Unveiling Sustainable Co-Creation Patterns in Entrepreneurial Ecosystems of Shanghai’s High-Density Urban Communities

Abstract

Communities in China’s high-density cities, like Shanghai, are evolving from traditional residential roles into vibrant centers of entrepreneurial innovation. This research delves into the development of community-supported entrepreneurial ecosystems (CSEEs) in the city, with a specific focus on the sustainable co-creation mechanisms facilitated by stakeholders, explored through a comparative study framework. By utilizing Kelly’s Repertory Grid Technique, 14 essential elements of co-creation are identified, which form the framework for classifying the cases into three distinct types. This study employs in-depth interviews and content analysis to analyze and contrast how these co-creation patterns are applied across cases representing each type. The results show that key factors, such as resource origins, the interaction between CSEEs and embedded communities, and participant selection strategies, significantly shape the variations in value co-creation mechanisms, processes, and outcomes. Recognizing the variety of co-creation models is crucial for enhancing both the vitality and efficiency of Shanghai’s CSEEs. Furthermore, this study offers valuable insights into managing co-creation efforts and predicting risks in similar contexts, contributing to the sustainable regeneration of urban areas through community-driven entrepreneurship and innovation.

1. Introduction

The explosive population growth and rapid urbanization in China have significantly depleted land resources, making urban stock regeneration increasingly recognized as an essential development strategie, particularly in the residential communities of densely populated cities [1]. In modern China, residential communities have evolved from meeting basic living needs to serving as multifaceted spaces catering to a broader range of social and commercial functions [2]. As one of China’s major metropolitan cities, Shanghai contains numerous aging communities whose infrastructure and services no longer meet the diverse needs of contemporary residents [3]. In response, the Shanghai Municipal Planning and Land Resources Bureau has introduced a community planning strategy aimed at enhancing living quality and promoting sustainable development. This strategy includes the implementation of the “15-Minute Community Life Circle” initiative, which integrates diverse infrastructure and services, such as healthcare, education, entertainment, dining, and elderly care, within a 15 min walking radius of each community [4]. This initiative has significantly accelerated the development of entrepreneurial projects embedded in communities, which aim to provide these lifestyle services and experiences. To develop entrepreneurial projects that benefit both local community welfare and the sustainable development of new businesses, universities and social enterprises have taken proactive measures [5]. They have established entrepreneurial ecosystems within community spaces to support the development and expansion of new enterprises, featuring interconnected elements, such as resource linkage channels, stakeholder engagement and collaboration, as well as mentoring and support services, and mentoring and support services [6].

With a strong foundation in community regeneration, Shanghai’s residential communities create a favorable environment for developing entrepreneurial ecosystems (EEs), underpinned by public acceptance of innovative services and a culture of active co-development of new experiences [7,8]. Historically, communities were primarily seen as spaces dedicated to daily life and leisure. In the traditional innovation chain—spanning invention, engineering, design, production, distribution, and consumption—communities were positioned at the consumption end, far removed from the innovation processes occurring in other urban functional zones [5]. This paradigm is undergoing a fundamental shift as communities begin to develop entrepreneurial ecosystems. Compared to urban centers or other functional zones, embedding an EE within a community offers advantages across various dimensions [9]. Initially, given the proximity of communities to everyday life issues, entrepreneurs are well-positioned to identify diverse entrepreneurial opportunities within user behavior, a core principle of the lean entrepreneurship [10]. Moreover, the diverse and complex needs of residents, who act as both users and customers, foster a profound understanding of user preferences. This understanding allows creative inventions and applied technologies to be translated into viable business prototypes under a co-development mechanism [11]. Additionally, as informal social spaces, communities, particularly older ones, actively seek external forces to boost local innovation and vitality, providing high flexibility for business prototype testing and tolerance for trial and error [12]. Driven by these internal and external advantages, communities have increasingly become platforms and active participants in entrepreneurial activities [13]. The diverse social interactions between entrepreneurs and community members not only provide strong support for identifying business opportunities, developing prototypes, and testing [14,15], but also stimulate momentum for bottom-up community regeneration [16]. Thus, the emerging entrepreneurial ecosystem in Shanghai, deeply integrated within the community and characterized by frequent resident interaction for research and testing, has been termed by researchers as a Community-Supported Entrepreneurial Ecosystem (CSEE) [5,17]. Within this framework, the community functions as a physical boundary, a platform for social interaction, and a resource provider [18,19].

Entrepreneurship is a creative process characterized by substantial interactions among agents and strategic actions [20]. Within entrepreneurial ecosystems, the co-creation by diverse participants is a crucial element in facilitating the sustainable generation of value [21]. As a broad concept encompassing collaboration among multiple stakeholders, co-creation has attracted considerable attention from both academia and practitioners in recent years [22]. It refers to the process in which stakeholders, including firms, consumers, and partners, collaboratively develop products, services, or solutions. This approach highlights the interactive and participatory nature of the innovation process [23]. However, the existing research on co-creation primarily focuses on action experiences within specific corporate management or organizational innovation contexts. The research on co-creation with entrepreneurial goals often focuses solely on static analyses of individual elements or their management methods, lacking an in-depth exploration of how these elements interact within operational mechanisms [24]. Additionally, many studies rely on single case study methodologies to analyze causal relationships between co-creation inputs and outputs, lacking typological analyses of co-creation models across various cases [25]. This limitation hinders the provision of comparative insights into community contexts with uneven resource availability and varying action capacities. As the social and economic values of Shanghai’s CSEEs become more prominent, increasing numbers of studies and practices have emerged. Researchers and practitioners need comprehensive theoretical guidance on co-creation mechanisms and references to typified action experiences. Therefore, clarifying the basic components of co-creation mechanisms and their differentiated patterns in Shanghai’s CSEEs is essential for optimizing resource utilization and enhancing the efficient conversion of co-creation outcomes.

To address these issues, this study poses three research questions: (1) What are the constituent elements of the co-creation mechanism in Shanghai’s CSEE cases? (2) What are the implementation patterns of co-creation mechanisms across different case types? (3) What are the variations in and underlying causes of these co-creation mechanism patterns? The structure of this study is as follows: First, 14 constituent elements of the co-creation mechanism in Shanghai’s CSEE cases are identified using the Kelly’s Repertory Grid Technique. Subsequently, based on the performance characteristics of the cases with respect to these 14 elements, the cases are classified into three distinct co-creation types. The implementation of co-creation mechanisms from representative cases are analyzed, and their dynamic patterns are visualized. Finally, the commonalities and differences in these mechanisms, along with their causes, are compared. The findings offer valuable insights into selecting management and improvement strategies for stakeholder co-creation in diverse CSEEs in high-density cities similar to Shanghai, thereby fostering sustainable community regeneration from the perspective of endogenous innovation.

2. Related Works

2.1. Defining the Community-Supported Entrepreneurial Ecosystem (CSEE)

As a conceptual foundation of CSEEs, the entrepreneurial ecosystem (EE) represents interrelated systems that influence entrepreneurs in creating and scaling new ventures sustainably [26]. As a typical regional entrepreneurial network, the EE enables the rapid exchange of talent, ideas, information, and resources, enabling entrepreneurs to acquire what they need at each stage and translate it into prompt action [27]. The goal of an EE is to promote the formation, survival, and growth of new ventures [28]. Its core outcomes include productive entrepreneurship and social entrepreneurial achievements [29]. Successful EE cases in Silicon Valley, London, and Barcelona show that well-established participant interaction mechanisms and services are critical to EE vibrancy and innovation quality [30]. A common feature of these cases is the adherence to the spatial principle, wherein geographical proximity enhances network formation and knowledge exchange among stakeholders [31,32]. Moreover, the key components of EEs include access to systems, human capital, funding and investment sources, regulatory frameworks, education and training, cultural support, physical infrastructure, incentive policies, and services that coordinate knowledge resources [33].

Given the vast spatial, cultural, and social resources within Chinese residential communities in high-density cities, like Shanghai, CSEEs and their multi-stakeholder co-creation have emerged as cutting-edge interdisciplinary research topics [5,17,34,35]. CSEEs allow entrepreneurs to seize opportunities emerging from user behaviors in real-life scenarios, fostering entrepreneurship driven by local needs and lifestyles. Simultaneously, this ecosystem enhances the vitality of underserved communities through a close interaction between external innovators and local stakeholders during the development of entrepreneurial projects [36]. Factors influencing the sustainable operation of CSEEs include complex issues, such as trust, entrepreneurial culture, transportation and infrastructure, responsible resource utilization, management services, and the unique characteristics of local communities, which research has shown can impact operational quality either positively or negatively [37]. Correspondingly, evaluating indicators for CSEEs cover various aspects, including the emergence and activity level of new startups [38], system resilience, participant interaction, digital transformation, shared value, system compatibility, and competitive relationship management [39]. A comparison between the characteristics of EEs and CSEEs based on the abovementioned studies is shown in

Table 1. Characteristics comparison between an EE and CSEE.

	EE	CSEE
Spatial features	Downtown areas or university surroundings.	Urban or suburban residential communities.
Outputs	Unlimited types of commercial entrepreneurship.	Commercial and social entrepreneurships driven by future community lifestyle, needs, and well-being.
Targets	Establishing new enterprises, controlling costs, and achieving profit growth.	Providing sustainable solutions for community issues, needs, and social relationships, accelerating market transformation through the utilization of community resources for prototype development.
Stakeholders	Entrepreneurs, anchor enterprises, and universities.	Entrepreneurs, universities, NGOs, and local stakeholders (such as residents, community managers, and workers) with implicit and explicit resources.
Investment	Investors and capital companies.	Investors, capital companies, and community crowdfunding.
References	[28,29,41]	[5,7,8,16,17,34,42]

[40].

2.2. Co-Creation in CSEEs

Co-creation is composed of two main dimensions: co-production and value in use. Co-production involves knowledge sharing, equity, and stakeholder interaction, while value in use includes experience, personalization, and relationships [43]. Co-creation centers on an interactive dialog and mutual learning, aiming to co-construct shared experiences. The value generated can include both products (goods or services) and brands. In this process, customers are not simply consumers, but active co-producers of value [44]. Recent studies have increasingly highlighted the pivotal role of co-creation in entrepreneurial ecosystems. Entrepreneurs are not isolated actors; instead, they collaborate with various stakeholders to co-create value from dispersed latent resources [45]. This collaborative approach enhances resource efficiency while also improving the ability to identify, exploit, and develop opportunities [46].

Co-creative activities within entrepreneurial ecosystems exhibit several defining features. First, co-creation outputs are multidimensional, going beyond product and service innovation to include organizational learning, relationship management, and value proposition development. It enables entrepreneurs to co-develop new markets and products through interactions with stakeholders, such as residents, businesses, and governmental entities [47]. Additionally, stakeholders, ranging from private developers to public institutions, contribute both tangible and intangible resources in the co-creation process. These contributions may be driven by profit motives or public interest, leading to a range of co-creation outputs, including both for-profit and non-profit ventures [46]. Secondly, the dynamic nature of stakeholder collaboration is a crucial aspect. The co-creation process evolves in response to environmental changes, demographic shifts, and emerging needs, requiring continuous adjustments in coordination and leadership to align with evolving market and technological conditions [48]. Thirdly, the complexity of factors influencing co-creation must be recognized. Factors such as the capabilities of involved actors, the effectiveness of intellectual property strategies, social commitments, and the availability of public funding are critical determinants of successful co-creation [22]. Additionally, the interplay among stakeholder bases, organizational foundations, and environmental contexts, along with the number of stakeholder and cultural and goal divergences, significantly influences the capacities of mediators or initiators in managing co-creation processes [21]. Also, dynamic resource exchange capabilities, such as upward innovation support, market leadership, and the expansion of corporate or organizational boundaries, alongside downward capabilities, like minimizing idle resources, maximizing potential, and enhancing identity synchronization, are integral to the quality and vitality of co-creation [49].

The co-creation mechanism refers to a systematic approach and workflow where multiple stakeholders collaborate and interact to jointly participate in the value creation process within organizational, community, or market environments [50]. Core elements of these mechanisms typically include multi-stakeholder participation frameworks, open communication and information-sharing channels, process management strategies, shared goals, and the integration of diverse resources [51,52]. Despite the advancements in co-creation theory and mechanisms in fields such as tourism and financial services, the research remains limited on how these mechanisms function within entrepreneurial ecosystems and across diverse entrepreneurial objectives [22,53]. Future research should focus on how cultural, goal-oriented, and participant differences impact co-creation processes and mechanisms. While the co-creation mechanism offers significant opportunities for participants, it also presents challenges, such as managing complex interactions, aligning stakeholder capabilities, conducting regular evaluations and iterative processes, and balancing diverse interests to improve co-creation outcomes [48]. Effective action strategies are needed to address these challenges.

2.3. Research Gap

Although the relevant literature has examined CSEEs and value co-creation with stakeholders, significant knowledge gaps persist. First, while studies have made progress in identifying co-creation components within entrepreneurial ecosystems, they predominantly focus on individual elements. The relationships among these elements and their interactions within entrepreneurship-oriented mechanisms are still insufficiently understood [54]. Compared to the well-established concepts of co-creation roles and elements in traditional entrepreneurial ecosystems, the research on co-creation within CSEEs has lagged behind policy and practical needs [55]. This is particularly evident from the lack of systematic studies on co-creation mechanisms, resulting in a deficiency in theoretical support for managing and improving co-creation processes. Consequently, existing community resources remain underutilized, and stakeholder engagement and capabilities are not fully activated [55,56].

Furthermore, although the literature emphasizes that stakeholders’ dynamic capabilities are crucial for cross-sector co-creation partnerships to generate social impact and address complex issues [57], the current research inadequately explores how managing these capabilities can improve co-creation quality. This gap limits efforts to harness participants’ latent resources to improve the effectiveness of co-creation mechanisms. Although intermediaries are crucial for identifying and integrating the resources and capabilities of diverse participants, limited research exists on how they coordinate, facilitate, and lead stakeholder actions within the co-creation process [58]. The existing literature often overlooks how intermediaries facilitate co-creation through strategies and methods, and it has not established an actionable paradigm to guide practice.

Methodologically, most studies have analyzed co-creation experiences through singular case studies or within specific industries [59]. However, typological studies on co-creation activities and mechanisms across diverse cultural contexts and entrepreneurial objectives remain scarce. Additionally, the empirical research on implementing co-creation theories across different sectors, goals, and interaction forms, as well as analyses of the relationships between these types and final outcomes, remains limited [24] Given the considerable variation in resource reserves, stakeholder capabilities, and structural foundations across different CSEEs, the absence of typological studies hinders practitioners in similar contexts from tailoring their approaches to their specific circumstances. Moreover, the unclear commonalities and differences in the factors influencing co-creation implementation types, along with the lack of correlation analyses between co-creation processes and outcomes, make it difficult for researchers and practitioners to perform comparative analyses.

To address these research gaps, a more typological and dynamic perspective is urgently needed to analyze the components of co-creation mechanisms and their implementation patterns. Further exploration of the causes of these differences is essential. This approach could provide valuable insights for CSEE actors across diverse social, economic, and cultural contexts, enabling a deeper understanding of dynamic management and implementation in co-creation.

3. Method

The study is structured into three key phases, with the research framework illustrated in Figure 1.

(a)

Phase 1: Identification of Case Characteristics through Stakeholder Interviews

In Phase 1, in accordance with the literature review, entrepreneurial community focal cases published by the local government [4], coupled with a high degree of social impact, such as awards received in social innovation competitions [60], six CSEE cases were selected. The case names and location characteristics are listed in

Table 2. The locations of selected cases.

Number	Case	Location Characteristics
C1	NICE 2035	Located in the Si Ping Community of Yangpu District, adjacent to Tongji University, the project is surrounded by several workers’ new village-style communities established in the 1960s.
C2	Xianxialai Cooperative	The project is situated in an underground air-raid shelter in the Hong Xian Community, Changning District. The space includes a community center offering daily services to residents and a co-creation center supporting young entrepreneurs.
C3	Kangjian Future Artistic Living Co-Creative Space	The project is situated in the Kangjian Community of Xuhui District, home to a permanent population of 87,500, with a high proportion of elderly residents, making it a typical large residential community in Shanghai’s central urban area. It is adjacent to the Shanghai Institute of Technology and several other higher education institutions, boasting abundant educational resources.
C4	GGB Community	Originally located on Handan Road in Yangpu District (now relocated due to spatial adjustments), the project benefits from proximity to several universities, attracting a significant influx of young entrepreneurs from nearby communities. The aim is to support the incubation of youth entrepreneurship projects through community-based experimentation.
C5	Upbeing	The project is located in the Dinghai Community of Yangpu District, once a gathering area for industrial workers, with heritage buildings from a century-old industrial civilization. The project aims to guide public engagement in action learning, both online and offline, based on social issues or visions, thereby fostering social entrepreneurship incubation and co-creating a sustainable community.
C6	Xinhua Community Building Center	Situated in the Xinhua Road community of Changning District, the project encompasses a rich tapestry of historical districts, heritage buildings, old residential complexes, high-end commercial housing, and cultural and office parks. The aim is to catalyze community regeneration and foster entrepreneurial initiatives through a series of community engagement activities and the stimulation of internal community dynamics.

, with detailed case introductions provided in Supplementary Table S1. Comprehensive case descriptions were constructed using a content analysis approach [61], drawing from semi-structured interviews with key stakeholders of selected cases (22 project participants and 10 residents, totaling 25.5 h), and secondary data analysis, including official media interviews, annual reports, and academic speeches. The analysis covered co-creation stakeholders, participation rules, resources, targets, and outputs, which were used as elements in subsequent experiments.

(b)

Phase 2: Classification of Co-Creation Typologies Using Kelly’s Repertory Grid Technique

In Phase 2, the experiment utilizing Kelly’s Repertory Grid Technique (RGT) [62] was conducted for identifying key components and typologies of the case co-creation mechanism. RGT, a cognitive modeling tool for analyzing specific phenomena, aids in elucidating complex system compositions, assessment indices, and features for a particular class of research subjects. It consists of three components, involving elements, constructs, and links between them [63]. The detailed information of experiment participants is provided in

Table 3. The respondents’ information during research Phase 2 and Phase 3.

Research Phase	Respondents Number and Roles
Phase 2 (The RGT process)	18 researchers participated in RGT constructs elicitation and Likert scale scoring. They had an average of over 6 years of experience in researching community business, co-creation, and entrepreneurial ecosystem issues. (Among them, 5 individuals participated in the CVR verification process for the selection of representative case studies.)
Phase 3 (Co-creation mechanism analysis)	16 respondents (involving co-founders, project managers and tutors, and designers from selected cases).
	5 external experts from universities and research institutions participated in CVR for coding results. They had an average of 4 years of experience in researching entrepreneurial ecosystem, stakeholder co-creation, and Shanghai community planning and regeneration.

.

First, during the experiment, descriptions of the selected cases from Phase 1 were presented to 18 participants as experimental elements. Participants were required to read carefully, and researchers clarified any ambiguous or confusing parts until all elements were fully understood. Subsequently, through individual interviews, the participants were asked to select three cases randomly using the triad method [64]. During this process, when the participants identified the similarities in the co-creation mechanism components of two cases as the positive aspect of the construct, they also identified the difference between the third case and the first two, representing the negative pole of the construct. The rule of triad method is illustrated in Figure 2. The experiment was concluded after repeating the aforementioned process until no new constructs emerged, with each participant’s duration ranging from 20 to 45 min. These constructs were categorized and statistically analyzed for frequency [64]. Finally, following RGT principles [65], 14 pairs of constructs were determined, representing the essential components of the stakeholder co-creation mechanism in Shanghai’s CSEE cases.

According to RGT data analysis rules, the 18 participants were then asked to rate the 14 pairs of constructs on a five-point Likert scale, where scores closer to 5 indicated an alignment with the positive construct and scores closer to 1 indicated an alignment with the negative construct. Data analysis was performed using Repertory Plus (V 2.0). The clustering and principal component analysis was conducted to derive the results of case typologies and the relationships between components and cases. Based on this analysis, a representative case from each type was selected, and the Content Validity Ratio (CVR) was calculated by a panel of five experts to ensure that the selected cases were representative of the three types [66].

(c)

Phase 3: Visualization and Comparative Analysis of Co-Creation Mechanism Patterns Using a Typical Case Analysis

In Phase 3, in-depth interviews were conducted with a total of 16 respondents from three representative cases to examine the co-creation relationships, inputs, outputs, implementation approaches of representative cases from each type. The co-creation mechanisms encompassing these aspects were identified through data coding and content analysis [61]. Following the coding phase, a panel of five external experts was invited to conduct a CVR [66] assessment of the coding results. Detailed information about the 16 respondents and 5 experts is also provided in Table 3. Subsequently, a comparison of selected cases was conducted, focusing on both the implementation initiatives under the 14 components of co-creation mechanism and the visualized action patterns. Finally, this analysis was further inquired to elucidate the commonalities and disparities observed among the cases, as well as the underlying causes that contributed to these variations.

4. Results

4.1. Co-Creation Mechanism Components and Case Typology Identification

In the RGT experiment, case descriptions were used as experimental elements to elicit participants’ perceptions of co-creation mechanisms in the selected cases. Constructs, representing participants’ interpretations of the relationships among these elements, were collected using a triadic method, resulting in 179 pairs of preliminary constructs. After semantic analysis, classification, and frequency statistics [64,65], the final 14 pairs of high-frequency constructs were identified, representing the essential components of the co-creation mechanism (as shown in

Table 4. The final constructs covering positive and negative descriptions.

Item	Components	Positive Construct (+)	Negative Construct (–)
A01	Co-creation goals (CCGs)	Clear objectives unify collaboration and resource sharing.	Unclear co-creation objectives lead to inconsistent participant collaboration and resource use.
A02	Open communication channels (OCCs)	Open and transparent communication channels foster trust and efficiency.	Unopen communication channels result in co-creation misunderstandings and a lack of trust.
A03	Resource integration and allocation (RIA)	Effective integration, circulation, and optimal allocation of resources support co-creation potential realization and quality.	Ineffective integration, circulation, and optimal allocation of resources hinder co-creation potential realization and quality.
A04	Stakeholder capability management (SCM)	Effective management of stakeholder capabilities ensures full utilization of expertise, enhancing co-creation professionalism and efficiency.	Poor management of stakeholder capabilities leads to an ineffective utilization of professional skills, reducing co-creation effectiveness.
A05	Participant empowerment strategies (PESs)	Effective empowerment strategies ignite participants’ innovation, potential, and engagement.	Ineffective empowerment strategies limit participants’ innovation, contribution, and engagement.
A06	Collaborative work mechanism (CWM)	A clear collaborative work mechanism fosters team cooperation and innovation efficiency.	The unclear collaborative work mechanism leads to poor cooperation and low innovation efficiency.
A07	Innovation incentive system (IIS)	Effective innovation incentive system stimulates participants’ enthusiasm and creativity.	Insufficient innovation incentive system hinders the participants’ enthusiasm and creativity.
A08	Coordination intermediary (CI)	An efficient coordination intermediary promotes information sharing, enhancing mutual trust and ensuring participants’ cultural and policy adaptation.	An inefficient coordination intermediary leads to poor information transmission, lack of trust, and barriers to cultural and policy adaptation.
A09	Benefit sharing mechanism (BSM)	A clear and fair mechanism for sharing benefits strengthens co-creation cohesion and willingness.	The absence or unreasonableness sharing benefits leads to internal conflicts and co-creation obstacles.
A10	Dynamic access and exit mechanism (DAEM)	A flexible access and exit mechanism maintain the vitality and adaptability of the ecosystem.	A rigid access and exit mechanism is not conducive to the vitality and adaptability of the ecosystem.
A11	Ongoing evaluation and feedback (OEF)	Effective evaluation and feedback mechanism on the co-creation process and results promote continuous improvement.	Ineffective evaluation and feedback mechanism means that co-creation problems remain unresolved for a long time.
A12	Community manager policy support (CMPS)	Managers provide clear and positive policy support, guiding and safeguarding co-creation activities, and enhancing the stability and predictability of the ecosystem.	Insufficient clear policy guidance or insufficient support from managers leads to uncertainty and potential risks in co-creation activities.
A13	Co-creation process management (CCPM)	Managers provide clear and positive policy support, safeguarding the stability of co-creation activities.	The lack of clear or insufficient policy support leads to uncertainty and potential risks co-creation activities.
A14	Co-creation education and training (CCET)	Effective pre-education and adaptive training enhances participants’ co-creation capabilities and efficiency.	Ineffective pre-education and adaptive training results in insufficient co-creation capabilities and efficiency of participants.

).

Subsequently, based on the rules for clustering associations between elements and constructs in RGT [67], a questionnaire was developed using the 14 final constructs, as shown in Supplementary Table S2. All 18 expert participants completed the questionnaire. To ensure data accuracy, the researchers conducted individual follow-ups with each participant to confirm their understanding of the relationship between questionnaire responses and case correspondences [67]. Additionally, to assess the questionnaire’s reliability, participants were asked to comment on the elements within the clustering results that showed a high degree of consensus (exceeding 80%). All participants easily interpreted the relationships among these elements, and their interpretations aligned with the results of the cluster and principal component analyses, indicating high validity and reliability of the questionnaire data.

Cluster analysis was performed based on the correlation between elements and constructs. Elements or constructs with a similarity exceeding 80% indicated a high level of coherence between adjacent rows (elements) and columns (constructs) [64]. As shown in Figure 3a, the resulting relationship tree visually delineates the grouping of experimental elements and constructs, reflecting consistency within each category of cases in terms of their co-creation mechanism components. As depicted in the figure, C1 and C3 represent Type 1, C2 and C6 represent Type 2, and C4 and C5 represent Type 3. Subsequently, principal component analysis (PCA) was employed as a dimensionality reduction technique to simplify data interpretation by linearly combining elements and constructs to elucidate their interrelationships. The PCA results, shown in Figure 3b, placed elements and constructs within a two-dimensional coordinate system based on their loadings on the two principal components. Elements and constructs positioned in the same quadrant can be interpreted as distinct clusters. The closer the distance between elements (cases), the more similar their scores in the principal component analysis, indicating a higher consistency in the constructs (components of the co-creation mechanism). The cumulative explanatory variance of the first and second principal components is 85.8%, exceeding the 80% threshold [67]. The PCA results align closely with those from the cluster analysis, categorizing all cases into three types.

As illustrated, Type 1, encompassing C1 and C3, is characterized by a well-defined collaborative work mechanism, effective intermediary involvement, and robust stakeholder capability management. These attributes are linked to their origins as university-led initiatives that leverage the community’s potential resources and benefits. The quality of co-creation is maintained through university-driven communication and a rigorous value-sharing system. Considering factors such as social impact, entrepreneurial activity, and prototype testing frequency, C1 was selected as a representative case for further analysis. Type 2, consisting of C2 and C6, shares key features in their co-creation mechanisms, notably effective participant empowerment strategies and methods for resource integration and allocation. These features are linked to their role as incubators for entrepreneurial projects that address the specific needs of aging communities. C2 was selected as the representative case for an in-depth analysis, based on social impact, industry exemplarity, and participant experience diversity. Type 3, comprising C4 and C5, exhibits similarities in co-creation mechanism characteristics, particularly in stakeholder capability management and co-creation education and training methods. These aspects align with their role in training and empowering entrepreneurs to access resources and expand their social impact. Given the diversity of entrepreneur types and the richness of data, C5 was chosen as the representative case for further analysis.

To ensure the representativeness of the selected case studies, a questionnaire was developed using the Content Validity Ratio (CVR) method, and five experts who participated in the RGT experiment were selected for evaluation. The questionnaire assessed aspects such as the case’s representativeness within its type, the innovativeness of the co-creation model, its leadership within similar ecosystems, and the data richness. The evaluation of the three representative cases across four dimensions of inquiry yielded a Scale Content Validity Index/Universal Agreement (S-CVI/UA) of 91.7%, demonstrating a high level of representativeness in the case selection.

4.2. Co-Creation Mechanism Pattern Comparison

Based on the three identified types, C1, C2, and C5 were selected as representative cases for in-depth interviews. The data collection process involved two main components: first, the co-creation mechanisms were analyzed through content analysis by coding stakeholder relationships, interactive actions, contributions, and collaboration outputs. Subsequently, the coding results were validated by five experts using the Content Validity Ratio (CVR) method. A total of 12 coding items across four dimensions were evaluated for each of the three cases to assess the accuracy, uniqueness, and timeliness of the co-creation actions. The S-CVI/UA result was 83.3%. Following this validation, a comparative analysis for three representative cases’ implementation initiatives under the framework of 14 co-creation mechanism components was performed, as summarized in

Table 5. The comparison for three cases’ implementation initiatives under 14 components of the co-creation mechanism.

	C1	C2	C5
A01 (CCG)	To create a 2035 community vision emphasizing “innovation, entrepreneurship, and creation”, focusing on future lifestyle clusters, entrepreneurs, designers, residents, and other resource providers collaboratively develop and test business prototypes. Iterative testing optimizes interactions with users and capital, establishing meaningful co-creation links.	Develop community-oriented entrepreneurial projects that align with local habits and needs, supporting young entrepreneurs in designing and testing business concepts. The co-creation goal emphasizes collaboration with residents, social media practitioners, external consumers, and social organizations for sustainable community development.	Project-based learning rooted in community issues drives resident engagement in social innovation, co-creating learning methods and tools. This approach fosters community governance, empowers facilitators, and incubates social entrepreneurship through an open source action network.
A02 (OCC)	Encourage open discussions and feedback through social media discussion groups established by the initiator team. Regular member meetings ensure project transparency and effective communication.	Conduct large-scale home visits and participatory research to ensure community needs and suggestions are heard and aligned with entrepreneurial products and capabilities. Regular meetings are led by the entrepreneur self-governance group to enhance efficiency.	During the development and testing of prototypes, organize reflective sessions with entrepreneurs and users. Participants in online communities are encouraged to provide feedback, which enhances the conversion process of entrepreneurial projects.
A03 (RIA)	As the leader, the university consolidates resources for prototype development and testing, including entrepreneurial mentors, co-design participants, resident insights and enthusiasm for product testing, and promotion channels from external users and local government.	The initiating social organization repurposes unused basement spaces, integrates management policies, and leverages social media promotion resources to create a platform for testing and promoting entrepreneurial projects.	The initiator’s long-term engagement in community social innovation-built trust, integrating real-life problems into learning issues and leveraging residents’ and entrepreneurs’ problem-solving skills as resources to incubate entrepreneurial projects focused on local social innovation.
A04 (SCM)	Recognizing participants’ prior project experiences and capabilities, entrepreneurs are categorized into thematic modules such as future food, learning, working, and housing. The initiator team cultivates and enhances their entrepreneurial capacities through regular concept co-design and joint decision making.	The initiator team cultivates target user social media groups, leveraging opinion leaders to enhance community engagement and sustain group activity through regular tasks and rewards, thereby creating a resource pool for future prototype design and testing.	Co-create project-based learning tasks with local residents based on real-life issues, transforming these into social entrepreneurial projects.
A05 (PES)	Empowering residents in prototype development and testing allows them, as active contributors, to influence the design and improvement of products, services, and business models, as well as entrepreneur evaluation and selection.	Empower young entrepreneurs and residents by establishing co-management teams for daily operations, decision making, and periodic evaluations. These teams hold regular meetings to set pricing guidelines and service regulations, ensuring parallel supervision.	Empower entrepreneurs to conduct research in communities. Grant open access rights to entrepreneurs in similar situations, allowing them to learn from action experiences through toolkits, while collecting market feedback and co-building social influence.
A06 (CWM)	Establish collaboration tasks among living labs based on unique knowledge resources, such as pairing future sound labs with digital media design studios to develop sound-based products, interactive sound art performances, and music-centered maker courses.	Divide co-design tasks with residents for space regeneration, including layouts, signage, and functionality. In daily operations, responsibilities are divided among entrepreneurs, with oversight and joint marketing efforts.	Assign tasks, such as learning task design, outcome evaluation, resident interaction, and experience documentation based on each entrepreneur’s capabilities. Additionally, residents are matched with entrepreneurs whose project goals align with the prototype tests.
A07 (IIS)	The lab cluster effect has created an innovative system image of mutual benefit, attracting more partners. Regular roadshow events present commercially viable prototypes, incorporating real-life feedback, to investors and government agencies, motivating innovation by connecting directly with resource providers.	For successful projects with positive social feedback, organize themed pop-up events and lifestyle festivals with social media platforms, like Xiaohongshu, featuring live-streaming and co-promotion with influencers and bloggers, to support entrepreneurs in expanding their influence and encouraging sustainable innovation.	Incentivizing ongoing innovation and participating motivation through the entrepreneurial project and personal growth record maps, project impact accounts, online community feedbacks, and project development rewards.
A08 (CI)	University designers act as coordinators and co-founders, offering consultations on prototype design, improvement, and test implementation, recruiting residents for testing, and assessing participants’ co-creation behavior and attitudes to plan incubation pathways.	The initiating team serves as a coordinator, overseeing community-themed co-creation, execution, resident feedback analysis, as well as project diagnostics and evaluations.	The initiating team acts as a coordinating intermediary, managing tasks, expanding impact, and facilitating knowledge sharing between residents, entrepreneurs, and community managers.
A09 (BSM)	Cooperation agreements clarify the rights and interests of all parties, especially regarding co-design testing activities and joint products, ensuring the sharing of external user attention and internal user data within the community.	Develop a community agreement and benefit-sharing system through extensive research and meetings conducted by the co-management team, focusing on social attention and revenue sharing.	Through the co-promotion with actors from online communities, enable entrepreneurs to share benefits like social attention and procurement opportunities of other community managers.
A10 (DAEM)	Develop entry mechanisms based on ecosystem distribution and alignment with future community lifestyles. Set exit mechanisms based on feedback from prototype testing activities, adjusting participants flexibly according to co-creation behaviors, attitudes, and contributions.	Implement entry mechanisms through community visits and local research, testing participant capabilities through trial operations. For exit mechanisms, eliminate and replace participants based on performance, co-creation quality, and user feedback.	Develop entry mechanisms based on the alignment between participants’ project plans and local development issues. For exit mechanisms, assess the incubation quality and completion based on the project growth cycle and feedbacks, and adjust the participants.
A11 (OEF)	Regularly organize assessment meetings that involve comprehensive considerations based on market feedback, such as social media shares, comments, new followers, and backend inquiry activity levels. Conduct horizontal comparisons and periodic reviews of the content richness, inclusivity, and attractiveness of lab-provided activities.	Conduct monthly open experience events to showcase and test products and services, reflecting on market reception and execution quality. Initiators refine participant selection criteria to align with local needs.	Track participants’ long term for their project operation. Develop growth maps for entrepreneurs and projects to periodically assessing and analyze their project problems and values.
A12 (CMPS)	Community managers provide space usage rights, promote resident engagement, and facilitate investor connections.	Community managers provide space access and co-develop entry policies to ensure projects align with community plans and activities meet community needs.	Community managers provide feedback channels for living problems and needs for co-create learning tasks transformation, and support to build platforms for resident recruitment and promotion.
A13 (CCPM)	Pre-plan regular co-creation of new products and services among labs while organizing flexible prototype co-creation and collaborative R&D based on holidays, celebrations, and emerging community needs, delivering co-creation outcomes for social promotion to investors and media partners.	Based on entrepreneurs’ strengths, establish and manage online “content circles” through social media chat groups to analyze user needs and feedback, assess market acceptance, and inform activity planning within these groups.	Manage the online database and toolkits of entrepreneurial action experience, facilitating internal and external knowledge flow.
A14 (CCET)	Organize co-learning and co-design workshops involving residents on topics such as community fitness, arts, healthy dining, and skill acquisition, aiming to cultivate residents’ co-creation awareness, willingness, and capabilities.	Regularly organize community meetings and events to provide co-creation training, enhancing the capabilities of residents and entrepreneurs to engage in community building.	Based on local needs, conduct solution discussions and workshops to facilitate both residents and entrepreneurs in becoming familiar with and understanding the co-creation process and targets.

. The visualized mechanism patterns, including stakeholder relationships, co-creation interactions, and outcomes, are visualized in Figure 4, Figure 5, and Figure 6, representing C1, C2, and C5, respectively.

Figure 4 illustrates that C1 represents a co-creation model driven by experiments in future lifestyle scenarios. This model was implemented through pre-designed community lifestyle modules, followed by the involvement of key anchor entrepreneurs to establish living labs. These labs facilitated interactions through experiments, exhibitions, roadshows, and pop-up events aimed at collaboratively developing and testing business prototypes. In this model, the university served as both a leader and designer, orchestrating co-creation efforts. Initially, it worked with participants to co-plan the operational framework for the labs. During the collaborative design phase, the labs engaged in periodic interactions within content module groups and flexibly adapted to cross-group collaborations in response to market trends and emerging demands. In the prototype testing phase, participant feedback and brand support were encouraged, resulting in the co-branding and market launch of the prototypes.

C2 represents a co-creation model driven by the needs of the local community. This model involved extensive community visits and surveys to identify and co-plan local needs with community managers and selected entrepreneurs for pilot implementation. During the prototype development phase, core services and business models were collaboratively designed to address various local needs, with residents’ experiences and demands serving as key co-creation drivers. These needs encompassed areas such as dining, entertainment, fitness, and the arts. During the testing phase, small-scale community businesses and retail ventures were utilized to trial the entrepreneurial prototypes in real-life settings. Regular assessments of operational quality were conducted by a co-governance group of entrepreneurs, and collaborative promotion was carried out with industry leaders and influencers, such as self-media bloggers, to leverage advantages in similar communities.

C5 represents a co-creation model aimed at addressing local community development challenges. This involved gaining deep insights into community development issues, which were then transformed into project-based learning themes. By empowering facilitators of project-based learning, who shared their experiences from social entrepreneurship prototype experiments across various contexts, the collaborative refinement of practices with other entrepreneurs was facilitated. This led to the development of a co-constructed and co-delivered body of knowledge. The initiator of this model is a social enterprise dedicated to the education, empowerment, and incubation of social entrepreneurs. It served as a co-creation coordinator, documenting the growth and experiences of entrepreneurs and translating these into entrepreneurial guidance, thus supporting the collaborative expansion of influence.

4.3. The Shaping Factors of Commonalities and Differences in Representative Cases

The commonalities observed in the co-creation mechanisms of the three representative cases reveal several key patterns. Each case demonstrates a well-defined co-creation objective, with project initiators acting as coordination intermediaries to manage stakeholder capabilities and allocate resources. Furthermore, all three cases emphasize the strategic use of community manager policy support and adapt co-creation process management methods to the capabilities and expectations of participating entrepreneurs. A notable commonality is the initiators’ deep understanding of and empathy for the community’s pressing issues and social conflicts. This, combined with a nuanced understanding of the community’s social dynamics and a strategic interpretation of local development plans, contributes to effective co-creation patterns that are finely tuned to the community’s needs. Moreover, all three projects focus on integrating, classifying, and sustainably cultivating stakeholder capabilities, particularly in online and offline community engagement, which supports user recruitment, business prototype promotion, and insight generation into emerging opportunities.

The differences in co-creation mechanisms among the cases arise from variations in resource sources, the relationship between CSEEs and the located community, and stakeholder screening perspectives. A primary factor is the nature of the resource sources of CSEE initiators, which influences strategies for resource integration in co-creation, particularly in activities involving prototype development and influence expansion. Specifically, C1 and C5 are spearheaded by universities and knowledge-intensive social enterprises, both with extensive experience in social learning. As a result, they can directly guide the development and refinement of entrepreneurial prototypes or allocate knowledge resources based on their educational and training capabilities, thereby supporting the co-creation of business opportunities aligned with participants’ strengths. Their direct access to knowledge resources ensures the advancement of co-creation, facilitates the rapid resolution of developmental and iterative challenges in prototypes, and provides sustainable support for market entry and advantage replication. In contrast, C2 is initiated by a social organization dedicated to community sustainable development, with resources derived more indirectly from the trust established with community managers and collaboration with self-media bloggers from prior regeneration projects. Consequently, its co-creation pattern focuses on providing channels for entrepreneurs to expand their social influence and emphasizes empowering participants to actively engage and develop a sense of belonging in the ecosystem. This approach fosters spontaneous and stable motivation and capability for co-creation.

Secondly, the relationship between CSEEs and the host community varies. In C1, this relationship exemplifies symbiotic development, with CSEEs transforming the community from passive recipients of solutions into active engines of endogenous innovation. This mutually beneficial relationship influences co-creation process management by assessing stakeholders’ complementary strengths within the framework of envisioned future lifestyle modules and their ability to attract additional social attention and investment resources. In C2, the relationship is characterized by interdependent development, focusing on enhancing the adaptability of entrepreneurial projects to local needs through systematic design and implementation. Entrepreneurs in this model enter the community as solution providers, improving social action capabilities and gaining access to channels for expansion into other communities with similar social environments. Consequently, co-creation process management emphasizes adaptability to community needs, residents’ behaviors, and living patterns. In C5, the relationship is one of embedded development, where entrepreneurs engage in project-based learning around local community issues, serving as a supportive context for incubating social entrepreneurship projects. Entrepreneurs must possess a deep understanding of the community’s social structure, resident composition, and systemic issues, along with strong empathy and the ability to interpret policies in order to develop social entrepreneurial projects. By leveraging the community’s long-term experience, robust online experience sharing, and toolkit data, C5 established an efficient flow of social entrepreneurial knowledge, facilitated by growth records and a repository of action experiences.

Finally, the perspective on participant screening differs, leading to variations in the content of co-creation outputs. C1 evaluates participants based on their potential to create new products, services, and value, emphasizing complementary capabilities and the product potential of entrepreneurs within future lifestyle frameworks. C2 focuses on aligning entrepreneurial projects with community needs and the frequency of business prototype trials. Both C1 and C2 are highly inclusive regarding the categories of business prototypes. In contrast, C5 emphasizes the social entrepreneurship attributes and the operational efficiency of participants. It assesses participants based on their values, engagement duration, and project alignment, evaluating their motivation to participate in co-creation, ability to independently execute projects in other communities, and capacity to document and share project growth. This approach ensures that the final outputs of social entrepreneurship align with community development goals and collective social learning. However, the diversity of co-creation outputs in C5 is relatively limited.

5. Discussion

As an emerging urban innovation mechanism, CSEEs shift residential communities from a peripheral role in the urban innovation chain to a central position, actively attracting resources and incubating entrepreneurial projects. This study investigates the characteristics and patterns of multi-stakeholder co-creation mechanism within the context of Shanghai’s CSEEs. While focused on Shanghai, the principles discussed here are applicable to entrepreneurial ecosystems in other regions with similar socio-economic and community support structures. For instance, cities like Beijing, Shenzhen, and other high-density urban areas exhibit comparable characteristics in resource integration, stakeholder collaboration, and innovation facilitation. Therefore, this theoretical model is also applicable to entrepreneurial ecosystems in high-density urban communities in other regions.

Using in-depth interviews and the RGT, the study examines essential components of the co-creation mechanism, case typologies, and dynamic co-creation implementation patterns. Consistent with prior research, the findings indicate that co-creation activities extend beyond the outputs of commercial and social entrepreneurship projects to encompass resource integration and improvements in local welfare within communities [9,16]. Additionally, in line with the existing studies, it was found that clear objectives, open communication, efficient intermediaries, and continuous evaluation mechanisms are crucial for supporting entrepreneurial incubation driven by diverse co-creation efforts [23,54].

Contrary to previous views that position entrepreneurs as the central actors [68], this study highlights their role within the CSEE co-creation mechanism as part of a collaborative group. Traditionally, the research on entrepreneurial co-creation has portrayed entrepreneurs as the primary drivers of the entrepreneurial process [45]. This study redefines entrepreneurs’ roles within the CSEE co-creation mechanisms as peer participants alongside other stakeholders. External influencers, such as social media bloggers and entrepreneurial mentors, along with internal community residents and managers, play significant roles in providing entrepreneurial insights, creation, and development. This shift from viewing entrepreneurs as the sole drivers underscores that co-creation within the entrepreneurial ecosystem results from multi-stakeholder collaboration rather than individual actions. Entrepreneurs are seen as collaborators and resource integrators, relying on a broad network of knowledge and resources within the community.

Additionally, this study challenges the characterization of co-creation intermediaries as merely neutral regulators or basic communication platforms [69]. It highlights the pivotal roles of universities, embedded social organizations, and social enterprises in business prototype design, impact enhancement, and resource allocation. These intermediaries serve as crucial connectors and resource providers, whose value extends beyond merely linking parties and communicating their needs and capabilities [70]. They also play significant roles as leaders and implementers, facilitating dynamic stakeholder interactions that advance business prototypes toward market implementation. Consequently, these intermediaries not only influence co-creation outcomes but also engage directly in resource provision and utilization.

Furthermore, previous studies have often focused on the value analysis of co-creation inputs, like financial or technological investments, and the outputs, such as services, competitiveness, and network expansion, from a static perspective [71]. However, there has been insufficient analysis of the dynamic mechanisms of co-creation and their relationships with outputs. This lack of a comprehensive perspective hinders effective resource mobilization and the balancing of participant motivations and benefits [43]. This study employs visualized models of the co-creation process and mechanisms to provide a systematic view, illustrating how the various components of co-creation activities interact and evolve over time to ultimately create value.

Methodologically, in contrast to previous studies that primarily focus on single-case analyses of experiences [72,73], this study examines multiple case typologies to explore how co-creation quality is ensured through the enhancement and management of stakeholder capabilities, formulation of diverse resource allocation strategies, and implementation of co-creation evaluation and collaborative mechanisms. By visualizing the flow of knowledge, funding, materials, and policy resources among participants in various cases, this study enhances the understanding of co-creation diversity and variance in Shanghai’s CSEE cases. It provides insights for actors in similar contexts to review their co-creation activities based on these visual patterns and the correlations between inputs and final outputs observed in case experiences.

6. Conclusions

The theoretical model proposed in this study offers actionable insights and guidance for high-density urban CSEE practices. By comparing different case experiences, the study enhances the efficiency and quality of co-creation management by considering diverse stakeholder characteristics and resource availability, while mitigating social risks, such as context exclusion, resource waste, and trust issues among residents caused by a simplistic replication of successful models. The key contributions of this study are summarized as follows:

(1)

Implementation process of co-creation mechanisms: This research highlights that successful implementation relies on explicit and implicit resource connections, contextual adaptations, and the strategic engagement of social media channels. These channels not only support prototype testing and promotion, but also uncover new entrepreneurial opportunities, enhancing the scalability of co-creation practices.

(2)

Variation in co-creation mechanisms: This study demonstrates that intermediaries such as universities, social organizations, and social enterprises introduce varying degrees of differences in co-creation goals, process management, and evaluation methods. These variations are shaped by resource origins, entrepreneurial positioning, and participant selection, which in turn influence entrepreneurial outputs, such as user targeting, prototype conversion efficiency, and capital interactions.

(3)

Community benefits and local role development: Effective CSEE co-creation mechanisms not only address small-scale community needs, but also contribute significantly to enhancing community vitality and attracting social investment. The findings also emphasize the importance of empowering local stakeholders to ensure the sustainability of innovation after the departure of external advisory teams.

The theoretical contribution of this study addresses the knowledge gap related to the diversity and typology of co-creation mechanisms within Shanghai’s CSEEs. By proposing a replicable mechanism composition and implementation paradigm, and correlating these with co-creation outputs and influencing factors, this study broadens the application scope of the value co-creation theory under an entrepreneurial target. From a practical perspective, this study provides valuable references for co-creation management and improvement in CSEEs, and offers empirical data to support decision making regarding co-creation regulations and participant selection. Socially, this research helps unlock the latent resources within residential communities in Shanghai and other high-density cities, promoting sustainable development through proactive social entrepreneurship and knowledge spillover effects.

However, co-creation activities within entrepreneurial ecosystems are inherently complex. This study acknowledges certain limitations, including a small sample size and a limited observation period. Future research could expand the case base, conduct longitudinal analyses across diverse urban contexts, and integrate social network data and expert evaluations to provide more robust and generalizable evidence for both academics and practitioners.