Abstract
Over the past two decades, rural areas within Chinese coastal megacities have been influenced by the rapid development of e-commerce platforms that capitalize on the social and physical advantages of local rural villages, turning them into clusters of entrepreneurship. These once-rural ensembles have embraced the opportunities offered by digital platforms, shifting their mode of production from agriculture to manufacturing and delivering their products directly to final customers in other parts of the world. Despite the magnitude of a phenomenon currently reshaping vast regions of China, the physical transformation of the rural landscape catalysed by platformization remains underinvestigated. Grounded in the notion of platform ruralism, this interpretive study focuses on the first generation of TaoBao Villages (TBVs) in the Greater Bay Area of China (GBA) and selects three for in-depth investigation, employing a mixed-methods approach. Morphological analysis is coupled with in-depth semi-structured interviews and field observations to provide a nuanced understanding of the ongoing multifaceted spatial materialization of platform capitalism. Findings show that, with the penetration of e-commerce, the rural landscape has undergone overwhelming spatial restructuring at both the village and building scales. The expansion of the industrial landscape, along with an expanded road network, was superimposed without regard for existing topography, historical context, morphological continuities and the inhabitants’ centuries-old agricultural traditions. By unveiling such rural landscape transformations, this article aims to contribute to the current critical debate on the power of platform capitalism to reshape rural areas within rising megalopoles, such as the GBA.
Introduction
E-commerce platforms have been established as indispensable digital infrastructures that facilitate the gig economy, exerting multifaceted influences on the real world (Lin, 2019; Repenning and Hardaker, 2024). Their penetration into the economic landscape produces a novel spatial fix in the age of platform capitalism (Harvey, 1981, 2001; Montalban et al., 2019; Talamini et al., 2022), mobilizing production space with powerful and effective algorithmic control to meet the needs of capital expansion and cope with its inherent crisis tendency. In the last decades, such a spatial fix has produced profound impacts in both urban and rural realms worldwide (Caprotti and Liu, 2020; Graham, 2020; Leszczynski, 2020; Talamini et al., 2022).
In China, the TBV, a platform-dependent industrial ensemble, is rapidly emerging within former rural economic landscapes as a new mode of production. This e-commerce village, which specializes in a single manufacturing production and sells and delivers via an e-commerce platform, is named after China’s largest e-commerce transaction platform, TaoBao (also known as Alibaba). To qualify as a TBV, a rural ensemble must meet two conditions: within the village boundaries, 1) the annual e-commerce transaction revenue must be at least 10 million yuan, and 2) the number of active online shops must be at least 100, or the active online shops must make up 10% or more of the local households (AliResearch, 2016). Since Alibaba Group started officially listing TBVs, their numbers have skyrocketed: from 20 in 2013 to 7,780 in 2022 (AliResearch, 2020, 2022).
In response to the digitalization trend, China launched the ‘Digital Village Strategy’ in 2018, positioning e-commerce as a strategic instrument for rural development (CCCPC and SCPRC, 2018). Subsequently, a series of policies were introduced: the Development Plan of Digital Agriculture and Village in 2019, the Guidelines for Developing Digital Villages in 2021, and the Action Plan for Digital Village Development in 2022 (MOARA and OCCAC, 2020; OCCAC et al., 2021, 2022). China is not unique in embracing the construction of digital villages; recently, digital-based initiatives were reported in regions and nations across the globe, such as the Smart Villages in Europe (European Commission, 2017), the Digital Villages Initiative in Africa (FAO, 2023) and the Digital India Flagship Initiative in India (National Portal of India, 2021). The application and development of these policies will benefit from a nuanced understanding of the entanglement between rural e-commerce development, landscape transformation and inhabitants’ livelihoods.
The impact of the spatial fix of platformization in rural areas has recently been under scrutiny (Wang et al., 2022). Scholarly works focused on local economic development (Li and Song, 2023; Wang, 2025; Wu et al., 2020; Zhang et al., 2024), household income (Chao and Biao, 2021; Li, 2017; G. Li and Qin, 2022; Li et al., 2024; Liu and Zhou, 2023; Lin et al., 2016; Tang et al., 2022; Zhang et al., 2024), population migration (Y. Wang et al., 2025; Wei et al., 2025) and gender equality (Liu, 2022; Yu and Cui, 2019). Although some studies have explored the phenomenon from a spatial perspective (Li and Song, 2023; Lin, 2019; Wang et al., 2021; Zang et al., 2023), a systematic scrutiny of the platform-mediated landscape transformation remains necessary. Thus this research aims to fill a critical knowledge gap by answering the following research questions: How has the rural landscape evolved with the rise of platform capitalism? Specifically, what are the morphological changes of TBVs since the establishment of the e-commerce platform TaoBao?
This article proceeds as follows. The next section reviews the recent critical debate on the spatial fix of platform capitalism and existing research on its impact in rural areas. The ‘Methods and materials’ section introduces the research design and the study area, along with data collection, interpretation and analysis. The ‘Findings’ section reports the results from morphological analysis and supplemental information from semi-structured interviews and field observations. Finally, the ‘Discussion’ and ‘Conclusion’ sections discuss the results, novelty, limitations and implications for policymakers, planners and designers.
Literature review and theoretical framework
Spatial fix and platform capitalism
The spatial fix conceptualizes the ‘insatiable drive’ of capitalism to alleviate its inherent crisis by spatially releasing over-accumulated capital into external markets (Harvey, 1981, 2001). It sheds light on the process by which capital arranges or fixes space, particularly through geographic expansion and restructuring to ‘soak up’ over-accumulated capital surpluses (Barratt, 2024; Schoenberger, 2004). The term ‘fix’ conveys double meanings, as Harvey (2001: 24) further clarifies: the first meaning refers to something ‘attaching’ or being ‘secured in a particular locus’ or absorbing capital into large infrastructure investment. Another meaning refers to ‘fix[ing] a problem’ or addressing the crisis embedded in capital accumulation. Notably, the fix provides only temporary relief rather than a permanent solution to the over-accumulation of capital. In recent decades, spatial fix has progressively become a mature theoretical framework for explaining phenomena such as global capital flows (Apostolopoulou, 2021), capital-labour relations (Carswell et al., 2026; Czirfusz, 2021), unequal regional development (Chen and Jensen, 2025) and urban sprawl (De Jong et al., 2024; Miao and Phelps, 2022).
Stepping into the digital era, platformization is seen as the leading edge of emerging business models and increasingly acknowledged as a pivotal feature of contemporary capitalism (Rahman and Thelen, 2019; Yeşilbağ, 2022). The platform, as a new type of firm, takes full advantage of extracting and leveraging data that conventional capitalism was not well designed to exploit (Srnicek, 2017: 28). It positions itself as the intermediary that enables different groups of users to interact and is therefore considered flatter and more participatory than conventional forms of capitalism (Liang et al., 2022; Srnicek, 2017: 30). In contrast to conventional firms, the platform is more efficient at pursuing capital accumulation due to reduced direct investment and its capacity to mobilize the ‘crowd’ of producers or consumers (Montalban et al., 2019). It achieves the ‘best of both worlds’ by exercising more fine-grained control over its outsourced production and manufacturing systems while escaping regulatory oversight (Rahman and Thelen, 2019).
Nevertheless, platform capitalism is not immune to the inherent crisis tendency of conventional capitalism and, similarly, attempts to resolve it via the spatial fix. Recent research interprets the profound impact of platform capitalism, focusing on the connotation of fixing. Kuang et al. (2024) mobilized the term ‘spatial-digital fix’, conceptualizing how platform capital puts a sophisticated algorithm system as the core to restore equilibrium between global market trends in digital space and product supply in physical space. Similarly, Repenning and Hardaker (2024) proposed the concept of ‘platform fix’, highlighting the role of digital platforms in addressing economic challenges through redefining spatial interactions. Concurrently, other scholars focused on the production of space within the fixing process. Observing the reduced physical investment, Stehlin et al. (2020) envisioned a ‘fragile spatial fix’ resulting from a data-centred accumulation strategy, thereby absorbing over-accumulated capital into data production infrastructure without risking the construction of fixed assets at a slow pace. Moreover, Wang et al. (2022) conceptualized digital platforms as ‘techno-spatial fix’ or ‘infrastructures’ sprawling into rural areas to construct new markets and harness new labour forces.
The profound impact of platformization in rural China
E-commerce, as a digital infrastructure, shows a tremendous social impact in rural areas, as reported in recent literature; it is considered a ‘technical catalyst’ for the development of the village’s industrial structure and the employment of local inhabitants (Zhang et al., 2022). Regarding the local economy, e-commerce-enabled rural growth is prominent in non-agricultural industries, particularly manufacturing (Li and Song, 2023). Some cases were reported in Chinese TBVs, such as small-commodity manufacturing in Qingyanliu, outdoor products in Beishan and Bainiu and wooden furniture in Dongfeng (Li, 2017; Leong et al., 2016; Wang, 2025; Wu et al., 2020). Furthermore, platform-enhanced extra-regional linkages are considered capable of stimulating improvements in product quality and variety, as evidenced by the clothing industry in Wuxing and hardware accessories production in Yuyao (Zhang et al., 2023). Notably, a few studies also contended that e-commerce facilitated the expansion of family-run farms (Zhang et al., 2024).
Regarding employment patterns, the e-commerce-driven shift away from agriculture and improvements in household income were highlighted alongside industry restructuring in TBVs. With the introduction of rural e-commerce, villagers have shifted away from a reliance on farming toward diversifying their livelihoods through small-scale manufacturing and commercial services (Zhang et al., 2022). These activities transformed farmers into e-commerce entrepreneurs who prefer to open online shops from home rather than migrate to cities for work (Wang, 2025). Moreover, the growth of rural e-commerce has further boosted rural economies; a strong body of research consistently finds a positive link to higher household incomes (Li and Qin, 2022; Li et al., 2024; Tang et al., 2022; Zhang et al., 2024). As a result, local villagers have shown a decreased tendency to migrate for employment, a trend supported by household survey data from recent studies (Qi et al., 2019; Wei et al., 2025).
Recently, a few studies have begun investigating the platform-facilitated spatial transformation of the village landscape. Initially, Lin (2019) developed a conceptual framework for e-urbanism in TBVs, with three interwoven layers, including spatial production as a core layer. The empirical part of this research reported an e-commerce-affected, grid-pattern-organized urban form in the case of Lirendong Village. Subsequently, Li and Song (2023) documented a spatially linear pattern of the e-commerce industrial park, popularly termed ‘TaoBao street’, in Daji Town, where buildings for e-commerce shops, small garments and factories are distributed along the newly built main street of the town. In a pioneering attempt, Zang et al. (2023) took land use as the basic parameter in the empirical analysis of the physical landscape for TBVs. This study echoed the research by Li and Song (2023), showing a linear pattern in the expansion of built-up land along the transportation corridor in Dongfeng Village. Furthermore, the case of Qinyanliu Village reveals a more subtle layer of landscape transformation – the emergence of branding walls to promote its e-commerce development (Wang et al., 2021).
Current gaps and proposed theoretical framework
The current critical debate on platform capitalism either highlights the functional aspect of platforms in addressing the inherent capitalist crisis or conceptualizes platforms as a new type of data-centred infrastructure that absorbs over-accumulated capital. They both merely allude to the outward ‘spatial fix’ by the digital platforms, while overlooking the profound spatial restructuring of the mobilized ‘crowd’ that surrounds both sides of platform-based mediation. Empirical research on platform-driven modifications of the rural landscape remains limited, as it either narrowly focuses on a single TBV or investigates a singular spatial dimension.
In this research, the concept of ‘spatial fix’ coined by Harvey (1981, 2001) is employed as a theoretical framework to investigate and interpret the multifaceted spatial materialization of rural e-commerce development enabled by platform capitalism. Subsequently, the research investigates the specific agency of e-commerce platforms in rural areas as they mobilize and control the outsourced production of space through an algorithmic system.
Methods and materials
Research design
Theoretically framed by the spatial fix of platform ruralism, this mixed-methods interpretive study focuses on the multifaceted physical materialization of platform capitalism in the very first generation of TBVs in the GBA. This research echoes the advocacy for methodological triangulation as a critical strategy in urban planning research, recognizing that no single approach can comprehensively unfold the complexity of an urban issue (Ogunkan and Akinpelu, 2026). This strategy enables researchers to generate more nuanced and reliable insights by deliberately combining multiple methods and cross-validating them to investigate a single phenomenon.
Against such a background, this research triangulates: 1) morphological analysis to trace the rural landscape changes since the establishment of the TaoBao platform, 2) semi-structured interviews with knowledgeable local informants to provide supplementary information on landscape and livelihood changes along with the penetration of rural e-commerce and 3) field observations to validate and supplement the findings from the previous two dimensions of investigation. Those methods complement and cross-validate one another, yielding reliable findings on the transformation of the rural landscape catalysed by e-commerce penetration. Morphological analysis provides an overall understanding of longitudinal changes in three morphological elements: patches, streets and buildings. Semi-structured interviews and field observations contribute to a nuanced capturing of the building structure adaptation and the accompanying livelihood transformation.
Echoing that ‘urban form can only be understood historically’ (Moudon, 1997), morphological analysis was carried out by employing three observation time points: 2003, 2013 and 2022. These time points were strategically selected under the following considerations: 1) 2003 was the year of the TaoBao e-commerce platform establishment, 2) 2013 was one year before when the first-generation TBVs were identified in GBA by AliResearch, 3) 2022 was the year of the latest TBV list issuing as of the time of this investigation and 4) such selection allows for mapping and analysis of the TBVs at approximately equal intervals (10 years).
Study area
The first generation of TBVs in GBA was selected as the study focus for the following considerations: 1) located in GBA, a region with well-developed infrastructure and numerous TBVs, 2) informing the subsequent TBV generations and 3) being geographically representative in the region. First, GBA created good conditions for e-commerce development: the national highway and the expressway have reached every county-level administrative unit (Department of Transport of Guangdong Province, 2017, 2020), and broadband access has been extended to all administrative villages as early as 2014 (People’s Government of Guangdong Province, 2014). Under such circumstances, a total of 921 villages were identified as TBVs in 2022, accounting for 11.83% of the 7780 TBVs in China (AliResearch, 2022). Second, first-generation TBVs have a first-mover advantage due to their earlier exposure to e-commerce platforms, thus serving as a proxy for revealing the developmental trajectory of subsequent generations. Third, the first-generation TBVs are relatively evenly distributed across the region, serving as a representative microcosm of the surrounding TBVs.
A total of 32 villages were ultimately selected in accordance with the principles outlined above. Those first-generation TBVs are typically located in flat terrain areas in ten counties across four cities in GBA (see their spatial distribution in Figure 1). The majority are located in the suburbs near Guangzhou and Foshan, while the others are in Huizhou and Jiangmen in the eastern and western GBA, respectively. Each is typically specialized in one or two product categories sold on e-commerce platforms, as summarized in Table 1.

Study area – the first-generation TBVs (N = 32) in the GBA.
First-generation TBVs in GBA and their product specialization.
Data collection and interpretation
First, the morphological analysis of landscape changes in TBVs focuses on three urban form elements: patches, streets and buildings (Berghauser Pont et al., 2019; Conzen, 1960; Kropf, 2009; Moudon, 1994; Thornton et al., 2011; Whitehand, 2001; With, 2016). To obtain historical landscape data, a total of 96 groups of remote sensing images of the first-generation TBVs in GBA across three different years were collected from Google Earth satellite imagery, with spatial resolutions ranging from 0.27 m to 2.15 m. Manual interpretation was used to ensure accuracy in classifying landscape patches and in extracting urban elements (i.e., streets and buildings) from the satellite images. For categorizing landscape patches, 10 types of spatial landscapes were identified based on field observations and existing studies, as shown in Table 2 (Ding and Chen, 2022; Feng and Liu, 2015; Kraff et al., 2022; Lee, 2020; Li et al., 2021; Wen et al., 2020; Zhao, 2000). Following this, two experienced assessors manually interpreted and mapped the data in CAD using georeferenced satellite imagery and validated the results through cross-checking. Finally, the polygons of each category were imported into GIS for further morphological analysis. Regarding the extraction of streets and buildings, the two assessors, according to the procedures described above, manually mapped street networks and building footprints by referencing Baidu Maps.
Classification for landcover interpretation.
Satellite images of 100 m by 100 m.
Subsequently, the investigation was supplemented with 26 semi-structured interviews carried out in August 2024 and July-August 2025. Participants were recruited using a combination of purposive and snowball sampling strategies. Firstly, village cadres were interviewed to gain an overall understanding of landscape transformations and villagers’ livelihood changes over the last two decades, with a particular focus on the period following the certification as a TBV. Then, following the principle of snowball sampling, village cadres were invited to recommend other informants knowledgeable about e-commerce-related development in the village. Moreover, both e-commerce practitioners and ordinary villagers who were not necessarily involved in the e-commerce business were invited to complement and cross-validate the information retrieved from village cadres and government officials. The sampling process was discontinued when data saturation was reached (Guest et al., 2006). At last, a total of 26 informants from 11 villages or upper-level governments took part in the semi-structured interviews, including eight village cadres, two government officials, nine e-commerce practitioners and seven ordinary villagers. Each interview typically lasted between 10 and 60 minutes.
Lastly, field observations were carried out during the same fieldwork campaign, covering all 32 first-generation TBVs in the GBA. The observations aimed to gather first-hand information to validate the findings from the semi-structured interviews and supplement the morphological analysis of landscape changes. Three main activities were part of the field observations: 1) photographing the TBV landscape, 2) recording the activities of the e-commerce businesses and 3) observing respondents’ reactions to the interview questions. These observations are important for providing first-hand insight into the spatial conditions in TBVs and the villagers’ attitudes toward e-commerce involvement in their livelihoods.
Data analysis
Three levels of morphological analysis examined the physical transformation of the landscape: 1) land cover distribution, 2) street network and 3) accessibility to productive landscapes. First, considering the potential impact of the spatial structure identified in the preliminary screening on landscape land cover configuration, the study focused on morphological categorization. It then further analysed the changes by selecting a representative case from each category. The following spatial characteristics were used in morphological categorization: 1) the number and area of patches by land cover type and 2) the adjacency and relative location of residential areas and productive landscapes (i.e., croplands and factory buildings), referencing relevant literature (Çalışkan et al., 2023; Harris and Ullman, 1945; Hoyt, 1937; Jia and Wang, 2024). Three morphological categories emerged from the analysis (Figure 2): 1) cluster (N = 8), characterized by single-type land cover concentrations and segregation between different types; 2) sector (N = 8), characterized by manufacturing facilities distributed along transportation lines with agricultural and natural landscapes in between; and 3) multi-nuclei (N = 16), characterized by interwoven productive and natural landscapes centred around several discrete residential nuclei. Liwu, Baijiang and Heting were selected as representative cases, each exemplifying a distinct morphological category. Categorical maps of the 10 identified patches were used to report the landscape patch composition in each village and to track morphological changes in 2003, 2013 and 2022 (Cuba, 2015).

Morphological categorization of the first-generation TBVs (N = 32) in GBA.
Second, the regularity of the street network was analysed to understand how human intervention influences the morphological features of mobility infrastructure, as well as the penetration of the rural platform economy. Three aspects of the street network – 1) the straightness of streets, 2) the intersection angles and 3) the duplication of similar street patterns – were examined over time by overlaying three years of street systems with different colours.
Third, accessibility to two categories of productive land cover types – agriculture (i.e., cropland and fishpond) and manufacturing (i.e., factory building) – was examined to understand how manufacturing-led rural development influences villagers’ traditional agricultural livelihoods during the process of rural platformization. Specifically, the cumulative-opportunities accessibility measure was used, which calculates the ratio of landscape surface area to the total area within a reachable convex hull of each residential building (Berghauser Pont et al., 2019; Bhat et al., 2000). Using network analysis, street-based convex hulls were created from origins by applying a 1000 m distance threshold to simulate the basic radius of rural daily life (Liu et al., 2022; Ulceluse et al., 2022; Wang et al., 2024). The accessible agriculture index (AAI) and accessible manufacturing index (AMI) are then computed with each individual building as the basic unit by the following formulas.
where S(R;D) is the surface of the reachable area within the distance threshold D from each residential building R; S(CF) is the total surface of the cropland and fishpond; S(FLS) is the total surface of the factory building landcover type. To longitudinally trace the changes in accessibility to two categories of productive landscape, AAI and AMI were normalized per village.
Findings
The physical transformation of the rural landscape
First, the land cover of enhanced manufacturing and degraded agriculture is equally present across three observed morphological categories, as shown in Figure 3. The surface areas of natural and agricultural landscapes decreased by 3.63% and 14.86%, respectively, from 2003 to 2022. This decline was largely offset by the expansion of residential areas by 7.43% and industrial areas by 11.81%. Specifically, the Cluster type (i.e., Liwu) expanded its factories through gradual encroachment into nearby cropland, maintaining a coherent and single patch. The factory building surface area more than quintupled from 4.58% to 25.59%, while cropland decreased by one-third from 48.03% to 31.23%. The Sector type (i.e., Baijiang) shows an approximately linear distribution of factories along the existing road in 2003, after which the factories diffused and upgraded along the newly built transportation line in 2013; as of 2022, the total factory building surface increased slightly by 3.21%. The Multi-nuclei type (i.e., Heting) distributes its factories not only close to the edge of residential nuclei but also expands into the middle of the agricultural and natural environment, indirectly abandoning the fishponds and fragmenting the cropland and forest with factory building patches. The surface area of fishponds dramatically decreased from 22.10% in 2003 to 6.56% in 2022, while cropland and vegetation saw minor decreases of 2.64% and 2.18%, respectively, in 2022.

Landcover transformation in three TBV morphological categories.
Similarly, the decontextualization of the street network’s spatial structure is observed across all three morphological categories, particularly Cluster and Sector, as shown in Figure 4. Specifically, the street networks of both Cluster and Sector underwent significant expansion from 2003 to 2013, a period during which the grid pattern was initially established. This regularized street system was further enhanced in the following decade with minor developments. Similarly, minor street extensions along the main road are observed in the Multi-nuclei type, exhibiting a mixture of fishbone and grid patterns.

Spatial structure changes of the street network.
Lastly, the shift in residential proximity from agriculture to manufacturing is consistently observed across all three categories and is quantitatively examined through changes in accessibility from residential buildings to these two productive-landscape categories, as illustrated in Figure 5. In general, AAI experienced a comparably gradual decrease of 49.09% from 2003 to 2022; however, AMI sharply increased first, more than tripling from 0.154 to 0.495 in 2013, and was followed by a minor growth (0.023) in the later decade. Specifically, Cluster shows the largest change, with an average AAI decrease of 84.26% over the last two decades, followed by Sector and Multi-nuclei, at 51.09% and 34.00%, respectively. Similarly, Cluster shows the most significant increase in average AMI (255.32%), followed by Sector (228.56%) and Multi-nuclei (210.72%).

Residential accessibility to two categories of productive landscapes – agriculture (above) and manufacturing (below).
On-the-ground perspectives from local inhabitants
The semi-structured interviews add evidence on the industrialization of rural landscapes and the de-agriculturalization of local livelihoods that occurs alongside platformization. When asked about their observations of rural landscape changes, all interviewees explained that platform-based businesses expanded the urbanized and industrialized areas within the villages. At first, e-commerce platforms only caused small shifts: the reconfiguration of household spaces into self-operated small factories, as 19 interviewees reported (including e-commerce factory owners, workers, village cadres and residents from Baijiang, Dazha, Dongxi, Jili, Liwu, Liaozai and Rongzhou Taiyang Village, personal communication, 23–29 August 2024, 23–24, 25–27 July 2025). One villager said that: This [factory] was converted from the ground floor of my house. . .and many other [village buildings] have been converted in the same way. My house is located in the inner part of the village. No customers come here to buy [in person]. All products are sold online [via e-commerce platforms] and can be customized.. . . Zhongtong Express [staff] comes here to pick products up every day. (A furniture e-commerce factory owner, Dazha Village, personal communication, 26 August 2024)
This adaptation process was met with tacit consent from administrative bodies at both the town and village levels: ‘as long as it [converting ground space to a small factory] doesn’t make major impacts, we [town government] will not intervene’ (government official, Baihua Town, personal communication, 24 July 2025); ‘only public security and fire protection are strictly monitored. . .especially fire protection, you must meet the requirements’ (village committee cadre, Liwu Village, personal communication, 26 July 2025). Subsequently, nearby village industrial parks underwent gradual expansion to accommodate relocating factories seeking to scale their operations and facilitate their e-commerce activities (village cadres, Baijiang, Dongxi, Liwu, Taiyang and Tianxin Village, personal communication, 23–24, 26–27, 29 July 2025, 2 August 2025). Recently, local governments of villages with thriving e-commerce have been actively investing market capital into building e-commerce industrial parks and live-streaming offices to meet the rising demand for space and property services, such as Volor Furniture E-commerce Park and Chaoo E-commerce Park (e-commerce service centre officer and shop owner, Dayuan Village, personal communication, 26 August 2024, 29 July 2025). The rural character of the village landscape has been significantly diminished in these e-commerce parks: ‘If you did not mention, I would not consider here a village. It does not look like a village at all. I feel I am working in the city’ (live-streaming shop employee, Dazha Village, personal communication, 26 August 2024).
When asked about the current livelihoods of the local residents, the process of de-agriculturalization is confirmed by the local village committee members and all villagers interviewed. Concerning the existing farmland and fish ponds, they are primarily cultivated by the previous generation – parents or relatives of current e-commerce workers (village cadres, e-commerce practitioners and ordinary villagers, personal communication, Dongxi, Liaozai, Liwu, Rongzhou, Taiyang, Tianxin Village, 22–25, 28 August 2025, 29 July 2025, 2 August 2025).
There are almost no young people left [in our village] to farm; they’ve all gone to work in nearby factories or in the county. A few elderly people [still] do farming, but they just for themselves to eat and don’t sell. If it were not for the Permanent Basic Farmland [Protection Policy], [those farmlands] cannot be developed, the farmland in our village would disappear. (Village committee cadre, Tianxin Village, personal communication, 29 July 2025)
Apart from the villagers directly involved in e-commerce business, some others are engaged in platform-derived business, such as express delivery: ‘I used to work outside [of the village]. . .and I came back to [work for] express delivery. Most of the packages are theirs [e-commerce factories]’ (Express staff, Liaozai Village, personal communication, 23 August 2024).
When it comes to the motivation behind these livelihood changes, reasons such as increased income, industrial infrastructure, accessible logistics facilities and e-commerce platform support were highlighted: ‘This is my own house, and I don’t need [to pay for a] rental fee.. . . Selling these [tires] makes more money than farming.’; ‘Our village has been doing this [tire recycling] business for a long time. There is a large tire market nearby.. . . It is very convenient to have the express delivery come here to pick up (tires).’ (Second-hand tire shop owners, Liwu Village, personal communication, 24 August 2024). During the interview, the researcher heard a notification sound from the online order platform (Qianniu), and the shop owner smiled and explained: ‘Business coming’ (Second-hand tire shop owner, Liwu Village, personal communication, 24 August 2024).
Immediate slices from field observations
An industrial, urban-oriented landscape was observed at both village and building levels through field observations. At the village level, landscape changes were mainly driven by the creation of ‘e-commerce parks’ and ‘industrial districts’ by local governments, typically at the county level. These physical transformations are ongoing, continually replacing traditional rural features (e.g., Dazha Village Industrial Park and the warehouse and logistics park under construction; see Figure 6). At the building level, field observations confirmed the subtle transformation of functional vertical differentiation noted from interviews.

Platform-enabled industrial landscape at the village level: industrial district, e-commerce logistics park and live-streaming parks.
The ground space, originally designed as the villagers’ living area, has now been adapted into a small manufacturing facility (Figure 7). Scattered unfinished products along the sidewalks, electric cargo tricycles parked on the streets and the distinct smell of metal processing in the air collectively create an industrial atmosphere. The upper floors remain as residential spaces, serving as homes for owners and rental housing for migrant workers (Figure 7).

Platform-enabled industrial landscape at the building level: vertical, functionally differentiated and mixed-use village houses with self-owned factories on the ground floor and residences above.
Moreover, a notable trend that comes with the growth of rural e-commerce is the rise of spatial uniformity at both the village and building levels. At the village level, the street network in TBVs, especially in the industrial district, has been designed without clear regard for the local context and morphological features (e.g., rigid-geometry gridiron and concentric semi-circular patterns with a central axis, shown in Figures 8(a1) and (b1)). At the building level, e-commerce office buildings and standardized factories or warehouses (Figures 8(a2) and (b2)), funded by external investment, tend to become similar by losing the rural traits of the local traditional landscape. This spatial standardization is seen as a side effect of prioritizing efficiency in plot allocation and building development.

Generic spatial patterns and landscapes in TBVs. (a1) Spatial pattern of a concentric semi-circle street network with a central axis in China Ceramics City (Huaxia Taoci Bolancheng) at Jili Village. (a2) A typical street with e-commerce shops on both sides. (b1) Grid pattern in Liwu Village industrial park. (b2) A common building structure for logistics and warehouses.
Discussion
This interpretive study, theoretically anchored in the spatial fix of platform ruralism, traced transformations in the rural landscape during the process of economic platformization of the GBA. The findings reveal a three-dimensional expansion of the industrial landscape, showing how platform capital gradually occupies the rural area, shaping an industrial, urban-oriented landscape in rural China (Srnicek, 2017). The decontextualized street network and the shifting of rural residential proximity to manufacturing potentially highlight the power of platform capital in indirectly restructuring land use to serve the needs of a platform-mobilized production of space (Montalban et al., 2019). The de-agriculturalization of local residents’ livelihoods and their involvement in the platform economy also clearly indicate the rise of platform ruralism in China (Barns, 2019; Liang et al., 2022). Applying the Marxist-grounded geographical perspective of the spatial fix helps understand that the fixation of capital in rural areas through platformization has a dual nature: on one hand, it manifests in new physical infrastructure; on the other hand, it diminishes private space and time. Additionally, recent discussions on platform urbanism reveal that socio-technical changes driven by the rise of digital platforms have already impacted society and space (Leszczynski, 2020). The spatial effects of platformization are extensive, affecting not only specific locations but also territorial configurations (Caprotti and Liu, 2020). This research further offers empirical support for these theoretical ideas, based on solid evidence from TBVs in rural China.
Unlike existing studies that have incorporated the spatial dimension in their investigation of TBVs but overlooked the morphological perspective (Li and Song, 2023; Lin, 2019; Zang et al., 2023), this research identified three landscape composite types (Figure 3) that consistently showed the expansion of the manufactural landscape and the shrinkage of agriculture since the year the TaoBao platform was established. The scholarly contribution of this study is multifaceted. First, as the manufacturing landscape already occupied a relatively small area in 2003, this study posits that the industrial foundation was one factor that spurred the formation of TBVs (Li et al., 2024; Lin, 2019). Second, the continuous expansion of manufacturing and the degradation of agricultural landscapes from 2003 to 2022 further support the idea that platformization significantly impacts the rural landscape by facilitating rural industrialization. Third, the linear platform-based expansion of the productive landscape along the main road, as shown in the Sector type, aligns with the findings in the research by Li and Song (2023) and Zang et al. (2023). Finally, this research further enriches the understanding of platform-mediated rural landscape spatial patterns and evolving trajectories by identifying three landscape composite types.
Furthermore, this research showed how the layout of street networks in TBVs is often done with little regard for existing topographic and historical conditions, leading to a significant increase in the gridiron pattern (Figure 4). This pattern is more evident in the first decade, supporting the idea that TBVs generally have an industrial basis (Lin, 2019; Wei et al., 2020). Additionally, the active construction of ‘e-commerce parks’ and ‘industrial parks’ by county-level governments can be inferred from findings in semi-structured interviews and field observations. These findings highlight the important role of local governments in shaping the landscape during the platformization of the rural economy (Chu et al., 2024; Cui et al., 2019; Li, 2017). However, the results also show an emerging homogeneity in spatial patterns during this process. On the one hand, this trend can be seen as a spatial by-product of conventional practices aimed at increasing plot allocation efficiency. On the other hand, it reflects how rural space, as a mobilized platform, adapts to the needs of platform-mediated manufacturing, revealing the materialization of generic villages within the platformization of rural China. This trend mirrors developments seen in urban areas of Asia over recent decades (Koolhaas, 2013).
Another key finding is that residential buildings enhance their connectivity to manufacturing areas at the expense of agricultural ones (Figure 5), indicating an ongoing shift of local residents’ livelihoods from agriculture to manufacturing. This research demonstrates the transition of residents’ daily lives between the two types of productive landscapes during the process of platformization, confirming the platform-driven de-agriculturalization of the territorial space (Lin et al., 2022; Wang et al., 2024; Zhang et al., 2022).
Lastly, this research highlights the two levels – village and building – of landscape transformation during rural platformization, drawing on first-hand information from semi-structured interviews and field observations. Inference can be drawn from the fact that the industrial landscape not only materializes in terms of land cover but also appears more subtly – the vertical functional differentiation at the building level (Figure 7). Additionally, the emerging homogeneity is embedded in both scales – the regularized street network pattern at the village level and the commonized spaces within individual buildings (Figure 7). This research scrutinizes the platformized landscape across multiple scales, thereby addressing the limitations of single-scale investigations in prior studies (Li and Song, 2023; Lin, 2019; Zang et al., 2023).
Conclusion
Rural areas are not immune to the current wave of widespread e-commerce growth. Over the past few decades, platform capital has significantly impacted the rural landscape in the GBA of China, resulting in a spatial fix of platform ruralism that has gradually created platform-dependent economic systems, such as the TBV. However, the transformation of the rural landscape with e-commerce penetration currently results from a lack of a comprehensive strategy. By examining the first generation of TBVs in GBA, this interpretive study provides a detailed snapshot of physical changes and shifts in residents’ livelihoods during rural platformization, using a methodological triangulation. The findings show that, with the rise of e-commerce, the rural landscape has experienced extensive spatial reorganization. At the village level, this restructuring involves the expansion of the industrial areas, the de-contextualized street network and the increased distance between residences and industrial sites. At the building level, it is reflected in the vertical functional differentiation of village houses and the construction of generic buildings for logistics and warehouses.
By examining the spatial changes in the first-generation TBVs of GBA, this study aims to add to the ongoing critical discussion on platform capitalism in rural China. Theoretically, its novelty lies in using the spatial fix of platform ruralism as an interpretive lens. Empirically, the innovation is in the three-level morphological analysis and highlighting that such physical changes occur at both village and building scales. Additionally, combining morphological analysis with semi-structured interviews and field observations goes beyond just land cover changes to include social and cultural shifts. Lastly, using methodological triangulation across various scales not only produces solid findings but also presents a novel framework for future research. Nevertheless, this study is not free from limitations. The way cases are chosen within each morphological category may influence comparisons. A systematic morphological analysis of all first-generation TBVs is needed and may be undertaken by future studies.
This research explores the contemporary spatial order influenced by platformization and highlights that e-commerce is a key force shaping the territorial landscape. Although only first-generation TBVs in the Chinese GBA have been examined, the findings provide insights into future TBVs and other e-commerce villages, both within the GBA and across China and the world. In light of findings, policies should be developed to 1) regulate the rapid growth of industrial zones and minimize their ecological and hydrological impacts, 2) strictly safeguard agricultural production to ensure food safety and security and 3) plan and design new industrial districts considering local topography and cultural landscapes.
Footnotes
Acknowledgements
The authors express their gratitude to all the interviewees for their time and willingness to participate and anonymous reviewers for their thoughtful comments and constructive critiques. The authors acknowledge the use of Grammarly for grammar editing of the article.
Ethical considerations
The appropriate ethics review board (City University of Hong Kong Human Subjects Ethics Sub-Committee) approved the study design.
Reference No.: 26-2022-57-E
Application No.: H003058
All the interviewees gave their consent to participate in the study.
Consent to participate
The study comprises 26 interviewees. All of them provided verbal informed consent to participate. The first author conducted all the interviews, informing participants about the research purpose, duration, procedures, and how the data would be used, and asking them to provide either written or verbal consent to participate. Fifteen interviewees were audio-recorded with previous consent and later manually transcribed verbatim and translated. The first author recorded the remaining 11 interviews (from interviewees who disagreed to be audio-recorded) in handwritten notes and later transcribed them.
Consent for publication
Not applicable: The dataset was anonymised (removing all individual/identifiable details, images or videos). Personal/identifiable information is not necessary to the study.
Author contributions
Weike Li: Conceptualization; Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Writing – original draft. Jinyi Hu: Data curation, Formal analysis, Investigation, Visualisation, Writing – original draft. Gianni Talamini: Conceptualization, Funding Acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft.
Funding
The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: The work described in this paper was fully supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. CityU 11611422).
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
This study comprises three datasets: 26 interview transcriptions (*.docx), on-site photographs depicting landscape transformation in 32 villages (*.jpg) and geospatial data of 32 remote sensing imagery covering all 32 villages (*.tif), alongside nine shapefiles of manually interpreted land-use types, nine shapefiles of street network structures and 18 shapefiles of residential accessibility to productive landscapes in the three selected villages (*.shp). The datasets are part of a larger data collection campaign aimed at building a database for an ongoing, funded research project. Part of the data used for this study will be further analysed for different objectives; therefore, they must remain confidential during this critical research phase. Nevertheless, the authors are willing to make the datasets available to the editors and reviewers upon request within the scope of the review process.
