Business Ecosystem Embeddedness and SRDI SME Growth: The Roles of Organizational Resilience and Digital Leadership

Highlights

Please indicate how your work links to systems science via your contributions to systems practice, theory, and/or methodology.

• This study adopts a systems perspective by viewing SRDI SME growth as an outcome of interactions within a business ecosystem characterized by interdependence, complementarity, and coordinated value creation among multiple actors.
• It contributes to systems theory by explaining the cross-level process through which business ecosystem embeddedness develops firm-level proactive and reactive organizational resilience, and how digital leadership shapes this process under environmental turbulence.

What are the main findings and/or the implications of the main findings?

• Business ecosystem embeddedness is positively associated with SRDI SME growth, while proactive and reactive organizational resilience partially mediate this relationship at different stages of environmental disruption.
• Higher digital leadership strengthens the positive effects of business ecosystem embeddedness on both proactive and reactive organizational resilience, thereby increasing its indirect effects on SRDI SME growth through both resilience pathways.

Abstract

Against the backdrop of intensifying geopolitical conflicts and deglobalization, how Specialized, Refined, Differentiated, and Innovative small and medium-sized enterprises (SRDI SMEs) can overcome resource constraints and achieve sustained growth has become an important research issue. Although prior research has recognized that business ecosystem embeddedness provides SMEs with access to external resources and growth opportunities, the mechanisms through which such embeddedness translates into the growth of SRDI SMEs in turbulent environments remain insufficiently understood. Drawing on resource orchestration theory, this study develops a theoretical model linking business ecosystem embeddedness to SRDI SME growth and examines the mediating roles of proactive organizational resilience and reactive organizational resilience, as well as the moderating role of digital leadership. Using two-wave survey data from 367 Chinese SRDI SMEs, this study tests the proposed hypotheses through regression analysis and bootstrapping. The results show that business ecosystem embeddedness is significantly and positively associated with SRDI SME growth and that proactive organizational resilience and reactive organizational resilience each partially mediate this relationship. Moreover, digital leadership strengthens the relationships between business ecosystem embeddedness and both dimensions of organizational resilience, thereby strengthening its indirect effects on SRDI SME growth through proactive organizational resilience and reactive organizational resilience. The findings provide theoretical insights and practical implications for SRDI SMEs seeking sustained growth in turbulent environments.

1. Introduction

Against the backdrop of intensifying geopolitical conflicts and deglobalization, firms face growing uncertainty in both supply and demand, while risks such as supply chain disruptions and fluctuations in market demand have become increasingly pronounced [1]. Specialized, Refined, Differentiated, and Innovative (SRDI) enterprises have long focused on specific technological fields and niche markets, building competitive advantages through continuous innovation. These enterprises play important roles in supplying critical components and intermediate goods and in supporting complementary activities across industrial chains, making them key contributors to supply chain resilience [2,3]. However, compared with large enterprises, SRDI SMEs generally face structural constraints arising from their limited resource bases and narrow market scope [4]. Their strong specialization creates a distinctive growth tension: sustaining competitive advantage requires concentrating scarce resources in narrow technological domains, whereas continued growth increasingly depends on accessing complementary knowledge, technologies, and market resources beyond firm boundaries. When technological trajectories shift rapidly, demand from key customers fluctuates, and the supply of critical resources is disrupted, SRDI enterprises are particularly vulnerable because of rigid resource structures, insufficient slack resources, and strong path dependence [5,6]. Such vulnerability may, in turn, constrain their survival and growth. Accordingly, a critical question is how SRDI SMEs can preserve the advantages of specialization while mobilizing and deploying complementary external resources to strengthen their resilience and sustain growth in turbulent environments.

Traditional theories of firm growth have mainly focused on internal drivers of growth [7]. Building on this view, subsequent studies have examined firm growth primarily in terms of internal factors, including strategic orientation, managerial characteristics, and innovation activities [8,9,10]. However, under persistent environmental turbulence, the accumulation of internal resources alone is no longer sufficient to support the sustained growth of SRDI SMEs. Environmental uncertainty, blurred organizational boundaries, and technological change have increased firms’ reliance on business ecosystems to access external resources and participate in value co-creation [11,12]. A business ecosystem is a value co-creation network composed of focal firms, suppliers, customers, complementors, and other stakeholders [11]. It provides firms with access to external knowledge, technologies, markets, and complementary resources. Unlike embeddedness in a single network, ecosystem embeddedness involves simultaneous participation in multiple networks that are larger, more complex, and functionally distinct [13]. This broader form of embeddedness allows firms to reach beyond organizational boundaries, access diverse resources, obtain complementary support, and co-create value with multiple actors [14,15]. However, the resource-access advantage created by ecosystem embeddedness does not automatically translate into firm growth. External resources may remain fragmented, underutilized, or poorly matched with organizational needs unless firms can effectively structure, bundle, and leverage them [16]. Thus, the unresolved issue is not simply whether ecosystem embeddedness gives SRDI SMEs access to valuable resources, but how those externally accessed resources are converted into organizational capabilities that support sustained growth. This resource-conversion problem is particularly salient for SRDI SMEs because their limited internal resource bases increase the value of ecosystem resources while simultaneously constraining their ability to absorb and deploy them. Accordingly, explaining the process through which business ecosystem embeddedness is transformed into SRDI SME growth constitutes the central theoretical problem addressed in this study.

According to resource orchestration theory, a firm’s competitive advantage depends not only on the resources it possesses but also on how it manages those resources over time [16,17]. This theory provides a useful lens for explaining how business ecosystem embeddedness contributes to the growth of SRDI SMEs. Specifically, business ecosystem embeddedness gives SRDI SMEs access to diverse external resources. However, these resources can support firm growth only when they are identified, selected, integrated, and reconfigured into organizational capabilities. Organizational resilience is one of the key capabilities through which this resource transformation occurs [18]. Organizational resilience refers to a firm’s ability to identify risks and respond quickly in turbulent environments. It enables firms to maintain stability while adapting to change and to capture emerging growth opportunities [19]. Rather than treating organizational resilience as a unitary capability, this study distinguishes between proactive and reactive organizational resilience, which reflect firms’ capabilities to cope with environmental disruptions at different stages. From a resource orchestration perspective, these two dimensions can emerge from different resource-bundling processes through which firms combine ecosystem resources with their internal resources. On the one hand, business ecosystem embeddedness broadens firms’ sources of information, improves their ability to sense environmental changes and interpret emerging trends, and strengthens risk anticipation and preparedness [20]. This allows firms to adjust their strategic direction and market positioning in advance, thereby reducing losses caused by resource misallocation. On the other hand, business ecosystem embeddedness facilitates resource sharing and coordination between firms and their ecosystem partners, thereby improving resource access and strengthening firms’ ability to adapt and respond during crises [21]. This enables firms to recover more quickly from disruptions and sustain their growth. Therefore, this study examines the mediating roles of proactive organizational resilience and reactive organizational resilience in the relationship between business ecosystem embeddedness and SRDI SME growth.

Moreover, whether business ecosystem embeddedness translates into organizational resilience also depends on a firm’s internal management and coordination capabilities. Prior research suggests that resource management requires managerial involvement and that top managers play an important role in identifying, allocating, and deploying resources [22]. This managerial role becomes particularly important in business ecosystems because deeper embeddedness increases not only the availability of external resources but also their heterogeneity, dispersion, and coordination complexity. In the digital economy, digital leadership refers to leaders’ ability to develop a digital vision, promote the adoption of digital technologies, and facilitate collaboration across organizational boundaries [23]. For SRDI SMEs, strong digital leadership can build internal consensus around a clear digital vision [24], enabling firms to identify and select relevant ecosystem resources more effectively. It can also encourage firms to use tools such as data analytics and digital platforms [25], thereby improving environmental scanning, risk identification, and the speed of decision making. Digital leadership can further strengthen digital connections between firms and their stakeholders [23], which improves knowledge sharing and the efficiency of coordinated responses. From a resource orchestration perspective, these functions make digital leadership more than a general leadership capability: it represents an orchestration-enabling condition that helps firms filter ecosystem information, align external resources with internal needs, and coordinate their deployment across organizational boundaries. Accordingly, this study examines digital leadership as a managerial boundary condition that shapes how effectively the resource-access advantages of ecosystem embeddedness are converted into proactive and reactive organizational resilience and, subsequently, into firm growth.

In summary, drawing on resource orchestration theory, this study develops a theoretical model linking business ecosystem embeddedness, organizational resilience, digital leadership, and SRDI SME growth. It examines the mechanisms and boundary conditions through which business ecosystem embeddedness affects the growth of SRDI SMEs. This study makes three main theoretical contributions. First, it extends research on SRDI SME growth from an internally oriented perspective to an ecosystem perspective. By focusing on resource-constrained SRDI SMEs, the study highlights how relationships with ecosystem actors can complement firms’ limited internal resource bases and support sustained growth. Second, by distinguishing proactive from reactive organizational resilience, this study explains how business ecosystem embeddedness is linked to firm growth through two resilience capabilities that perform different temporal and functional roles across stages of environmental disruption. This provides a more fine-grained understanding of the mediating role of organizational resilience in the relationship between ecosystem embeddedness and firm growth. Third, the study introduces digital leadership as an important internal contextual condition in the digital era and examines how it shapes the relationships between business ecosystem embeddedness and both forms of organizational resilience. In doing so, it further specifies an important internal managerial boundary condition of resource orchestration theory. This study provides theoretical guidance for SRDI SMEs seeking sustained growth in increasingly ecosystem-based and digitalized environments.

2. Theoretical Background and Hypothesis Development

2.1. Resource Orchestration Theory

Resource orchestration theory (ROT) extends the resource-based view by explaining how firms manage resources to create value. The traditional resource-based view holds that valuable, rare, inimitable, and non-substitutable resources are an important source of sustained competitive advantage [26]. Resource orchestration theory further argues that resource possession alone does not necessarily generate competitive advantage. Firms must create value by structuring their resource portfolios, bundling resources into capabilities, and leveraging those capabilities in the market [16]. Sirmon et al. (2011) further emphasize that firms need to adjust the configuration of internal and external resources across different stages of development and environmental conditions to sustain competitive advantage [17]. ROT is particularly appropriate for the present study because it provides a more process-specific explanation of resource management than the RBV while addressing a different theoretical question from Dynamic capabilities theory (DCT). Whereas the RBV primarily explains why particular resource endowments can underpin competitive advantage, and DCT emphasizes firms’ higher-order capacities to sense, seize, and reconfigure in changing environments, ROT focuses more explicitly on the managerial actions through which resource portfolios are structured, resources are bundled to build capabilities, and those capabilities are leveraged to create value [16,17]. Accordingly, the central argument of ROT is that competitive advantage depends not simply on a firm’s stock of resources, but on managers’ ability to configure and coordinate heterogeneous resources so that they can be transformed into usable organizational capabilities and subsequently deployed for value creation. Recent applications of ROT further suggest that such resource orchestration need not be confined to resources located within firm boundaries; resource-constrained firms may rely on external stakeholders and ecosystem relationships to access and combine complementary resources [27].

Drawing on resource orchestration theory, this study develops an analytical framework of “resource structuring–capability building–value realization.” Within this framework, business ecosystem embeddedness provides SRDI SMEs with access to external resources and therefore serves as a basis for resource structuring. Organizational resilience represents the capability developed by integrating internal and external resources to identify risks and respond rapidly, and thus reflects the outcome of resource bundling. SRDI SME growth, in turn, captures the value created when firms leverage these capabilities to sustain their operations and continue creating value. ROT also emphasizes the central role of managers in identifying, integrating, coordinating, and deploying resources [22]. Their digital leadership is therefore an important internal factor that can improve the efficiency with which firms identify, integrate, and deploy resources. Accordingly, this study develops a theoretical model linking business ecosystem embeddedness, organizational resilience, and SRDI SME growth, and examines the moderating role of digital leadership in the resource orchestration process.

2.2. Business Ecosystem Embeddedness

Business ecosystem embeddedness is an important concept for understanding how firms access resources across organizational boundaries and engage in value co-creation. Moore (1993) introduced the ecosystem concept into strategic management and described a business ecosystem as a community of economic actors that supports new product development, responds to customer needs, and promotes continuous innovation [11]. Iansiti and Levien (2004) further argued that business ecosystems consist of many loosely connected but interdependent participants whose survival and performance are closely linked [28]. From a structural perspective, Adner (2017) defined an ecosystem as an alignment structure among multiple partners organized around a focal value proposition [12]. Its defining feature is not simply the presence of interfirm ties, but the systematic alignment of activities, actors, positions, and links required to realize that value proposition. Jacobides et al. (2018) further emphasized that ecosystem actors exhibit varying degrees of multilateral and non-generic complementarity and are not subject to the hierarchical control of a single actor [29]. Taken together, these perspectives highlight interdependence, complementarity, and coordinated value creation as defining characteristics of business ecosystems and provide the conceptual basis for understanding firms’ embeddedness within them.

Embeddedness theory suggests that firms can gain economic opportunities from external relationships that cannot be easily obtained through market transactions or vertical integration. Such embedded relationships are characterized by trust, knowledge exchange, and joint problem solving [30,31]. Drawing on Riquelme-Medina et al. (2022, 2023) [14,32], this study defines business ecosystem embeddedness as the degree to which a firm is integrated into and becomes part of a business ecosystem. It comprises three dimensions: interdependence, value potential, and shared components. Interdependence indicates that the relationships between a firm and other ecosystem members extend beyond direct transactions and include both formal contractual ties and informal connections. Value potential refers to the complementary collaboration of ecosystem members around a shared value proposition, through which they pursue their own goals while contributing to value creation within the ecosystem. Shared components provide an important basis for ecosystem embeddedness. They include tangible elements, such as platforms, standards, data, and infrastructure, as well as intangible resources, such as a shared vision and collective goals that extend beyond individual interests. Thus, business ecosystem embeddedness reflects not only the extent of a firm’s connections with external actors, but also the depth of its participation in multi-actor collaboration and value co-creation.

2.3. Business Ecosystem Embeddedness and Firm Growth

This study proposes that business ecosystem embeddedness is positively related to SRDI SME growth. SRDI SMEs often face constraints arising from limited resource bases, insufficient market access, and difficulties in commercializing their innovations [4]. Business ecosystem embeddedness can alleviate these constraints through interdependence, value potential, and shared components, thereby creating favorable conditions for firm growth. First, interdependence enables SRDI SMEs embedded in business ecosystems to establish relatively stable relationships with focal firms, customers, suppliers, and complementors [33]. These relationships broaden firms’ access to resources and reduce the difficulty and cost of resource search, thereby providing greater resource availability and relational stability for growth [34]. Second, value creation within an ecosystem does not occur through a sequential process along a single value chain. Instead, it results from collaboration among multiple actors around a shared value proposition. Although SRDI SMEs rarely occupy dominant positions within an ecosystem, their specialized products, technological modules, or supporting services can become important components of the ecosystem’s value proposition. Once integrated into this value proposition, these offerings can reach a wider range of applications by drawing on the ecosystem’s collective market reach. For SRDI SMEs operating in niche positions, growth often occurs when their complementary value is recognized, integrated, and amplified within the ecosystem [35]. This value-amplification effect can strengthen customer retention and expand market coverage, thereby supporting firm growth. Finally, shared components, including platforms, technological standards, data resources, and infrastructure, can reduce the costs of resource search and matching among ecosystem partners. They also enable SRDI SMEs to accelerate technological innovation and product validation [36,37]. By using these shared components, SRDI SMEs can avoid duplicated investments, concentrate more resources on their specialized technological fields, and accelerate the commercialization of their innovations, thereby improving the efficiency with which limited resources are converted into growth opportunities. Taken together, greater business ecosystem embeddedness can expand SRDI SMEs’ access to complementary resources, enhance the market value of their specialized offerings, and reduce the costs of innovation and commercialization. These benefits provide resource and market conditions that are conducive to firm growth. Accordingly, this study proposes the following hypothesis:

Hypothesis 1.

Business ecosystem embeddedness is positively related to SRDI SME growth.

2.4. The Mediating Effect of Organizational Resilience

Organizational resilience is a key capability that enables firms to identify risks, absorb shocks, recover, and adapt in rapidly changing environments [38]. It provides an important foundation for firm survival and sustained growth [19]. More broadly, organizational resilience is not limited to crisis situations but also reflects an organization’s ongoing capacity to adapt to change [39]. Prior research views organizational resilience as a multidimensional construct that includes risk anticipation, preparedness, crisis response, recovery, and adaptation [40,41]. In line with the context of this study and drawing on Jia et al. (2020) [42] and Lin and Fan (2024) [41], we distinguish between proactive organizational resilience and reactive organizational resilience. Proactive organizational resilience refers to a firm’s ability to identify potential risks, anticipate their possible consequences, and prepare for them before a crisis fully unfolds [43]. Reactive organizational resilience refers to a firm’s ability to interpret unfolding events, mobilize resources, and restore operations during and after a crisis [42]. The distinction between these two forms of resilience is primarily temporal: proactive resilience emphasizes anticipation and preparedness before disruption, whereas reactive resilience emphasizes adaptation and recovery during and after disruption. Drawing on resource orchestration theory, this study argues that business ecosystem embeddedness does not lead directly to firm growth. Instead, it contributes to the development of organizational resilience through different resource-bundling processes, which in turn support firm growth.

First, business ecosystem embeddedness can strengthen proactive organizational resilience through an “information–sensemaking–preparedness” resource-bundling process. Because of their limited internal resources and information, SRDI SMEs often struggle to detect changes in technology and market demand in a timely manner. Through sustained interaction with multiple ecosystem actors, embedded firms can gain access to more diverse market information and knowledge of emerging technologies. This improves their sensitivity to environmental change and their ability to anticipate it [44,45]. Compared with conventional business networks, business ecosystems are characterized by multilateral complementarities and joint value creation [32]. As a result, the information provided by different ecosystem actors is often diverse and complementary. By comparing information from multiple ecosystem actors, firms can more accurately identify shifts in market demand, trends in technological substitution, and potential supply risks. This reduces judgment errors caused by reliance on a single source of information. On this basis, firms can integrate external signals into their internal data analysis, strategic planning, and contingency planning. This helps them develop stronger capabilities for environmental scanning, early risk warning, and advance preparation, thereby strengthening proactive organizational resilience.

Proactive organizational resilience further supports the growth of SRDI SMEs. Its role extends beyond risk reduction by enabling firms to allocate organizational resources in advance and clarify the direction of resource deployment. Firms with stronger proactive organizational resilience can detect changes in customer demand and trends in technological innovation at an early stage [46]. They can then adjust their R&D priorities and market positioning in advance, thereby improving the alignment between resource commitments and market opportunities. For resource-constrained SRDI SMEs, growth depends critically on avoiding the allocation of scarce resources to inappropriate technological paths, low-potential markets, or low-value customers. Proactive organizational resilience enables firms to build resource buffers, develop response plans, and adjust their strategic paths before external shocks fully materialize. These actions reduce the risks of misdirected R&D, market misjudgment, and resource misallocation [47,48]. Therefore, proactive organizational resilience converts the informational advantages gained through business ecosystem embeddedness into stronger environmental sensing and anticipation capabilities, which in turn support sustained firm growth. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 2a.

Proactive organizational resilience mediates the relationship between business ecosystem embeddedness and SRDI SME growth.

Business ecosystem embeddedness can also strengthen reactive organizational resilience by facilitating access to complementary resources and cross-organizational coordination during disruptions. Close relationships with ecosystem partners enable firms to obtain critical information quickly during a crisis, thereby reducing delays in action caused by information asymmetry [49]. Business ecosystem embeddedness also gives firms access to a pool of external resources needed for crisis response. When existing supply, technology, or market channels are disrupted, firms can rely on ecosystem partners to identify alternative sources of supply, obtain technical support, share platform resources, or adjust their market channels. The complementarity among ecosystem partners increases the variety of resources available for crisis response and facilitates the transfer of knowledge across organizational boundaries [50,51]. Shared understandings and established relationships within the ecosystem can further improve firms’ ability to interpret external knowledge and coordinate external resources with immediate internal needs [41]. When these external resources are rapidly integrated and mobilized alongside existing organizational resources, firms are better positioned to adjust operations and restore critical activities. Thus, whereas proactive resilience is supported primarily by information and knowledge that enable anticipation and preparedness, reactive resilience depends more strongly on timely access to, coordination of, and mobilization of complementary resources after disruption occurs.

Reactive organizational resilience further supports SRDI SME growth in two main ways: by sustaining growth during disruptions and facilitating post-crisis adjustment. For SRDI SMEs, external shocks that cause production stoppages, delivery delays, or customer loss can quickly weaken cash flow, customer trust, and market opportunities, thereby impeding firm growth. Firms with stronger reactive organizational resilience can reallocate resources, adjust operational processes, and restore critical business activities more quickly during a crisis. These actions reduce the adverse effects of operational disruptions on sales revenue and market share. Moreover, reactive organizational resilience encourages firms to reassess their technological weaknesses and process deficiencies during a crisis and learn from these experiences. Such learning can support the development of new products and services [47]. Thus, reactive organizational resilience not only enables firms to bounce back and recover in uncertain environments but also helps them bounce forward by adapting and improving beyond their pre-crisis state [52]. This, in turn, allows firms to achieve growth despite adverse conditions. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 2b.

Reactive organizational resilience mediates the relationship between business ecosystem embeddedness and SRDI SME growth.

2.5. The Moderating Effect of Digital Leadership

Digital leadership is an important condition for firm transformation and growth in the digital era [53]. It reflects leaders’ ability to develop a clear digital vision, promote the use of advanced digital technologies, and guide organizational members in adapting to and jointly advancing digital transformation [23,24]. From a resource orchestration perspective, business ecosystem embeddedness gives SRDI SMEs access to a broad base of external resources. However, as the number of resource sources increases, these resources also become more heterogeneous, dispersed, and complex. Digital leadership helps firms manage this complexity by improving the efficiency with which they identify, integrate, and deploy these resources. A clear digital vision provides strategic criteria for resource selection, while digital technologies and digitally enabled coordination facilitate the integration of ecosystem-derived resources with firms’ existing resource bases. Accordingly, this study conceptualizes digital leadership as an internal managerial boundary condition that shapes the extent to which business ecosystem embeddedness is associated with the formation of proactive and reactive organizational resilience.

Digital leadership can strengthen the positive effect of business ecosystem embeddedness on proactive organizational resilience. Business ecosystem embeddedness enables firms to obtain information from multiple external actors. However, information from different actors may be fragmented, complex, or even conflicting. Without clear goals and strategic direction, access to a broader range of resources may lead to information overload. This makes it more difficult for firms to distinguish strategically relevant signals from less valuable information. Firms with stronger digital leadership have a clear digital vision that provides direction and criteria for resource identification. This helps them more accurately identify resources related to emerging technologies and market needs [24]. Leaders can also promote the use of digital tools, such as big data analytics and digital platforms, to analyze and compare ecosystem resources. These tools help firms assess the fit between the value of a resource and their organizational needs more accurately [23,25,54], allowing them to prioritize valuable resources and incorporate relevant external information into internal planning and decision making. Digital leadership also promotes cross-functional information sharing and coordination [23], enabling firms to integrate ecosystem-derived knowledge with internal R&D, production, and market information [55]. This integration supports earlier recognition of potential threats, contingency planning, and advance resource preparation [42]. Consequently, when digital leadership is stronger, the informational and knowledge resources associated with ecosystem embeddedness can be more effectively incorporated into organizational anticipation and preparedness, strengthening the formation of proactive organizational resilience. Conversely, when digital leadership is weak, firms may have greater difficulty filtering and integrating heterogeneous ecosystem information, limiting the extent to which ecosystem embeddedness supports anticipation and preparedness. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 3a.

Digital leadership positively moderates the relationship between business ecosystem embeddedness and proactive organizational resilience, such that the positive effect of business ecosystem embeddedness on proactive organizational resilience is stronger when digital leadership is stronger.

Digital leadership can also strengthen the positive effect of business ecosystem embeddedness on reactive organizational resilience. When a crisis occurs, firms can draw on their business ecosystems to access alternative resources. However, the contribution of these resources to crisis response depends on firms’ ability to identify relevant resources rapidly, match them with immediate operational needs, and coordinate internal and external actors. Digital leadership encourages firms to use technologies such as digital platforms to share information and monitor the availability and status of ecosystem resources in real time, thereby improving the efficiency of resource search and matching. In addition, digital leadership facilitates clear strategic decision making and cross-functional consensus [56,57]. This promotes the rapid flow and sharing of resources within the organization and reduces delays in action caused by departmental barriers. Digital leadership can also strengthen coordination between firms and their stakeholders through real-time communication and cross-location collaboration [58]. This improves the efficiency with which external resources are incorporated into organizational decisions and operations. Thus, when digital leadership is stronger, firms are better able to coordinate and mobilize the complementary resources available through ecosystem relationships, strengthening their capacity to adapt and recover after disruptions. Conversely, when digital leadership is weak, limited real-time visibility and inefficient cross-functional and interorganizational coordination may delay resource matching and mobilization, thereby weakening the contribution of ecosystem embeddedness to reactive organizational resilience. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 3b.

Digital leadership positively moderates the relationship between business ecosystem embeddedness and reactive organizational resilience, such that the positive effect of business ecosystem embeddedness on reactive organizational resilience is stronger when digital leadership is stronger.

2.6. The Moderated Mediation Effect of Digital Leadership

Building on H2a and H3a, this study proposes that digital leadership moderates the mediating effect of proactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth. Specifically, when digital leadership is strong, firms are better able to identify valuable technological and market information within the ecosystem, evaluate its relevance, and integrate it with internal knowledge and strategic planning. This strengthens the relationship between business ecosystem embeddedness and proactive organizational resilience. In turn, stronger proactive organizational resilience enables firms to anticipate environmental changes, prepare resources in advance, and align resource commitments with emerging opportunities, thereby strengthening its mediating role in the relationship between business ecosystem embeddedness and firm growth. Conversely, when digital leadership is weak, firms may have greater difficulty filtering and integrating heterogeneous ecosystem information, limiting the extent to which ecosystem embeddedness contributes to proactive organizational resilience. Consequently, the mediating effect of proactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth is weaker. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 4a.

Digital leadership positively moderates the mediating effect of proactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth, such that this mediating effect is stronger when digital leadership is stronger.

Building on H2b and H3b, this study proposes that digital leadership moderates the mediating effect of reactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth. Specifically, when digital leadership is strong, firms can quickly identify resource gaps after a crisis occurs and use digital technologies and cross-organizational coordination to identify, match, and mobilize alternative resources provided by ecosystem partners. This strengthens the relationship between business ecosystem embeddedness and reactive organizational resilience. In turn, stronger reactive organizational resilience enables firms to adjust operations, restore critical activities, and limit disruption-related losses more effectively, thereby strengthening its mediating role in the relationship between business ecosystem embeddedness and firm growth. Conversely, when digital leadership is weak, limited real-time information and inefficient internal and interorganizational coordination may delay the identification, matching, and mobilization of ecosystem resources, limiting the extent to which ecosystem embeddedness contributes to reactive organizational resilience. Consequently, the mediating effect of reactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth is weaker. Based on the above arguments, this study proposes the following hypothesis:

Hypothesis 4b.

Digital leadership positively moderates the mediating effect of reactive organizational resilience on the relationship between business ecosystem embeddedness and SRDI SME growth, such that this mediating effect is stronger when digital leadership is stronger.

Accordingly, this study develops the research model shown in Figure 1.

3. Method

3.1. Sample

This study focuses on SRDI SMEs officially recognized at the national, provincial, or municipal level. Given China’s multi-level SRDI enterprise recognition system, the sample was drawn from official lists of SRDI cultivation enterprises and SRDI “Little Giant” enterprises. The target firms were selected from these official lists, with middle- and senior-level managers familiar with firm strategy and operations serving as respondents. Screening questions were used to confirm firms’ participation in business ecosystems. Before the formal survey, a pilot survey was conducted with middle- and senior-level managers from several SRDI SMEs in Changchun, Dalian, Shenzhen, and Beijing. Based on their feedback, the wording and structure of the questionnaire were revised to improve its clarity and reliability. Responses collected during the pilot survey were excluded from the subsequent analysis.

The formal survey used a multichannel data collection approach. First, field surveys were conducted among SRDI SMEs in Ningbo, Changchun, Shenyang, and other cities. Second, snowball sampling was used to recruit additional respondents through MBA students and alumni networks. Third, the questionnaire was distributed online to targeted respondents through the enterprise version of Wenjuanxing. To reduce potential common method bias, data were collected in two waves. The first wave was conducted from September to December 2024 and collected data on business ecosystem embeddedness, organizational resilience, digital leadership, and the control variables. A total of 700 questionnaires were distributed, of which 493 were returned. The second wave was conducted from September to December 2025. Firm growth data were collected from the firms that had participated in the first wave, and 448 firms completed the follow-up survey. A further 81 responses were excluded because they were incomplete, were completed in less than three minutes, contained logical inconsistencies, or showed patterned responding. The final sample consisted of 367 valid responses, yielding a valid response rate of 52.43%. The characteristics of the final sample are reported in

Table 1. Sample Profile.

Characteristic	Category	Frequency	Percentage	Characteristic	Category	Frequency	Percentage
CEO Gender	Male	226	61.6%	Industry	Non-metallic mineral products	36	9.8%
	Female	141	38.4%		Metal products	28	7.6%
CEO Age	20 years old or younger	0	0.0%		Automobile manufacturing	41	11.2%
	21–30 years old	30	8.2%		General and special-purpose equipment manufacturing	41	11.2%
	31–40 years old	176	47.9%		Biotechnology, chemicals, and pharmaceuticals	30	8.2%
	41–50 years old	95	25.9%		Instrumentation manufacturing	26	7.1%
	Over 50 years old	66	18.0%		Electrical machinery and equipment manufacturing	47	12.8%
CEO Education	Junior high school or below	0	0.0%		Software and information technology services	63	17.2%
	High school or technical secondary school	21	5.7%		Other	55	15.0%
	Junior college or bachelor’s degree	184	50.2%	Firm Age	≤5 years	76	20.7%
	Master’s degree or above	162	44.1%		6–10 years	122	33.2%
Firm Size (Number of Employees)	1–100	22	6.0%		11–20 years	79	21.5%
	101–300	119	32.4%		21–30 years	56	15.3%
	301–500	113	30.8%		>30 years	34	9.3%
	501–1000	79	21.5%	Ownership	Private	192	52.3%
	>1000	34	9.3%		Joint	82	22.3%
Ownership	State-owned	58	15.8%		Foreign-invested	35	9.5%

.

3.2. Measurement of Variables

All constructs were measured using well-established scales widely adopted in prior research. The items were adapted where necessary to fit the context of SRDI SMEs. All items were measured on a 7-point Likert scale. 1 = strongly disagree; 7 = strongly agree.

Business ecosystem embeddedness. This construct was measured using the scale developed by Riquelme-Medina et al. (2022) [14]. The scale comprises eight items covering three dimensions: interdependence, value potential, and shared components.

Organizational resilience. Following Lin and Fan (2024) [41], this study adopted the organizational resilience scale developed by Jia et al. (2020) [42]. Organizational resilience was measured along two dimensions: proactive organizational resilience and reactive organizational resilience. Proactive organizational resilience was assessed with four items capturing firms’ ability to identify and prevent potential risks before a crisis occurs. Reactive organizational resilience was assessed with five items capturing firms’ ability to respond rapidly and recover during and after a crisis.

Digital leadership. Digital leadership was measured using the four-item scale developed by Benitez et al. (2022) [23], which captures leaders’ digital vision, digital skills, leadership capability, and capability to lead change.

Firm growth. The measure of firm growth was adapted from Altinay et al. (2016) [59], Du et al. (2008) [60], and Gilbert et al. (2006) [61]. After adjustments to the research context, a five-item scale was developed to assess growth in terms of market expansion and technological innovation.

Control variables. Following Ma et al. (2025) [62], firm size, firm age, industry, and ownership type were included as controls because they may influence firm growth. In addition, CEO gender, age, and education were controlled for because these characteristics may affect digital leadership [24]. Accordingly, these seven variables were included as control variables.

3.3. Common Method Bias

To assess the potential influence of common method bias, this study conducted two tests using SPSS 27.0 and AMOS 23.0. First, Harman’s single-factor test was performed in SPSS 27.0 using an unrotated principal component analysis. The unrotated solution showed that the first factor accounted for 33.271% of the total variance, which was below the 40% threshold suggested by Hair et al. (1998) [63]. This result suggests that common method bias was unlikely to substantially affect the findings. In addition, the unmeasured latent method factor approach was applied in AMOS 23.0. After the latent method factor was added, the model exhibited acceptable fit (χ²/df = 1.561, GFI = 0.923, CFI = 0.967, TLI = 0.959, and RMSEA = 0.039). These fit indices differed only slightly from those of the baseline five-factor model (χ²/df = 1.659, GFI = 0.911, CFI = 0.958, TLI = 0.952, and RMSEA = 0.042). The addition of the latent method factor resulted in only modest changes in model fit (ΔGFI = 0.012, ΔCFI = 0.009, ΔTLI = 0.007, and ΔRMSEA = 0.003). The small magnitude of these changes indicates that introducing a common latent method factor did not substantially improve the overall model fit. Taken together, these results indicate that common method bias was unlikely to pose a serious threat to the validity of the study’s findings.

4. Data Analysis and Results

SPSS 27.0 and AMOS 23.0 were used to assess the reliability and validity of the measures and to conduct descriptive statistics, correlation analysis, and multiple linear regression analysis.

4.1. Reliability and Validity

The reliability and validity of the measures were assessed using SPSS 27.0, and the results are reported in

Table 2. Measurement Items and Results of Reliability and Validity Assessments.

Variables	Items	Loading	Cronbach’s α
Business Ecosystem Embeddedness CR = 0.9136 AVE = 0.5695	BEE1. Our firm and many members of the business ecosystem are interdependent and jointly support one another’s operational effectiveness and long-term development.	0.716	0.916
	BEE2. Our firm has interdependent relationships with some ecosystem members even when no direct transactions take place between us.	0.755
	BEE3. Our firm actively participates in broader business communities composed of multiple networks, such as partner networks and lead-firm networks, and functions as an integral part of these communities.	0.751
	BEE4. Our firm is deeply involved in several different business networks, each of which provides important value for specific purposes, such as acquiring technological knowledge or exchanging resources.	0.758
	BEE5. Our firm identifies potential partners that are critical to business success, such as suppliers, distributors, outsourcing firms, competitors, and research institutions.	0.754
	BEE6. Through its ecosystem networks, our firm actively explores potential opportunities in emerging markets, technologies, and products or services that do not yet exist.	0.745
	BEE7. Our firm and other ecosystem members conduct business using shared industrial infrastructure or platforms, such as industrial clusters, public services, tools, and core technologies.	0.800
	BEE8. Our firm and many of its ecosystem partners share a similar vision of the future business environment.	0.756
Proactive Organizational Resilience CR = 0.8212 AVE = 0.5347	POR1. Our firm has established internal mechanisms for recognizing potential disruptions and actively communicates such awareness to employees.	0.732	0.808
	POR2. Our firm systematically analyzes and assesses the likelihood of potential disruptions and their impact on core business operations.	0.710
	POR3. Through continuous investment, our firm strengthens its ability to prevent internal and external disruptions and adverse shocks.	0.760
	POR4. Our firm develops contingency plans for potential crises that may affect its business operations.	0.722
Reactive Organizational Resilience CR = 0.8514 AVE = 0.5343	ROR1. Our firm can quickly recognize crisis situations that pose a threat to its business operations.	0.748	0.844
	ROR2. Our firm can promptly collect and interpret relevant information to assess the scale, location, and causes of a disruption.	0.713
	ROR3. Our firm can quickly identify, develop, and evaluate multiple feasible responses to business disruptions.	0.721
	ROR4. When a crisis occurs, our firm can quickly form a response team composed of key personnel and coordinate actions across organizational levels.	0.693
	ROR5. Our firm responds effectively to crises and can maintain stable business operations during a crisis.	0.777
Digital Leadership CR = 0.8777 AVE = 0.6420	DL1. Our firm’s leaders are capable of exploring and leveraging emerging trends in information technology.	0.800	0.845
	DL2. Our firm’s leaders can effectively identify and successfully deploy information technology applications and services to enhance competitiveness.	0.808
	DL3. Our firm’s leaders are capable of developing innovative strategic business and operating models and of anticipating and driving changes that improve business performance.	0.804
	DL4. Our firm’s leaders can lead interdisciplinary teams and influence stakeholders across functional and geographical boundaries.	0.793
Firm Growth CR = 0.8250 AVE = 0.4864	FG1. Compared with our competitors, our firm has achieved faster growth in sales revenue.	0.634	0.826
	FG2. Compared with our competitors, our firm has achieved faster growth in market share.	0.694
	FG3. Compared with our competitors, our firm has achieved faster growth in net profit.	0.745
	FG4. Compared with our competitors, our firm has experienced faster growth in the number of employees.	0.745
	FG5. Compared with our competitors, our firm develops and iterates new products or technologies more rapidly.	0.662

. The exploratory factor analysis yielded a Kaiser–Meyer–Olkin (KMO) value of 0.913, while Bartlett’s test of sphericity was significant (p < 0.001), indicating that the data were suitable for factor analysis. As shown in Table 2, the Cronbach’s alpha coefficients ranged from 0.808 to 0.916, and the composite reliability (CR) values ranged from 0.8212 to 0.9136, all exceeding the recommended threshold of 0.70. The factor loadings ranged from 0.634 to 0.808 and were all above 0.50. The average variance extracted (AVE) values exceeded 0.50 for all constructs except firm growth, for which the AVE was 0.4864. According to Fornell and Larcker (1981) [64], a construct may still demonstrate acceptable convergent validity when its AVE is slightly below 0.50, provided that its CR exceeds 0.70 and its factor loadings remain at acceptable levels. Although the AVE of firm growth was marginally below the conventional 0.50 threshold, its CR was 0.8250 and all standardized factor loadings ranged from 0.634 to 0.745. This result suggests that the convergent validity of the firm growth construct is somewhat weaker than that of the other constructs but remains acceptable when considered jointly with its composite reliability and item loadings. Overall, the measurement scales demonstrated satisfactory reliability and convergent validity.

Confirmatory factor analysis was conducted using AMOS 23.0, and the results are reported in

Table 3. Confirmatory Factor Analysis Results.

Model	Factors Included	χ2/df	GFI	CFI	TLI	RMSEA
Five factors	BEE, POR, ROR, DL, FG	1.659	0.911	0.958	0.952	0.042
Four factors	BEE, POR + ROR, DL, FG	2.430	0.855	0.907	0.897	0.063
Three factors	BEE + POR + ROR, DL, FG	4.752	0.686	0.753	0.729	0.101
Two factors	BEE + POR + ROR + DL, FG	6.402	0.615	0.642	0.609	0.121
One factor	BEE + POR + ROR + DL + FG	7.240	0.577	0.585	0.549	0.131

Note: N = 367, BEE = business ecosystem embeddedness, POR = proactive organizational resilience, ROR = reactive organizational resilience, DL = digital leadership, FG = firm growth; the same abbreviations apply to subsequent tables. + indicates that two factors were combined into a single factor.

. The five-factor model exhibited a better fit than the alternative models, with χ²/df = 1.659, GFI = 0.911, CFI = 0.958, TLI = 0.952, RMSEA = 0.042. These results provide support for the discriminant validity of the measurement model.

To further assess discriminant validity, the heterotrait–monotrait ratio (HTMT) was calculated for all focal constructs. As shown in

Table 4. Heterotrait–Monotrait Ratio (HTMT) of the Constructs.

Variables	1	2	3	4	5
1 BEE	-
2 POR	0.446	-
3 ROR	0.438	0.595	-
4 DL	0.292	0.370	0.339	-
5 FG	0.553	0.574	0.598	0.392	-

, the HTMT values ranged from 0.292 to 0.598, all below the recommended threshold of 0.85. These results provide further evidence of satisfactory discriminant validity among the focal constructs.

4.2. Descriptive Statistics and Correlation Analysis

Before testing the hypotheses, this study conducted descriptive statistical analysis and bivariate correlation analysis for all variables. The results are presented in Table 5. The means and standard deviations of all variables fell within acceptable ranges. In addition, the absolute values of the correlations among the core variables were all below 0.70, suggesting that severe multicollinearity was unlikely. Specifically, business ecosystem embeddedness was positively correlated with proactive organizational resilience (r = 0.384, p < 0.01), reactive organizational resilience (r = 0.385, p < 0.01), and SRDI SME growth (r = 0.481, p < 0.01). Proactive organizational resilience was positively correlated with SRDI SME growth (r = 0.469, p < 0.01), as was reactive organizational resilience (r = 0.500, p < 0.01). These significant correlations provide preliminary support for the proposed hypotheses.

4.3. Hypothesis Testing

The hypotheses were tested using hierarchical regression analysis and bootstrapping in SPSS 27.0 with the PROCESS macro. In addition, the variables used to construct the interaction terms were mean-centered to reduce nonessential multicollinearity. All variance inflation factor (VIF) values were below 1.3, indicating that multicollinearity was not a serious concern.

4.3.1. The Main Effect

The direct effect of business ecosystem embeddedness on SRDI SME growth, as proposed in H1, was tested, and the results are reported in Table 6. As shown in Model M2, business ecosystem embeddedness was significantly and positively related to SRDI SME growth (β = 0.470, p < 0.01). Therefore, H1 was supported.

Table 5. Correlation Matrix.

Variables	1	2	3	4	5	6	7	8	9	10	11	12
1 BEE	(0.755)
2 POR	0.384 ***	(0.731)
3 ROR	0.385 ***	0.491 ***	(0.731)
4 DL	0.256 ***	0.304 ***	0.286 ***	(0.801)
5 FG	0.481 ***	0.469 ***	0.500 ***	0.328 ***	(0.697)
6 Lgen	0.109 **	0.104 **	0.052	0.067	0.152 ***	-
7 Lage	−0.009	0.020	0.008	−0.018	−0.051	0.177 ***	-
8 Ledu	0.060	0.006	−0.076	−0.046	0.010	0.106	0.133 ***	-
9 F-size	−0.020	−0.053	−0.055	−0.151 ***	−0.088	0.072	−0.010	0.078	-
10 F-age	−0.164 ***	−0.096	−0.039	−0.079	−0.099	0.034	0.089 *	0.085	0.230 ***	-
11 Ind	0.002	0.034	−0.006	−0.029	−0.003	−0.056	−0.114 **	−0.115 **	0.013	−0.012	-
12 Own	0.051	0.033	−0.045	0.006	−0.070	0.081	0.021	0.049	−0.055	−0.060	−0.061	-
Mean	4.142	4.226	4.254	4.143	4.184	0.62	3.54	3.38	2.96	2.59	5.49	2.26
SD	1.193	1.205	1.138	1.315	1.133	0.487	0.880	0.593	1.073	1.233	2.669	0.836

Note: Lgen = CEO gender, Lage = CEO age, Ledu = CEO Education, F-size = Firm Size, F-age = Firm Age, Ind = Industry, Own = Ownership. ***: p < 0.01, **: p < 0.05, *: p < 0.1.

Table 5. Correlation Matrix.

Variables	1	2	3	4	5	6	7	8	9	10	11	12
1 BEE	(0.755)
2 POR	0.384 ***	(0.731)
3 ROR	0.385 ***	0.491 ***	(0.731)
4 DL	0.256 ***	0.304 ***	0.286 ***	(0.801)
5 FG	0.481 ***	0.469 ***	0.500 ***	0.328 ***	(0.697)
6 Lgen	0.109 **	0.104 **	0.052	0.067	0.152 ***	-
7 Lage	−0.009	0.020	0.008	−0.018	−0.051	0.177 ***	-
8 Ledu	0.060	0.006	−0.076	−0.046	0.010	0.106	0.133 ***	-
9 F-size	−0.020	−0.053	−0.055	−0.151 ***	−0.088	0.072	−0.010	0.078	-
10 F-age	−0.164 ***	−0.096	−0.039	−0.079	−0.099	0.034	0.089 *	0.085	0.230 ***	-
11 Ind	0.002	0.034	−0.006	−0.029	−0.003	−0.056	−0.114 **	−0.115 **	0.013	−0.012	-
12 Own	0.051	0.033	−0.045	0.006	−0.070	0.081	0.021	0.049	−0.055	−0.060	−0.061	-
Mean	4.142	4.226	4.254	4.143	4.184	0.62	3.54	3.38	2.96	2.59	5.49	2.26
SD	1.193	1.205	1.138	1.315	1.133	0.487	0.880	0.593	1.073	1.233	2.669	0.836

Note: Lgen = CEO gender, Lage = CEO age, Ledu = CEO Education, F-size = Firm Size, F-age = Firm Age, Ind = Industry, Own = Ownership. ***: p < 0.01, **: p < 0.05, *: p < 0.1.

4.3.2. The Mediation Effect

First, the mediating role of proactive organizational resilience was examined. The results are reported in Models M4, M7, and M8 of Table 6. Models M4 and M7 show that business ecosystem embeddedness was significantly and positively related to proactive organizational resilience (β = 0.372, p < 0.01) and that proactive organizational resilience was significantly and positively related to SRDI SME growth (β = 0.453, p < 0.01), respectively. When business ecosystem embeddedness and proactive organizational resilience were entered simultaneously in Model M8, the coefficient of business ecosystem embeddedness decreased from 0.470 to 0.349 but remained significant (p < 0.01). The Bootstrap analysis further showed that the indirect effect of business ecosystem embeddedness on SRDI SME growth through proactive organizational resilience was 0.1215 (Boot SE = 0.0237, 95% CI = [0.0775, 0.1705]). Because the confidence interval did not include zero, the indirect effect through proactive organizational resilience was statistically significant. Meanwhile, the relationship between business ecosystem embeddedness and SRDI SME growth remained significant after proactive organizational resilience was included, indicating partial mediation. Thus, H2a was supported.

Similarly, Models M6, M9, and M10 in Table 6 report the results for the mediating role of reactive organizational resilience. Models M6 and M9 show that business ecosystem embeddedness was significantly and positively related to reactive organizational resilience (β = 0.398, p < 0.01) and that reactive organizational resilience had a significant positive effect on SRDI SME growth (β = 0.488, p < 0.01), respectively. When business ecosystem embeddedness and reactive organizational resilience were entered simultaneously in Model M10, the coefficient of business ecosystem embeddedness decreased from 0.470 to 0.326 but remained significant (p < 0.01). The bootstrap analysis further showed that the indirect effect of business ecosystem embeddedness on SRDI SME growth through reactive organizational resilience was 0.1439 (Boot SE = 0.0263, 95% CI = [0.0938, 0.1987]). Because the confidence interval did not include zero, the indirect effect through reactive organizational resilience was statistically significant. Meanwhile, the relationship between business ecosystem embeddedness and SRDI SME growth remained significant after reactive organizational resilience was included, indicating partial mediation. Thus, H2b was supported.

4.3.3. The Moderating Effect

To test the moderating effect of digital leadership, the interaction term between business ecosystem embeddedness and digital leadership was included in the regression models. The two variables were mean-centered before the interaction term was constructed to reduce potential nonessential multicollinearity. The results of the moderating effect tests are reported in Table 7. First, to test H3a, digital leadership was added to Model M4 in Table 6, yielding Model M11. The results show that both business ecosystem embeddedness (β = 0.318, p < 0.01) and digital leadership (β = 0.217, p < 0.01) were significantly and positively related to proactive organizational resilience. The interaction term between business ecosystem embeddedness and digital leadership was then added to form Model M12. The results show that business ecosystem embeddedness continued to have a significant positive effect on proactive organizational resilience (β = 0.289, p < 0.01), while the interaction term also had a significant positive effect on proactive organizational resilience (β = 0.140, p < 0.01). These results indicate that digital leadership positively moderated the relationship between business ecosystem embeddedness and proactive organizational resilience. Thus, H3a was supported.

Table 6. Results of the Direct and Mediation Analyses.

Variables	Firm Growth		Proactive Organizational Resilience		Reactive Organizational Resilience		Firm Growth
	M1	M2	M3	M4	M5	M6	M7	M8	M9	M10
Lgen	0.181 ***	0.129 ***	0.108 **	0.067	0.066	0.023	0.132 ***	0.107 **	0.148 ***	0.121 ***
Lage	−0.078	−0.068	0.012	0.019	0.008	0.016	−0.083 *	−0.074 *	−0.081 *	−0.074 *
Ledu	0.019	−0.011	0.007	−0.017	−0.076	−0.102	0.016	−0.006	0.056	0.026
Size	−0.089 *	−0.093 **	−0.040	−0.043	−0.050	−0.053	−0.071	−0.079 *	−0.065	−0.073 *
Age	−0.085	−0.004	−0.090 *	−0.027	−0.028	0.041	−0.044	0.004	−0.071	−0.019
Ind	−0.005	−0.011	0.043	0.038	−0.013	−0.018	−0.024	−0.023	0.001	−0.004
Own	−0.094 *	−0.108 **	0.018	0.007	−0.052	−0.064	−0.103 **	−0.111 **	−0.069	−0.085 **
BEE		0.470 ***		0.372 ***		0.398 ***		0.349 ***		0.326 ***
POR							0.453 ***	0.326 ***
ROR									0.488 ***	0.362 ***
R2	0.055	0.265	0.024	0.157	0.016	0.167	0.255	0.355	0.289	0.375
Adj-R2	0.036	0.249	0.005	0.138	−0.004	0.148	0.238	0.339	0.273	0.359
F	2.961 ***	16.173 ***	1.278	8.312 ***	0.809	8.939 ***	15.322 ***	21.859 ***	18.156 ***	23.757 ***
VIF	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3

***: p < 0.01, **: p < 0.05, *: p < 0.1.

Table 6. Results of the Direct and Mediation Analyses.

Variables	Firm Growth		Proactive Organizational Resilience		Reactive Organizational Resilience		Firm Growth
	M1	M2	M3	M4	M5	M6	M7	M8	M9	M10
Lgen	0.181 ***	0.129 ***	0.108 **	0.067	0.066	0.023	0.132 ***	0.107 **	0.148 ***	0.121 ***
Lage	−0.078	−0.068	0.012	0.019	0.008	0.016	−0.083 *	−0.074 *	−0.081 *	−0.074 *
Ledu	0.019	−0.011	0.007	−0.017	−0.076	−0.102	0.016	−0.006	0.056	0.026
Size	−0.089 *	−0.093 **	−0.040	−0.043	−0.050	−0.053	−0.071	−0.079 *	−0.065	−0.073 *
Age	−0.085	−0.004	−0.090 *	−0.027	−0.028	0.041	−0.044	0.004	−0.071	−0.019
Ind	−0.005	−0.011	0.043	0.038	−0.013	−0.018	−0.024	−0.023	0.001	−0.004
Own	−0.094 *	−0.108 **	0.018	0.007	−0.052	−0.064	−0.103 **	−0.111 **	−0.069	−0.085 **
BEE		0.470 ***		0.372 ***		0.398 ***		0.349 ***		0.326 ***
POR							0.453 ***	0.326 ***
ROR									0.488 ***	0.362 ***
R2	0.055	0.265	0.024	0.157	0.016	0.167	0.255	0.355	0.289	0.375
Adj-R2	0.036	0.249	0.005	0.138	−0.004	0.148	0.238	0.339	0.273	0.359
F	2.961 ***	16.173 ***	1.278	8.312 ***	0.809	8.939 ***	15.322 ***	21.859 ***	18.156 ***	23.757 ***
VIF	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3	<1.3

***: p < 0.01, **: p < 0.05, *: p < 0.1.

Similarly, to test H3b, digital leadership was added to Model M6 in Table 6, yielding Model M13. The results show that both business ecosystem embeddedness (β = 0.350, p < 0.01) and digital leadership (β = 0.191, p < 0.01) were significantly and positively related to reactive organizational resilience. The interaction term between business ecosystem embeddedness and digital leadership was then added to form Model M14. The results show that business ecosystem embeddedness continued to have a significant positive effect on reactive organizational resilience (β = 0.315, p < 0.01), while the interaction term also had a significant positive effect on reactive organizational resilience (β = 0.168, p < 0.01). These results indicate that digital leadership positively moderated the relationship between business ecosystem embeddedness and reactive organizational resilience. Thus, H3b was supported.

To illustrate the moderating effect of digital leadership more clearly, this study conducted a simple slope analysis following Aiken and West (1991) [65] and plotted the relationships between business ecosystem embeddedness and proactive organizational resilience, as well as between business ecosystem embeddedness and reactive organizational resilience, at different levels of digital leadership, as shown in Figure 2 and Figure 3. As shown in the figures, business ecosystem embeddedness had a positive effect on both proactive organizational resilience and reactive organizational resilience regardless of the level of digital leadership. However, the slopes were steeper when digital leadership was high than when it was low, indicating that the positive effects of business ecosystem embeddedness on proactive organizational resilience and reactive organizational resilience were stronger at higher levels of digital leadership. These findings provide further support for H3a and H3b.

4.3.4. The Moderated Mediation Effect

Following Hayes (2013) [66], this study further tested the moderated mediation effects proposed in H4a and H4b using Model 7 of the PROCESS macro in SPSS 27.0. The results are reported in Table 8 and Table 9. As shown in Table 8, the conditional indirect effects through proactive organizational resilience were 0.0525 (95% CI = [0.0020, 0.1063]), 0.0943 (95% CI = [0.0582, 0.1368]), and 0.1362 (95% CI = [0.0879, 0.1928]) at low, moderate, and high levels of digital leadership, respectively. These results indicate that the indirect effect through proactive organizational resilience increased as the level of digital leadership increased. Moreover, the index of moderated mediation was 0.0419 (95% CI = [0.0108, 0.0781]), and the confidence interval did not include zero, indicating that the moderated mediation effect was significant. Thus, H4a was supported.

Table 7. Results of the Moderating Effect Analysis.

Variables	Proactive Organizational Resilience		Reactive Organizational Resilience
	M11	M12	M13	M14
Lgen	0.054	0.041	0.012	−0.004
Lage	0.024	0.023	0.020	0.019
Ledu	−0.005	−0.009	−0.091 *	−0.097 *
Size	−0.011	−0.007	−0.025	−0.019
Age	−0.026	−0.032	0.041	0.034
Ind	0.045	0.050	−0.011	−0.006
Own	0.011	0.014	−0.060	−0.058
BEE	0.318 ***	0.289 ***	0.350 ***	0.315 ***
DL	0.217 ***	0.186 ***	0.191 ***	0.154 ***
BEE×DL		0.140 ***		0.168 ***
R2	0.199	0.216	0.199	0.224
Adj-R2	0.179	0.194	0.179	0.202
F	9.872 ***	9.815 ***	9.880 ***	10.263 ***
VIF	<1.3	<1.3	<1.3	<1.3

***: p < 0.01, *: p < 0.1.

Table 7. Results of the Moderating Effect Analysis.

Variables	Proactive Organizational Resilience		Reactive Organizational Resilience
	M11	M12	M13	M14
Lgen	0.054	0.041	0.012	−0.004
Lage	0.024	0.023	0.020	0.019
Ledu	−0.005	−0.009	−0.091 *	−0.097 *
Size	−0.011	−0.007	−0.025	−0.019
Age	−0.026	−0.032	0.041	0.034
Ind	0.045	0.050	−0.011	−0.006
Own	0.011	0.014	−0.060	−0.058
BEE	0.318 ***	0.289 ***	0.350 ***	0.315 ***
DL	0.217 ***	0.186 ***	0.191 ***	0.154 ***
BEE×DL		0.140 ***		0.168 ***
R2	0.199	0.216	0.199	0.224
Adj-R2	0.179	0.194	0.179	0.202
F	9.872 ***	9.815 ***	9.880 ***	10.263 ***
VIF	<1.3	<1.3	<1.3	<1.3

***: p < 0.01, *: p < 0.1.

Table 8. The Moderating Effect of Digital Leadership on the Mediating Effect of Proactive Organizational Resilience.

Conditional Indirect Effect(s) of X on Y at Values of the Moderator(s)
Moderator	Effect	Boot SE	Boot LLCI	Boot ULCI
DL (−1 SD)	0.0525	0.0263	0.0020	0.1063
DL (Mean)	0.0943	0.0204	0.0582	0.1368
DL (+1 SD)	0.1362	0.0269	0.0879	0.1928
Index of Moderated Mediation
Moderator	Index	Boot SE	Boot LLCI	Boot ULCI
DL	0.0419	0.0171	0.0108	0.0781

Table 8. The Moderating Effect of Digital Leadership on the Mediating Effect of Proactive Organizational Resilience.

Conditional Indirect Effect(s) of X on Y at Values of the Moderator(s)
Moderator	Effect	Boot SE	Boot LLCI	Boot ULCI
DL (−1 SD)	0.0525	0.0263	0.0020	0.1063
DL (Mean)	0.0943	0.0204	0.0582	0.1368
DL (+1 SD)	0.1362	0.0269	0.0879	0.1928
Index of Moderated Mediation
Moderator	Index	Boot SE	Boot LLCI	Boot ULCI
DL	0.0419	0.0171	0.0108	0.0781

Similarly, as shown in Table 9, the conditional indirect effects through reactive organizational resilience were 0.0582 (95% CI = [0.0052, 0.1152]), 0.1140 (95% CI = [0.0731, 0.1606]), and 0.1699 (95% CI = [0.1140, 0.2299]) at low, moderate, and high levels of digital leadership, respectively. These results indicate that the indirect effect through reactive organizational resilience increased as the level of digital leadership increased. Moreover, the index of moderated mediation was 0.0558 (95% CI = [0.0221, 0.0942]). Because the confidence interval did not include zero, the moderated mediation effect was significant. Thus, H4b was supported.

Table 9. The Moderating Effect of Digital Leadership on the Mediating Effect of Reactive Organizational Resilience.

Conditional Indirect Effect(s) of X on Y at Values of the Moderator(s)
Moderator	Effect	Boot SE	Boot LLCI	Boot ULCI
DL (−1 SD)	0.0582	0.0274	0.0052	0.1152
DL (Mean)	0.1140	0.0220	0.0731	0.1606
DL (+1 SD)	0.1699	0.0298	0.1140	0.2299
Index of Moderated Mediation
Moderator	Index	Boot SE	Boot LLCI	Boot ULCI
DL	0.0558	0.0183	0.0221	0.0942

Table 9. The Moderating Effect of Digital Leadership on the Mediating Effect of Reactive Organizational Resilience.

Conditional Indirect Effect(s) of X on Y at Values of the Moderator(s)
Moderator	Effect	Boot SE	Boot LLCI	Boot ULCI
DL (−1 SD)	0.0582	0.0274	0.0052	0.1152
DL (Mean)	0.1140	0.0220	0.0731	0.1606
DL (+1 SD)	0.1699	0.0298	0.1140	0.2299
Index of Moderated Mediation
Moderator	Index	Boot SE	Boot LLCI	Boot ULCI
DL	0.0558	0.0183	0.0221	0.0942

4.3.5. Robustness Test

To assess the robustness of the findings, this study further employed structural equation modeling (SEM) to re-examine the core relationships in the proposed model, as shown in Figure 4. The standardized path coefficient from business ecosystem embeddedness to SRDI SME growth was positive and significant (β = 0.224, z = 4.691, p < 0.01). The standardized path coefficients from business ecosystem embeddedness to proactive organizational resilience and reactive organizational resilience were also positive and significant (β = 0.297, z = 5.632, p < 0.01; β = 0.342, z = 6.380, p < 0.01, respectively). In addition, proactive organizational resilience and reactive organizational resilience were both positively and significantly associated with SRDI SME growth (β = 0.266, z = 4.582, p < 0.01; β = 0.277, z = 5.306, p < 0.01, respectively). Furthermore, the interaction between business ecosystem embeddedness and digital leadership was positively and significantly associated with proactive organizational resilience (β = 0.153, z = 3.028, p < 0.01) and reactive organizational resilience (β = 0.193, z = 3.685, p < 0.01). These results were consistent with the main analyses in terms of direction and statistical significance, providing further evidence for the robustness of the study’s findings.

5. Conclusions and Discussion

5.1. Conclusions

Drawing on resource orchestration theory and focusing on Specialized, Refined, Differentiated, and Innovative small and medium-sized enterprises (SRDI SMEs), this study developed a theoretical model along the logical pathway of “business ecosystem embeddedness–organizational resilience–firm growth” to explain how business ecosystem embeddedness promotes firm growth. It further examined digital leadership as a boundary condition in the process through which ecosystem resources are transformed into growth outcomes. The proposed model was empirically tested using two-wave survey data collected from 367 Chinese SRDI SMEs. The results show that business ecosystem embeddedness positively affects SRDI SME growth. Both proactive organizational resilience and reactive organizational resilience partially mediate this relationship. Moreover, digital leadership positively moderates the relationships between business ecosystem embeddedness and both dimensions of organizational resilience. It also strengthens the indirect effects of business ecosystem embeddedness on SRDI SME growth through proactive and reactive organizational resilience. The main findings are summarized as follows.

First, business ecosystem embeddedness was significantly and positively associated with the growth of SRDI SMEs. This finding is consistent with Zang et al. (2022) [51], which showed that ecosystem embeddedness benefits the performance of non-core firms. And the finding is also in line with recent ecosystem research emphasizing that interdependence, resource integration, and value co-creation can help SMEs overcome the limitations of their internal resource bases [67]. For SRDI SMEs, this relationship is particularly relevant because their resource are often concentrated in specialized technological domains, whereas continued growth requires complementary knowledge, market access, technologies, and application opportunities beyond organizational boundaries. Deeper ecosystem embeddedness can therefore broaden the external resource base available to these firms and increase the opportunities for their specialized offerings to be integrated into broader value-creation activities. However, this finding should not be interpreted as implying that greater embeddedness is universally beneficial. Earlier research on the paradox of embeddedness suggests that excessive reliance on embedded relationships may generate lock-in, restrict exposure to novel information, and reduce strategic flexibility [30]. The positive relationship observed in this study therefore suggests that, within SRDI SMEs, the resource-access and complementarity benefits associated with ecosystem embeddedness outweighed these potential constraints. Therefore, business ecosystem embeddedness contributes to the sustained growth of SRDI SMEs.

Second, both proactive organizational resilience and reactive organizational resilience partially mediate the relationship between business ecosystem embeddedness and SRDI SME growth, indicating that business ecosystem embeddedness is indirectly associated with sustained firm growth through two resilience capabilities corresponding to different stages of environmental disruption. This finding echoes Duchek’s (2020) capability-based framework of organizational resilience, which distinguishes among anticipation before a disruption, coping during a disruption, and adaptation after a disruption as three interrelated stages [68]. Proactive organizational resilience relies more heavily on firms’ access to diverse market, technological, and supply-chain information, which helps them identify emerging threats and prepare before disruptions fully materialize. In contrast, reactive organizational resilience depends more strongly on timely access to complementary resources and cross-organizational coordination, thereby facilitating rapid operational adjustment and recovery after disruptions occur. Recent research on SME resilience likewise suggests that organizational resilience is grounded in multiple capability foundations, including cognitive, technological, and value-chain-related capabilities, rather than representing a single undifferentiated response capacity [69]. For resource-constrained SRDI SMEs, these two resilience capabilities are particularly important for sustaining growth in turbulent environments. Proactive organizational resilience helps firms avoid committing scarce resources to inappropriate technological paths or low-potential markets. Reactive organizational resilience helps firms reduce operational losses and recover more quickly when disruptions cannot be anticipated or avoided.

Third, digital leadership strengthens the relationships between business ecosystem embeddedness and both dimensions of organizational resilience, and it also strengthens their mediating roles. This finding is consistent with previous research showing that leadership and digital technologies play important roles in enhancing organizational resilience in SMEs [70]. Recent studies further suggest that digital leadership can influence the extent to which organizational resources and capabilities are translated into subsequent organizational outcomes [71,72]. This implies that firms with similar access to resources may derive different benefits depending on their levels of digital leadership. In the context of SRDI SMEs, the present study further shows that firms with higher levels of digital leadership are better able to filter and integrate information obtained from the ecosystem, thereby supporting risk anticipation and advance preparation. Digital leadership also facilitates cross-organizational communication, coordination, and resource mobilization when disruptions occur. These functions further strengthen the relationships between ecosystem embeddedness and both dimensions of organizational resilience. Accordingly, as digital leadership increases, the relationships between business ecosystem embeddedness and the two dimensions of organizational resilience become stronger, and the corresponding mediating effects are also strengthened.

5.2. Theoretical Contributions

First, this study examines the drivers of SRDI SME growth from an ecosystem perspective, thereby extending traditional internally oriented explanations of firm growth. Existing research on the growth of SRDI enterprises has primarily focused on internal factors, such as leadership, specialization strategies, and organizational culture [73,74], whereas relatively little attention has been paid to the determinants of their growth from an ecosystem perspective [75]. The SRDI context is theoretically relevant because these firms typically concentrate their limited resources on specialized technological domains, while continued growth requires access to complementary knowledge, technologies, market opportunities, and other resources beyond organizational boundaries. By adopting an ecosystem perspective, this study responds to the call by Zeng et al. (2023) [75] to complement traditional firm growth research centered on internal factors. More specifically, the findings show that business ecosystem embeddedness represents an important external resource-access condition associated with SRDI SME growth. This shifts attention from what specialized firms possess internally to how their relationships with ecosystem actors can complement their constrained internal resource bases. Accordingly, this study not only broadens research on the determinants of SRDI SME growth but also highlights the theoretical relevance of ecosystem relationships for understanding the growth of resource-constrained specialized firms.

Second, this study advances research on organizational resilience by distinguishing between proactive and reactive organizational resilience and clarifying their distinct mediating roles in the relationship between business ecosystem embeddedness and SRDI SME growth. Previous studies have often treated organizational resilience as a unitary construct [20,76,77], which may obscure the temporal and functional differences among resilience capabilities. By separating proactive resilience from reactive resilience, this study provides a more fine-grained explanation of how ecosystem embeddedness is linked to firm growth through resilience capabilities that operate at different stages of environmental disruption. Proactive organizational resilience captures firms’ capacity to anticipate potential threats and prepare before disruptions occur, whereas reactive organizational resilience reflects their capacity to respond, adapt, and recover once disruptions materialize. This distinction is theoretically important because the two mediating paths reflect different ways in which ecosystem-derived information and resources can support firm growth under turbulence. Proactive resilience emphasizes anticipation and advance preparation, while reactive resilience emphasizes adjustment and recovery after disruptions occur. Accordingly, this study moves beyond a unitary view of organizational resilience and shows the value of examining differentiated resilience capabilities when explaining how business ecosystem embeddedness is associated with SRDI SME growth.

Finally, this study extends understanding of the boundary conditions of resource orchestration theory by identifying digital leadership as an important internal managerial condition that shapes the relationship between business ecosystem embeddedness and resilience capability formation. Previous studies have mainly treated digital leadership as an independent variable and examined its direct effects on digital transformation, organizational agility, and firm performance [23,57]. More recent research has begun to show that digital leadership can also play a moderating role in shaping the extent to which organizational resources and capabilities contribute to subsequent outcomes [71,72]. At the same time, recent resilience research has emphasized the joint role of leadership, organizational capabilities, and inter-organizational relationships in shaping SME resilience [78]. Building on these developments, this study incorporates digital leadership into the resource orchestration framework as an internal contextual condition and shows that it shapes the relationships between business ecosystem embeddedness and proactive and reactive organizational resilience. This finding indicates that the capability benefits associated with ecosystem resource access depend partly on firms’ internal leadership conditions. Accordingly, this study enriches resource orchestration theory by further specifying an important internal managerial boundary condition in the digital era.

5.3. Practical Implications

The findings of this study offer the following three practical implications for the sustained growth of SRDI SMEs in turbulent environments:

First, SRDI SMEs should proactively embed themselves in business ecosystems to overcome internal resource constraints in turbulent environments. Rather than relying solely on internally accumulated resources for growth, these firms should leverage their specialized technological strengths to establish stable and collaborative relationships with key customers, suppliers, research institutions, platform firms, and other complementary partners. Managers can regularly map critical ecosystem partners, identify gaps in external knowledge and resources, and establish mechanisms for information sharing, joint technology development, and market collaboration. By becoming embedded in business ecosystems, SRDI SMEs can gain access to a broader range of knowledge and resources, including technological expertise, market information, application opportunities, and complementary capabilities. These can help firms complement their limited internal resource bases and create more opportunities for sustained growth.

Second, SRDI SMEs should cultivate both dimensions of organizational resilience by developing anticipatory preparedness and rapid response capabilities in parallel. To strengthen proactive organizational resilience, firms should maintain continuous interactions with ecosystem partners, such as customers and suppliers, to monitor technological changes, shifts in market demand, and potential supply-chain risks. Managers can establish early-warning indicators, conduct scenario planning, and prepare contingency plans for critical technologies, markets, and suppliers. To strengthen reactive organizational resilience, firms should develop alternative supplier networks, cross-functional emergency teams, and rapid communication and resource-allocation procedures. When disruptions occur, these arrangements can support timely coordination with ecosystem partners, rapid resource redeployment, and the restoration of critical operations. By developing both anticipatory and response capabilities, SRDI SMEs can improve their preparedness before disruptions and their ability to adjust and recover after disruptions, thereby supporting sustained growth in turbulent environments.

Third, SRDI SMEs should attach greater importance to developing digital leadership. Business leaders should continuously improve their digital knowledge and strengthen their ability to understand how digital transformation affects technologies, markets, and business models. They should also develop a clear digital vision and communicate it effectively throughout the organization. To build internal consensus, leaders should encourage employees to participate in digital initiatives and promote collaboration across functional boundaries. Managers should also strengthen their ability to make data-informed decisions and use digital information to identify emerging risks and opportunities. Beyond the organizational boundary, leaders should actively promote communication and collaboration with ecosystem partners and develop the ability to coordinate external information and complementary resources. During periods of disruption, strong digital leadership can support faster decision-making and more effective coordination among internal departments and external partners. These efforts can help SRDI SMEs strengthen proactive and reactive organizational resilience and support sustained growth.

5.4. Limitations and Future Research

Although this study examines the mechanisms and boundary conditions linking business ecosystem embeddedness to the growth of SRDI SMEs, several limitations remain. First, although the data were collected in two waves, the research design does not constitute a fully longitudinal panel study because each focal construct was measured at only one point in time. Firm growth, however, is a long-term and dynamic process, and such a design may not fully capture the characteristics and mechanisms of growth at different stages. Future research could therefore adopt longitudinal case studies or other process-oriented research designs to examine how the effects of business ecosystem embeddedness vary across different stages of firm growth. Second, the sample is limited to officially recognized SRDI SMEs in China, which may constrain the generalizability of the findings. SRDI SMEs operate within China’s specific institutional and policy environment, while ecosystem relationships may differ across market, institutional, and cultural contexts. Future research could examine specialized SMEs in Europe, North America, and other economies to assess whether the proposed relationships hold across different contexts. Cross-country comparisons could further clarify how institutional and cultural differences shape the relationships among ecosystem embeddedness, organizational resilience, and firm growth. Third, this study focuses on the overall relationship between business ecosystem embeddedness and firm growth. However, different types of ecosystems and different forms of embeddedness may have distinct effects on growth. Future research could compare the effects of different ecosystem types, such as digital ecosystems, innovation ecosystems, and platform ecosystems, as well as different forms of embeddedness, including structural, relational, and cognitive embeddedness, to provide a more fine-grained understanding of how different ecosystem configurations are associated with organizational resilience and firm growth. Finally, this study examines only how an internal organizational factor influences the transformation of ecosystem resource advantages into organizational capabilities and firm growth. External contextual factors may also play an important role in this process. Future research could examine the contingent effects of environmental turbulence and complexity or integrate external and internal contextual factors to investigate their joint boundary effects.