuli and forming a dynamic feedback loop. Through this loop, agents continuously adapt to environmental changes, driving system evolution (Holland, 1996; Malina, 2017).
The basic model of CAS theory: The stimulus-response model.
For example, in the rural e-commerce ecosystem, government support policies (e.g., subsidies and tax incentives) act as stimuli. When perceived by farmers’ cooperatives, these policies prompt adjustments in product structure (e.g., increasing green and organic production), expanding online sales channels, and collaborating with e-commerce platforms. These actions enhance market supply, platform user activity, and brand influence, attracting more producers to adjust their strategies and prompting further government support. This feedback loop drives the high-quality development of rural e-commerce.
Based on this model, this paper references the “County Digital Rural Index System” developed by Peking University’s Digital Rural Project Group. By analyzing its first- and second-level indicators, we construct a research framework to explore the configurational effects of rural e-commerce ecosystem development under multi-dimensional digital rural stimuli, as shown in Figure 2.
Research framework: Multi-agent collaborative configuration effect on the development of the rural E-commerce ecosystem.
Figure 2 shows how, in the context of digital rural development, various entities receive environmental stimuli and respond, collaboratively driving the development of the rural e-commerce ecosystem.
Local Governments
Local governments are stimulated by national policies, such as the “Key Points of Digital Rural Development Work in 2024,” which propose enhancing rural network infrastructure, and promoting smart agriculture (M. Liu & Liu, 2024). Their responses include:
① Rural Digital Infrastructure Construction: Increasing efforts to transform and upgrade rural infrastructure, promoting the coverage of 5G networks and the Internet of Things.
② Digitalization of Rural Governance: Enhancing governance capabilities through digital technology, such as establishing a “Rural Brain” platform for precise operations and intelligent decision-making.
③ Public Services: Promoting the digitalization of rural public services, including online education and telemedicine, to improve rural residents’ quality of life (Leong et al., 2016; Lu et al., 2024; Y. Wang et al., 2024).
New Farmers
The spread of digital technology and improvements in e-commerce platforms and logistics networks stimulate new farmers to digitally transform agricultural production, rural life, and agricultural product sales (Duan et al., 2023; H. Lin et al., 2024; M. Liu & Liu, 2024; D. Zhang et al., 2024). Their responses include:
① Rural Digital Production: Using smart agricultural equipment and big data for precise planting and breeding to boost production efficiency.
② Rural Digital Marketing: Conducting live-streaming sales of agricultural products via short video and e-commerce platforms to expand sales channels.
③ Rural Digital Life: Actively participating in digital rural construction, improving digital literacy, and enjoying the convenience of digital technology.
E-Commerce Platforms
Driven by policy support and the growth of agricultural product e-commerce sales, e-commerce platforms respond through “rural digital marketing” and supply chain optimization (Chu et al., 2023, 2025; Duan et al., 2023; Qu et al., 2018). They provide precise marketing services for farmers and enterprises through big data analysis, helping agricultural products meet market demand. They also collaborate with logistics companies to enhance intelligent logistics and create an efficient supply chain.
Logistics Enterprises
Logistics enterprises are stimulated by increased rural logistics demand and policy encouragement for logistics network construction (L. Huang et al., 2021; D. Wang, 2023; Y. Wang & Gao, 2024; J. Zhou et al., 2021). Their responses include:
① Rural Digital Supply Chain: Strengthening logistics center construction in agricultural product origins, achieving digital management of logistics information, and improving delivery efficiency.
② Collaboration with E-commerce Platforms: Optimizing logistics distribution routes and reducing costs through cooperation.
Financial Institutions
Policies promoting digital inclusive finance and supporting rural revitalization stimulate financial institutions to respond with “rural digital finance” (Y. Wang et al., 2024; Wei et al., 2024). They use big data and artificial intelligence to optimize credit approval processes, reduce service costs, and improve the availability and convenience of financial services, supporting the digital upgrading of the agricultural industry chain.
Research Hypotheses
As illustrated in Figure 2, the process by which the interactions of multiple agents promote the development of the rural e-commerce ecosystem is based on the different responses of each agent to the environmental stimuli of digital rural construction and their mutual interactions, which form the operational foundation of the complex system of the rural e-commerce ecosystem. To deeply analyze the development patterns and internal mechanisms of this complex system, it is far from sufficient to study from the perspective of a single agent or a single element, as mentioned earlier, there are close interconnections, and synergistic effects between agents and elements. Therefore, it is necessary to introduce a configurational perspective, regarding the rural e-commerce ecosystem as an overall configuration composed of multiple interrelated elements, to grasp its development trend and influencing factors as a whole, and then put forward more scientific and reasonable research hypotheses, so as to more comprehensively and accurately reveal the multi-agent collaborative mechanism of the development of the rural e-commerce ecosystem. Based on this, the following research hypotheses based on the configurational perspective and the synergistic effects of specific elements are proposed in this study.
Research Hypotheses Based on the Configurational Perspective
For example, rural digital production (RDP) can enhance the production skills and digital application capabilities of new farmers, promoting agricultural development. However, without the support of rural digital infrastructure (RDI), such as incomplete information infrastructure, agricultural products cannot be effectively sold through e-commerce platforms. Similarly, rural digital governance (RDG) can improve the transparency of policy implementation and government service capabilities, but without the cooperation of rural digital supply chains (RDC), the efficient circulation and timely delivery of agricultural products cannot be guaranteed, and the rural e-commerce ecosystem cannot be formed and developed.
Based on the principle of causal equifinality (El Sawy et al., 2010; Fiss, 2007, 2011), on the one hand, there may be a configuration where rural digital production levels are high, digital infrastructure is complete, digital governance is efficient, digital supply chains are smooth, and digital financial support is strong. Under this configuration, new farmers can use advanced production technologies and complete infrastructure to sell high-quality agricultural products through efficient supply chains and sufficient financial support, thereby promoting high-quality development of the rural e-commerce ecosystem. On the other hand, there may be another configuration where rural digital life is highly active, digital marketing is innovative, digital infrastructure has a certain foundation, and digital governance performs well in government services. In this case, by attracting consumers through active digital life and innovative marketing, and with the help of acceptable infrastructure and governance environment, the development of the rural e-commerce ecosystem can also be promoted, although the development path and focus are different from the previous configuration.
The principle of asymmetry means that the element configuration that promotes high-level development of the rural e-commerce ecosystem is not simply the opposite of the configuration that leads to low-level development (Fiss, 2007, 2011; Greckhamer, 2016; Misangyi et al., 2017). For example, a county with advanced digital infrastructure, efficient digital supply chains, and active digital marketing has a well-developed rural e-commerce ecosystem due to the synergistic effect of these elements. However, this does not mean that backward digital infrastructure, unsmooth supply chains, and weak marketing will definitely lead to poor development of the rural e-commerce ecosystem in that county. There may be other factors, such as strong support policies from the local government and special innovative measures by new farmers, that are at play, making the development situation not completely symmetrical.
Hypotheses Based on the Synergistic Effects of Specific Elements
Rural digital production provides a wealth of digital agricultural products and production data for rural e-commerce, while rural digital infrastructure is the prerequisite for these resources and data to circulate and monetize smoothly. When the degree of synergy between the two is high, new farmers can use the complete infrastructure to efficiently display, promote, and sell the digital agricultural products produced, promoting the prosperity of the rural e-commerce ecosystem.
Rural digital governance can optimize the construction and development environment of rural digital supply chains through policy guidance and resource integration, while the efficient operation of rural digital supply chains can provide data support and feedback for digital governance, helping the government to better adjust governance strategies. When the two interact positively, they can effectively improve the circulation efficiency of agricultural products in the rural e-commerce ecosystem, reduce costs, and enhance competitiveness.
Rural digital finance provides financial security and risk management tools for rural e-commerce marketing activities, supporting marketing innovation and market expansion. Rural digital marketing, on the other hand, creates more application scenarios and business opportunities for digital finance through innovative marketing models and channels. A high degree of integration between the two can provide strong support for rural e-commerce in both funding and market aspects, thereby effectively promoting the outward expansion and scale expansion of the rural e-commerce ecosystem.
On the one hand, the improvement of rural digital life brings a broader consumer market and richer product sources to the rural e-commerce ecosystem, and also provides new marketing channels and innovation models. On the other hand, the development of the rural e-commerce ecosystem can further improve the level of rural digital life services, such as providing more life services and conveniences through e-commerce platforms. The two promote each other, forming a virtuous cycle, jointly promoting the continuous development of the rural e-commerce ecosystem, and the continuous upgrading of rural digital life.
Research Methods and Data Sources
Research Methods and Sample Selection
Research Method: fsQCA (Fuzzy-Set Qualitative Comparative Analysis)
Qualitative Comparative Analysis (QCA) constitutes a methodological framework encompassing several technical variants, including crisp-set QCA (csQCA), multi-value QCA (mvQCA), and fuzzy-set QCA (fsQCA; Ragin, 2010; Rihoux & Ragin, 2009). Among these, fsQCA employs fuzzy-set theory to calibrate partial set membership—assigning values continuously between 0 and 1 rather than binary categories (Fiss, 2007; Greckhamer & Gur, 2021; Ragin, 2010). This approach addresses both qualitative kind (set membership) and quantitative degree (membership strength) simultaneously, thereby capturing nuanced causal relationships (Fiss, 2007).
fsQCA advances beyond traditional case-study methods by systematically analyzing causal complexity: it identifies configurations of antecedent conditions that generate outcomes, examines interactions among causal factors, and elucidates core generative pathways underlying observed phenomena (Fiss, 2007; Tekic & Tekic, 2024).
The specific steps of the fsQCA method are as follows:
① Case Selection and Data Collection: Select research subjects and collect relevant data. The QCA method is applicable to small samples (10 or 15 and below), medium samples (10 or 15–50), and large samples (over 100).
② Variable Selection and Calibration: Determine the conditional and outcome variables in the study and calibrate these variables. Calibration is the process of transforming qualitative data into quantitative data, typically using fuzzy-set calibration methods.
③ Constructing the Truth Table: Based on the calibrated data, construct a truth table that lists all possible combinations of conditions and their corresponding outcomes.
④ Logical Simplification: Use Boolean algebra algorithms to logically simplify the truth table, identifying key condition combinations (configurations) and their impact on outcomes.
⑤ Interpretation and Validation: Interpret the results after logical simplification and validate their applicability and explanatory power in actual cases.
The fsQCA approach is chosen to investigate the complex causal mechanisms behind the development of rural e-commerce ecosystems. fsQCA can accurately capture the complex characteristics of multiple connection causal relationships, causal equifinality, and causal asymmetry contained between variables (Greckhamer et al., 2018), which is highly consistent with the configurational perspective adopted in this study. It provides a powerful analytical tool and methodological support for in-depth analysis of the interrelationships between various elements in the rural e-commerce ecosystem and their comprehensive impact on the development performance of the rural e-commerce ecosystem.
To efficiently and accurately implement fsQCA, this part fully relies on the fsQCA3.0 software for operation, strictly following the standardized process proposed by Ragin (2010) and Verweij (2012). It successively carries out key steps such as measurement calibration, necessity analysis, and sufficiency analysis to ensure the scientificity, rigor, and reliability of the research process and results. Thus, it deeply explores the complex causal logic and key element configuration patterns behind the development of the rural e-commerce ecosystem.
Sample Selection
In practice, the development of China’s rural e-commerce ecosystem mostly appears in the form of Taobao villages, which are mostly distributed in the eastern coastal provinces, and the categories operated by these provinces’ Taobao villages are mostly non-agricultural products. Therefore, from the provincial scope, the development differences of the rural e-commerce ecosystem are not obvious, especially for the construction and development of the “agricultural products going up” rural e-commerce ecosystem, the reference significance is limited.
Therefore, this study focuses on the county level, paying special attention to the “E-commerce Entering Rural Areas Comprehensive Demonstration Counties” with outstanding performance in the field of “agricultural products going up,”“Top 100 Taobao Villages Counties,” and “Top 100 Agricultural Products Digitalization Counties.” After comprehensive comparison of the statistical lists of the three types of counties, it was found that the “Top 100 Agricultural Products Digitalization Counties” have typical “agricultural products going up” characteristics and significant rural e-commerce ecosystem development characteristics. Therefore, this paper selects 100 “Top 100 Agricultural Products Digitalization Counties” as the case sample range.
The sample data comes from the “Digital Business for Rural Revitalization: High-Quality Development of Agricultural Products E-commerce from Alibaba Platform” research report jointly written by the Management Cadre College of the Ministry of Agriculture and Rural Affairs and Alibaba Research Institute in April 2022. The report comprehensively considers the national digital rural pilot areas, national modern agricultural industrial parks, strong agricultural industry towns, geographical indication products, regional public brands, and Alibaba’s agricultural products e-commerce transactions, as well as the digital inclusive finance data of Webank, to form the list of top 100 agricultural products digitalization counties (Note: Considering that the economy of each municipal district is relatively developed, this ranking temporarily does not include municipal districts in the ranking). Because there are nine counties with missing digital rural index data, the final number of county samples left is 91.
Data Sources and Calibration
Data Sources
The outcome variable is the development of the rural e-commerce ecosystem (REED), for which there is currently no unified measurement standard in academia. Since this paper focuses on the development level of the rural e-commerce ecosystem that promotes “agricultural products going up” within the county scope, the development level of the rural e-commerce ecosystem in a county is often reflected by its digital management status of agricultural products. Therefore, this paper directly adopts the ranking situation of the “Top 100 Agricultural Products Digitalization Counties” in the research report “Digital Business for Rural Revitalization: High-Quality Development of Agricultural Products E-commerce from Alibaba Platform” as the original data of the outcome variable, with the order of ranking representing the high and low levels of its rural e-commerce ecosystem development.
The data of the seven condition variables all come from the “County Digital Rural Index” (2020) developed by the Digital Rural Project Group of the Institute of New Rural Development at Peking University, as well as relevant research reports released by Alibaba Research Institute and the Ministry of Agriculture and Rural Affairs. The specific measurement indicators are shown in Table 1. Considering the lag of the results, all condition variables are indicators of the year 2020, while the outcome variable is the data of the year 2022.
Selection and Measurement Indicators of Condition Variables.
Source. Compiled by the authors.
Data Calibration
Drawing on existing research (Rihoux & Ragin, 2009; S. Zhang et al., 2024), and combining the corresponding theoretical knowledge, the direct method is adopted for variable calibration. The 0.9, 0.5, and 0.1 quantiles of the descriptive statistics of each variable are used as the full membership anchor points, crossover points, and full non-membership anchor points, respectively. The specific situation is shown in Table 2. The calibration rule for non-high-level rural e-commerce ecosystem development is the opposite, taking the non-set of high-level rural e-commerce ecosystem development.
Variable Calibration Results.
Source. Created by the authors.
Results
Data Analysis and Results
Necessity Analysis of Single Variables
By evaluating the consistency and coverage of individual variables, it can be determined whether a condition is a necessary condition (L. Huang et al., 2024; Ragin, 2010). Using the fsQCA3.0 software, this study analyzed whether any single factor constitutes a necessary condition for the development of the rural e-commerce ecosystem. If the consistency of a condition variable exceeds 0.9, it is considered necessary, meaning it has the ability to independently explain the outcome variable (Rihoux & Ragin, 2009). The software calculation results are shown in Table 3. As shown in Table 3, the consistency of the prerequisite variables for both high-level and non-high-level development of the rural e-commerce ecosystem is below 0.9. This indicates that no single factor is a necessary condition for the development of the rural e-commerce ecosystem. In fsQCA, conditions usually do not operate alone but interact with other conditions to influence the outcome. Therefore, it is necessary to consider the interactions between these conditions and other conditions, and analyze how different condition combinations affect the outcome through configurational analysis. The high-level development of the rural e-commerce ecosystem is comprehensively influenced by multiple dimensions of digital rural construction. Therefore, subsequent research is needed to explore the configurational effects of causal variables.
Analysis of Necessary Conditions.
Source. fsQCA 3.0 software output.
Note. The symbol “∼” indicates “non-high level.” The same applies below.
Configurational Analysis
A configuration represents the arrangement and combination of different conditions that produce similar outcomes. Based on existing research practices and considering the number of cases in this study, the original consistency threshold is set at 0.8 (Fiss, 2007; Ordanini et al., 2014), the PRI (proportional reduction in inconsistency) at 0.75 (Greckhamer et al., 2018), and the case frequency at 1. The software calculates and outputs three types of solutions: complex solutions, parsimonious solutions, and intermediate solutions. Following the common practice of QCA analysis, this paper mainly reports the intermediate solutions and refers to the parsimonious solutions to determine the core and peripheral conditions in the configurations. Core conditions are those that appear in both the intermediate and parsimonious solutions; peripheral conditions are those that only appear in the intermediate solutions. Through analysis using the fsQCA software, the configurational paths for the development of high-level rural e-commerce ecosystems were determined, as shown in Table 4.
Configurational Paths for High and Non-High-Level Development of Rural E-commerce Ecosystems (Frequency = 1).
Source. fsQCA 3.0 software output
Note. A large ● indicates the presence or high level of a core condition (a condition that appears in both the intermediate and parsimonious solutions), while a small • indicates the presence or high level of a peripheral condition (a condition that only appears in the intermediate solution); a large ⊗ indicates the absence or low level of a core condition, and a small ⨂ indicates the absence or low level of a peripheral condition; – indicates that the presence or absence or the level of the condition does not affect the outcome.
As shown in Table 4, the software effectively identified 12 paths, of which 11 are configurational paths for the development of high-level rural e-commerce ecosystems and one is for the development of a low-level rural e-commerce ecosystem. Considering the similarity of core variables, the 11 high-level configurational paths are further categorized into eight configurations. The county cases corresponding to each configurational path are shown in Table 5.
Sample Distribution of High and Non-High-Level Development Configurational Paths.
Source. fsQCA 3.0 software output
Note.* indicates the logical “AND,”∼ indicates the logical “NOT.”
The results of the configurational analysis indicate that there are 11 configurational paths for the development of high-level rural e-commerce ecosystems, with the overall solution consistency index above 0.85 and the overall solution coverage at 0.540. This means that these 11 configurational paths can explain 54% of the reasons for the high-level development of rural e-commerce ecosystems. The consistency of each configurational path condition exceeds 0.9, indicating that these conditions have strong explanatory power. The results show that the empirical analysis is valid.
Different configurations mark various pathways that drive or hinder the high-level development of rural e-commerce ecosystems, with each conditional variable being regarded as an element in each configuration. Looking at the distribution of elements across all configurations, “rural digital marketing (RDM)” and “rural digital production (RDP)” are the most critical driving factors for the high-level development of rural e-commerce ecosystems, followed by “rural digital supply chain (RDC).” Meanwhile, “rural digital infrastructure” and “rural digital finance” are the essential foundations for the high-level development of rural e-commerce ecosystems. Among them, the original coverage rates of the S3, S4a, and S4b configurational paths are relatively high, at 0.250, 0.241, and 0.270, respectively. This indicates that with the basic support of “rural digital infrastructure” and “rural digital finance,” the high-level development of rural e-commerce ecosystems relies on the dual driving forces of rural digital marketing and rural digital supply chain. In addition, rural digital production and rural digital life are also important factors for the high-level development of rural e-commerce ecosystems. It should be noted that the original coverage rates of the 11 paths for the high-level development of rural e-commerce ecosystems all exceed the single coverage rate, indicating the existence of multiple compound causal pathways.
From the fsQCA results, it can be seen that no single element can independently support the high-quality development of the rural e-commerce ecosystem. For example, in paths S1a and S1b, both include low-level rural digital infrastructure (∼RDI) and rural digital finance (∼RDF), indicating that a single element such as digital governance (RDG) or digital marketing (RDM) cannot independently promote the development of the rural e-commerce ecosystem. This is consistent with Hypothesis
The fsQCA results show 12 element combination paths, each representing different development configurations at different levels. This indicates that different element combinations can indeed lead to different levels of development, supporting Hypothesis
Paths S4a and S4b show configurations for high-level development, while path N shows a configuration for low-level development. In these paths, the element combinations are not entirely symmetrical. For example, S4a includes RDI, RDF, and RDM, while N also includes RDI and RDF but lacks RDM. This indicates that the configurations for high and low levels of development are not simply opposite situations, supporting Hypothesis
Except for path S1a, RDP and RDI appear simultaneously in the remaining paths, and these paths correspond to a higher level of development in the rural e-commerce ecosystem. This indicates that the degree of synergy between rural digital production (RDP) and rural digital infrastructure (RDI) positively affects the development of the rural e-commerce ecosystem, supporting Hypothesis
Apart from paths S2 and S4b, the other high-level development paths all include both RDG and RDC. This demonstrates that the interaction between rural digital governance (RDG) and rural digital supply chain (RDC) has a significant impact on the operational efficiency of the rural e-commerce ecosystem, thereby supporting Hypothesis
With the exception of path S3, RDF and RDM are concurrently present in all other high-level development paths. This indicates that the integration level between rural digital finance (RDF) and rural digital marketing (RDM) is crucial for the expansion capability of the rural e-commerce ecosystem, thus upholding Hypothesis
Excluding S4a, RDL is found in the remaining 10 high-level development configurations, with six of these corresponding to a high level of RDL. This suggests that RDL plays a promotional role in the development of the rural e-commerce ecosystem and that there is a certain causal feedback loop, thereby supporting Hypothesis
The element configuration of the non-high-level path N reveals that even with high levels of “rural digital infrastructure (RDI)” and “rural digital finance (RDF),” if the other five conditional variables are underdeveloped, high-level development of the rural e-commerce ecosystem cannot be achieved. This inversely highlights the foundational role of “rural digital infrastructure (RDI)” and “rural digital finance (RDF),” as well as the critical importance of “rural digital governance (RDG),”“rural digital life (RDL),”“rural digital production (RDP),”“rural digital marketing (RDMY),” and “rural digital supply chain (RDC).”
The significance of elements is determined by their critical nature. Analysis of the distribution of configurational elements reveals that rural digital infrastructure and rural digital finance are foundational. This highlights the pivotal roles of the government in building rural digital infrastructure and financial institutions in supporting rural digital finance. However, these factors are not the key drivers of performance differences in rural e-commerce ecosystem development. Instead, the critical factors influencing development levels across counties are “rural digital marketing (RDM),”“rural digital supply chain (RDC),”“rural digital production (RDP),”“rural digital life (RDL),” and “rural digital governance (RDG).” This underscores the importance of collaborative patterns among new farmers, e-commerce platforms, logistics companies, and the government.
In summary, the fsQCA analysis confirms that no single element can independently drive rural e-commerce ecosystem development; multiple elements must combine. Different configurations lead to varying development levels, with high and low levels exhibiting unique and asymmetric configurations. Synergistic effects of specific elements, such as RDP with RDI, RDG with RDC, RDF with RDM, and RDL, significantly impact development. Among the seven conditional variables, RDI and RDF are foundational, while RDM, RDC, RDP, RDL, and RDG are key factors. New farmers, e-commerce platforms, logistics companies, and the government are key entities, and their collaborative patterns determine the development level of a county’s rural e-commerce ecosystem. These findings provide robust empirical support for understanding the multi-agent collaborative mechanism in rural e-commerce ecosystems.
Robustness Testing
Currently, there are three main methods for robustness testing in fsQCA (Du et al., 2021; Fiss, 2011; Schneider, C.Q., Wagemann, C., 2012): ①changing the case frequency; ②adjusting the calibration thresholds; ③varying the consistency thresholds. This study employed methods ① and ③ for robustness testing.
The first method, in line with the majority of literature, involves adjusting the consistency threshold for testing. The consistency threshold was increased from 0.8 to 0.85, and the software calculations yielded results consistent with the original findings, demonstrating the strong robustness of the results.
The second method, aimed at further enhancing the generalizability of the study results, involves changing the case frequency. The case frequency was adjusted from the original 1 to 2, while maintaining the original consistency threshold at 0.8 and the PRI consistency at 0.75, and the configurational calculations were performed again. The software outputted four valid configurational paths, three of which were high-level configurations and one was a non-high-level configuration, with the non-high-level configuration being consistent with the original. The corresponding configuration situations and sample distributions are shown in Table 6. Comparing Table 4 with Table 6, it can be seen that the configurations of M1, M2, and M3 are essentially the same as the original S4b, S6, and S2 configurations, respectively. That is, the results after adjusting the case frequency are a subset of the original results, which once again indicates the robustness and generalizability of the results.
Configurational Paths for High and Non-High-Level Development of Rural E-commerce Ecosystems (Frequency = 2).
Source. fsQCA 3.0 software output
As shown in the configurational analysis results in Table 6, after adjusting the case frequency to 2, there are three configurational paths for the development of high-level rural e-commerce ecosystems, with the overall solution consistency index at 0.861, above 0.85, and the overall solution coverage at 0.396. This means that these three configurational paths can explain 39.6% of the reasons for the high-level development of rural e-commerce ecosystems. The consistency of each configurational path condition exceeds 0.85, indicating that the three configurational paths have strong explanatory power.
Looking at the sample distribution, the L1 configuration explains the most samples, with an original coverage of 0.270 and a unique coverage of 0.146, indicating that the L1 configuration can explain 27% of the sample cases, and 14.6% of the sample cases can be uniquely explained by L1. This reveals that the L1 configuration is the main path for the development of high-level rural e-commerce ecosystems, and its multi-agent collaborative model has important reference value.
Regarding the distribution of elements, the results before and after adjusting the case frequency are basically consistent. That is, “rural digital infrastructure (RDI)” and “rural digital finance (RDF)” remain foundational elements, while “rural digital production (RDP),”“rural digital marketing (RDM),” and “rural digital supply chain (RDC)” are key core elements.
From the perspective of the combination of key entities, both analyses before and after adjusting the frequency show the characteristics of multi-agent collaboration, that is, a single entity cannot drive the development of the rural e-commerce ecosystem alone. From the QCA analysis results, new farmers, e-commerce platforms, and logistics companies have played a leading role in the development of the rural e-commerce ecosystem. These entities promote the upward movement of agricultural products and the high-quality development of the rural e-commerce ecosystem through collaborative cooperation. New farmers appear in almost every configurational path for high-level development, indicating their key agency in the rural e-commerce ecosystem. This result also expresses from another side that the digital transformation of traditional farmers is a prerequisite for the development of the rural e-commerce ecosystem.
Identification and Case Analysis of Multi-Agent Collaborative Patterns
Combining the robustness test configurational analysis results after adjusting the case frequency in the previous fsQCA analysis, and taking into account the distribution of samples (Tables 5 and 6) and the generalizability of the research results, the three universal paths in Table 6 are summarized into three patterns: “Quadruple-Ring Synergistic Drive Type,”“Production-Marketing Dual-Core Drive Type,” and “Core Triple-Ring Drive Type” using the conditional identification method and analyzing the actual case situations of key participating entities and county development.
M1: “Quadruple-Ring Synergistic Drive Type” (Typical Case: Wuyishan City)
In the M1 configuration, “rural digital production (RDP),”“rural digital supply chain (RDC),”“rural digital marketing (RDM),” and “rural digital finance (RDF)” are core conditions at high levels. “Rural digital infrastructure (RDI)” and “rural digital life (RDL)” are auxiliary conditions at high levels, while “rural digital governance (RDG)” can be present or absent. This indicates that in high-level rural e-commerce ecosystems, digital production, supply chain optimization, marketing innovation, and financial support are key drivers. Infrastructure and digital life, though important, are not decisive. This configuration highlights the synergistic effect of production, supply chain, marketing, and finance, driving rural e-commerce development. Thus, it is named the “Quadruple-Ring Synergistic Drive Type.” Counties like Jiashan, Wuyishan, and Xianyou exemplify this model, with Wuyishan City being the most representative. Below is a case analysis of Wuyishan City.
Case Analysis of M1 Model: Wuyishan City
Wuyishan City, under Nanping City in Fujian Province, has achieved significant success in rural e-commerce. As of 2023, it hosts over 3,800 e-commerce enterprises, 11,000 online stores, and 15,000 employees. From January to May 2023, online retail sales reached 1.801 billion yuan, with a 15.3% year-on-year increase, including 1.27 billion yuan from agricultural products, up 11%. These figures indicate a strong development foundation and market competitiveness.
Rural Digital Production (RDP): Wuyishan City integrates digital technology into tea planting through the National Digital Planting Industry Innovation Application Base Construction Project. This includes intelligent management of production, processing, quality tracing, and digital testing, significantly improving product quality and efficiency. This demonstrates the core role of digital production in enhancing agricultural competitiveness.
Rural Digital Supply Chain (RDC): Wuyishan City has established a comprehensive rural logistics network covering county, township, and village levels, creating a fast logistics channel for industrial and agricultural products. By integrating passenger and cargo postal lines, logistics efficiency is further optimized, reducing costs, and improving circulation. This highlights the key role of the digital supply chain in supporting rural e-commerce.
Rural Digital Marketing (RDM): With over 3,700 e-commerce enterprises, half of which are tea-related, Wuyishan City has embraced new business models like “special agricultural products + live broadcast sales” through platforms like Kuaishou and Douyin. Initiatives such as the “Scan to Buy Wuyishan” mini-program and rural revitalization live broadcast bases have created several successful agricultural e-commerce brands, significantly boosting market competitiveness and sales volume.
Rural Digital Finance (RDF): Wuyishan City has implemented financial reward policies, providing over 3.3 million yuan in direct support to more than 160 enterprises. Training programs like “E-commerce New Farmers” and “Rural Broadcasters” offer “one-stop” entrepreneurial guidance, demonstrating the key role of digital finance in supporting rural e-commerce development.
Rural Digital Infrastructure (RDI): Achieving “broadband access to every village” since 2009, Wuyishan City has further improved rural network coverage and data processing capabilities through 5G smart ecological tea gardens. This provides essential support for the efficient operation of the rural e-commerce ecosystem.
Rural Digital Life (RDL): By building rural revitalization live broadcast bases and e-commerce talent training bases, Wuyishan City enhances the digital life experience of rural residents. This not only expands the consumer market for rural e-commerce but also provides new marketing channels and innovation models.
In summary, Wuyishan City’s practice verifies the effectiveness of the “Quadruple-Ring Synergistic Drive Type.” Through the synergistic effect of digital production, supply chain optimization, marketing innovation, and financial support, Wuyishan City has successfully promoted high-quality development of the rural e-commerce ecosystem. Collaboration among new farmers, e-commerce platforms, logistics enterprises, governments, and financial institutions has improved operational efficiency and enhanced overall competitiveness, providing strong support for rural revitalization and agricultural modernization.
M2: “Production-Marketing Dual-Core Drive Type" (Typical Case: Qixia City)
In the M2 configuration, “rural digital production (RDP)” and “rural digital marketing (RDM)” are core conditions, while “rural digital finance (RDF)” plays a supplementary role. This indicates that new farmers and e-commerce platforms are the key drivers, emphasizing that digital production and marketing are central to rural e-commerce ecosystem development. This configuration is thus named the “Production-Marketing Dual-Core Drive Type.”
The M2 path shows that even without high-level “digital supply chain (RDC)” and “digital infrastructure (RDI),” a high-level rural e-commerce ecosystem can be established. New farmers, stimulated by technological innovation and policy orientation, actively engage in digital production, and marketing. E-commerce platforms enhance their market competitiveness and expand sales channels, promoting high-level ecosystem development despite weak digital infrastructure.
The model’s consistency is 0.900, with an original coverage of 0.189 and unique coverage of 0.069. This path explains 18.9% of high-level cases, fully accounting for 6.9% of them. County cases include Gutian County, Qixia City, and Menghai County, with Qixia City being the most representative. Located in Shandong Province, Qixia City is known for its apples and has promoted rural e-commerce through a “digital production and sales model.”
Case Analysis of M2 Model: Qixia City
Rural Digital Production (RDP): Qixia City has focused on apples, promoting smart orchard construction. IoT sensors and a Smart Orchard Planting Cloud Platform enable real-time monitoring of tree growth, supporting scientific planting and management. The city has built or is building 8,000 acres of digital orchards, improving production efficiency and product quality for e-commerce sales.
Rural Digital Marketing (RDM): Qixia City has innovated sales models by creating the “Fruit Capital Selection” WeChat Mall, integrating products from 42 villages and leveraging platforms like Alibaba, JD.com, and Pinduoduo. In 2023, the city had over 4,000 online shops, with online retail sales reaching 1.263 billion yuan, up 42% year-on-year. The city also cultivated local e-commerce talents through the “Internet celebrity farmers, Internet celebrity products” project, promoting live broadcast sales.
Rural Digital Finance (RDF): Qixia City has provided financial support through policy guidance and innovation. For example, Qixia Rural Commercial Bank launched “apple” loans tailored to the apple industry chain, solving financing difficulties for purchasers. This service reduces credit risks and guarantees agricultural storage and sales.
In summary, Qixia City’s “Production-Marketing Dual-Core Drive Type” successfully established a high-level rural e-commerce ecosystem by strengthening digital production and marketing, supported by digital finance and external stimuli. Despite lacking high-level digital supply chains and infrastructure, Qixia City promoted ecosystem development through active digital production and e-commerce sales by new farmers. This model increased farmers’ income, promoted rural economic sustainability, and provides a replicable experience for other counties.
M3: “Core Triple-Ring Drive Type” (Typical Case: Shuyang County)
In the M3 configuration, “rural digital production (RDP),”“rural digital supply chain (RDC),” and “rural digital marketing (RDM)” are core conditions, while “rural digital life (RDL)” is an auxiliary condition. Digital infrastructure (RDI) and digital finance (RDF) are missing core conditions, and digital governance (RDG) is a missing auxiliary condition. This indicates that digital production, supply chain, and marketing are key drivers, forming the core support of the rural e-commerce ecosystem. Thus, this path is named the “Core Triad Drive Type.” The presence of “rural digital life (RDL)” as an auxiliary condition suggests that enhancing digital literacy and promoting digital technology in rural areas can provide a social foundation for e-commerce development. This model includes Donghai County and Shuyang County, with Shuyang County being the focus of the case analysis.
Case Analysis of M3 Model: Shuyang County
Shuyang County, known for its floriculture e-commerce, has cultivated new rural industries through the “floriculture + e-commerce + live streaming” model, building a high-quality e-commerce ecosystem. With floriculture as the core industry, Shuyang County is China’s largest floriculture production and distribution center, employing over 300,000 people, with live streaming sales accounting for one-third of the national total. The county has explored this model through platforms like Douyin, Taobao, and JD.com, providing personalized guidance services and incubating cross-border e-commerce enterprises. It has also improved practitioners’ digital literacy and professional skills through training and talent cultivation, optimized the e-commerce industry chain layout, and created a good e-commerce environment through collaborative supervision and crackdowns on counterfeit goods.
(1) Rural Digital Production (RDP): Shuyang County has promoted the digital transformation of floriculture through smart agriculture technology. The “Digital Yanxia” project in Yanji Town uses video surveillance and sensor monitoring for precise control of temperature, humidity, and fertilization, increasing efficiency by 15% and benefits by 8%. The county has also built 28 agricultural IoT technology application demonstration bases, with IoT technology covering 25% of large-scale facility agriculture.
(2) Rural Digital Supply Chain (RDC): Shuyang County has established a three-level mail delivery logistics system and the Suqian Postal Mail Processing Center with a daily throughput capacity of 500,000 pieces. It has built express logistics parks in multiple townships, with 22 express logistics enterprises stationed. The county also has 16 Taobao towns and 104 Taobao villages, with an annual shipment volume of about 60 million pieces of floriculture e-commerce, effectively solving the “last mile” problem of agricultural product transportation.
(3) Rural Digital Marketing (RDM): Shuyang County has promoted floriculture products to the national market through e-commerce platforms and live streaming sales. As of the end of 2023, the county had over 40,000 floriculture e-commerce businesses, with online sales ranking first in the country for several consecutive years. Yanxia Village adjusted product strategies based on market trend data, achieving an annual sales volume of 515 million yuan. The county has also enhanced brand influence through floriculture festivals and e-commerce live streaming bases.
(4) Rural Digital Life (RDL): Shuyang County has provided agricultural science and technology information services to farmers through platforms like “Benefit Farmers Information Society” and “Agricultural Technology Cloud” APP, enhancing digital literacy. 12,000 new agricultural business entities and farmers share “accurate, timely, and all-weather” agricultural science and technology services through these platforms, providing a good social foundation for e-commerce development.
(5) Policy Support and Alternative Mechanisms: Despite the relative insufficiency of digital infrastructure and digital finance, Shuyang County has compensated through policy support and platform cooperation. The county government has introduced measures to promote the e-commerce industry, including building e-commerce industrial parks, providing entrepreneurship subsidies, and small loans. It has also optimized the business environment to attract e-commerce enterprises, forming a complete e-commerce ecosystem.
In summary, Shuyang County has achieved high-level development of the rural e-commerce ecosystem through the “Core Triad Drive Type” model, driven by digital production, supply chain, and marketing. Despite the insufficiency of digital infrastructure, digital finance, and digital governance, the county has achieved significant results through policy support and platform cooperation, becoming a model for rural e-commerce development in China.
Discussion
The fuzzy-set Qualitative Comparative Analysis (fsQCA) conducted in this study provides valuable insights into the development of rural e-commerce ecosystems through the lens of Complex Adaptive Systems (CAS) theory. The findings underscore the importance of a configurational perspective in understanding the multifaceted and interdependent factors that drive the development of rural e-commerce ecosystems. This section discusses the implications of the study’s findings, their alignment with the proposed hypotheses, and the broader theoretical and practical contributions.
Validation of Hypotheses
The fsQCA results strongly support the proposed hypotheses, emphasizing the complex interplay of digital rural elements in shaping rural e-commerce ecosystems.
Configurational Pathways and Their Implications
The identified configurational pathways offer nuanced insights into the diverse mechanisms through which rural e-commerce ecosystems can develop. For instance, the “Four-Ring Synergistic Drive” model (M1) emphasizes the critical role of digital production, supply chain optimization, marketing innovation, and financial support in driving high-level development. This pathway underscores the need for a comprehensive approach that integrates multiple digital elements to achieve robust ecosystem development.
In contrast, the “Production-Marketing Dual-Core Drive” model (M2) highlights the potential for high-level development even in the absence of advanced digital infrastructure and supply chains. This pathway suggests that strong digital production and marketing capabilities, supported by financial services, can compensate for deficiencies in other areas. This finding has significant practical implications for regions with limited resources, indicating that targeted investments in production and marketing can still yield positive outcomes.
The “Core Three-Ring Drive” model (M3) further illustrates the flexibility in achieving high-level development through a focus on digital production, supply chains, and marketing, even when digital infrastructure and finance are less developed. This pathway highlights the adaptability of rural e-commerce ecosystems and suggests that alternative strategies can be employed to overcome specific deficiencies.
The Role of Key Actors and Their Synergistic Interactions
The study’s findings also highlight the importance of multi-stakeholder collaboration in driving rural e-commerce ecosystem development. New farmers, e-commerce platforms, logistics companies, governments, and financial institutions all play crucial roles in shaping the development trajectory. For example, in the case of Wuyishan City, the collaborative efforts of these stakeholders led to significant advancements in digital production, supply chain optimization, marketing innovation, and financial support. This case demonstrates how synergistic interactions among key actors can catalyze high-level development.
The analysis of different configurational pathways further underscores the need for tailored strategies that leverage the strengths of each stakeholder group. For instance, in the “Production-Marketing Dual-Core Drive” model, new farmers and e-commerce platforms play dominant roles, while in the “Core Three-Ring Drive” model, logistics companies and local governments contribute significantly. This finding suggests that effective ecosystem development requires a nuanced understanding of stakeholder capabilities and the strategic alignment of their efforts.
Theoretical and Practical Contributions
Theoretical Contributions
This study significantly advances the understanding of rural e-commerce ecosystems by integrating Complex Adaptive Systems (CAS) theory with configurational analysis through fuzzy-set Qualitative Comparative Analysis (fsQCA). Unlike traditional linear analyses that focus on individual factors (S. Bai & Jia, 2023; W. Wu et al., 2020), this study highlights the importance of a holistic and context-specific approach, emphasizing the interdependencies and synergies among multiple digital elements. The findings challenge the traditional linear perspectives by identifying multiple configurational pathways that lead to high-level development of rural e-commerce ecosystems. This approach aligns with the broader literature on complex adaptive systems, demonstrating how configurational analysis can be applied to study the development of socio-economic systems in rural contexts (Fan, 2023; Khouja et al., 2008; Preiser et al., 2018).
The study also contributes to the literature on rural e-commerce ecosystems by providing a robust framework for analyzing the interplay of rural digital production (RDP), rural digital infrastructure (RDI), rural digital governance (RDG), rural digital supply chains (RDC), rural digital finance (RDF), rural digital marketing (RDM), and rural digital life (RDL) within the overall e-commerce ecosystem. This framework underscores the importance of a comprehensive and synergistic approach to ecosystem development, which is consistent with the findings of previous studies that highlight the importance of multi-stakeholder collaboration and the adaptability of development strategies (Leong et al., 2016; Paddeu et al., 2018; W. Wang & Yan, 2025).
Practical Contributions
Practically, this study offers actionable insights for policymakers, practitioners, and stakeholders involved in rural e-commerce development. The identification of multiple configurational pathways provides a flexible framework for designing targeted interventions that can be tailored to the specific conditions and strengths of different regions. For example, regions with limited digital infrastructure can focus on strengthening digital production and marketing capabilities, while those with advanced infrastructure can leverage synergies across multiple elements to achieve higher levels of development. This finding is particularly relevant given the current emphasis on digital transformation and the need for context-specific strategies to support rural economic growth (Qu et al., 2018; W. Wang & Yan, 2025; Wei et al., 2024).
The study also highlights the importance of multi-stakeholder collaboration, suggesting that coordinated efforts among new farmers, e-commerce platforms, logistics companies, governments, and financial institutions are essential for catalyzing high-level development. This aligns with the broader literature on sustainable development partnerships, which emphasize the need for integrating the knowledge and resources of multiple stakeholders to address complex socio-economic challenges (Bhattacharya & Fayezi, 2021; Cao & Liang, 2025; Zang et al., 2023). The findings provide practical guidance for policymakers and practitioners to foster collaborative efforts that leverage the strengths of each stakeholder group, thereby enhancing the overall effectiveness of rural e-commerce ecosystem development.
Conclusion
The development of rural e-commerce ecosystems is a complex and multifaceted process influenced by a variety of digital rural elements and the interactions among key stakeholders. This study, grounded in Complex Adaptive Systems (CAS) theory and employing fuzzy-set Qualitative Comparative Analysis (fsQCA), has provided a comprehensive understanding of the mechanisms underlying rural e-commerce ecosystem development. The findings highlight the importance of a configurational approach in analyzing the interdependencies and synergies among multiple factors, challenging the traditional linear and reductionist perspectives.
Summary of Key Findings
This research has demonstrated that no single element, such as digital infrastructure or digital governance, can independently drive the development of rural e-commerce ecosystems. Instead, the synergistic interactions among multiple elements, including digital production, supply chains, marketing, and financial support, are crucial for achieving high-level development. The identified configurational pathways, such as the “Four-Ring Synergistic Drive,”“Production-Marketing Dual-Core Drive,” and “Core Three-Ring Drive” models, illustrate the diverse mechanisms through which rural e-commerce ecosystems can thrive. These pathways emphasize the flexibility and adaptability of development strategies, suggesting that regions can leverage their strengths to compensate for deficiencies in other areas.
The study also underscores the critical role of multi-stakeholder collaboration, involving new farmers, e-commerce platforms, logistics companies, governments, and financial institutions. The synergistic efforts of these stakeholders are essential in catalyzing high-level development, as demonstrated in the case studies of Wuyishan City, Qixia City, and Shuyang County. The findings further reveal the unique and asymmetric nature of the configurations associated with high and low levels of development, highlighting the complexity and non-linearity of rural e-commerce ecosystem development.
Limitations and Future Research Directions
While this study provides valuable insights into the development of rural e-commerce ecosystems, several avenues for future research are suggested. First, the dynamic nature of rural e-commerce ecosystems calls for longitudinal studies that examine their evolution over time. Future research could incorporate time-series data to capture the dynamic interactions among digital elements and stakeholders. Second, the study’s focus on digital elements could be expanded to include socio-cultural and environmental factors, which may also play significant roles in shaping rural e-commerce ecosystem development. Third, comparative studies across different regions and countries could provide a broader understanding of the contextual factors influencing the development of rural e-commerce ecosystems. Finally, future research could explore the potential of emerging technologies, such as blockchain and artificial intelligence, in enhancing the resilience and competitiveness of rural e-commerce ecosystems.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is supported by the National Social Science Foundation of China (23XGL019).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.