Achieving Sustainable Innovation: A Fit Model of Digital Platforms and Absorptive Capacity

Abstract

Although researchers have identified a strong correlation between digital platform usage and corporate sustainable innovation, a disconnect persists between the two. The primary factor causing this disconnect lies in the isolated application of digital platform capabilities and organizational capabilities. To bridge this gap, the study proposes that during the enhancement of sustainable innovation, corporate digital platform capabilities and absorptive capacity exhibit a synergistic effect. Based on resource orchestration theory, this study posits that the fit mechanism between digital platform capabilities and absorptive capacity drives corporate sustainable innovation. It analyzes the mediating role of the knowledge duality between this fit mechanism and sustainable innovation. Analysis of primary survey data reveals that the fit mechanism of digital platform capabilities and absorptive capacity effectively enhances sustainable innovation capabilities, with dual knowledge capabilities mediating between this fit mechanism and sustainable innovation. These findings enrich the theoretical framework for leveraging digital platforms to achieve sustainable innovation, significantly supplementing research on the relationships among absorptive capacity, knowledge ambidexterity, and sustainable innovation. This study provides theoretical support for sustainable innovation research while offering practical guidance for practitioners.

1. Introduction

Discussions on sustainable development have grown exponentially and garnered increasing attention in the field of corporate sustainability management [1,2,3]. Rising public awareness, stricter government regulatory requirements, and market pressures have compelled many companies to integrate sustainability into their strategic development agendas [4,5,6]. Sustainable innovation (SI) is defined as “the introduction of new or significantly improved products, production processes, management practices, or business models that deliver economic, social, and environmental outcomes” [6,7], and is considered a crucial pillar for realizing corporate sustainability [8]. SI focuses on leveraging corporate resources and capabilities to identify existing internal and external knowledge. Better exploration and exploitation of knowledge can help enterprises achieve strategic objectives [9], namely, the knowledge ambidexterity.

Knowledge ambidexterity primarily encompasses exploration and exploitation. Knowledge exploration refers to the learning process that helps enterprises acquire/create, share, absorb, and store new knowledge; knowledge exploitation is the learning process through which enterprises reuse, transform, apply, and utilize existing knowledge and new knowledge [10]. Many researchers have argued that merely refining existing operations or developing incremental innovations based on current knowledge or technology is sufficient to enhance the knowledge ambidexterity capabilities required for SI [11,12]. As emphasized by Seuring & Gold [13] and Reficco et al. [3], the effectiveness of innovation processes hinges on a company’s ability to manage external knowledge. Firms must enhance their capacity to exploit shared information and knowledge management among partners and strengthen knowledge management capabilities [14] to cultivate synergies that advance sustainable objectives [15]. This compels enterprises to shift toward exploiting digital technologies with greater potential to access the resource and capability foundations supporting SI [2,16].

Existing research confirms that the use of DPs by enterprises significantly enhances the accessibility of corporate information, facilitates information sharing among partners, and promotes resource coordination [17,18,19]. However, McKinsey’s [20] research indicates that despite the widespread adoption of information platforms today, only a minority of enterprises demonstrate operational evidence that DP usage contributes to sustainable performance. Consequently, a disconnect persists between DP utilization and SI within organizations. In particular, Chinese enterprises that are representative in terms of scale and growth rate within industrial DPs are deeply entrenched in an efficiency dilemma concerning SI and development [21,22].

To advance SI, this study identifies DP capability—the extent to which enterprises utilize their digital information integration systems—as a prerequisite for SI. It further examines the underlying mechanisms linking DP capability to SI. Numerous factors contribute to the disconnect between DP capability and SI [7,23], with the primary factor being the isolated application of DP capability and organizational capability [21,24]. Existing research predominantly analyzes the role of either DP capabilities or knowledge absorption capacity (AC) in corporate innovation in isolation. As highlighted in reports by Salesforce [25] and McKinsey [20], such fragmented analysis and technology adoption frequently traps enterprises in dual operational dilemmas: technological drift (where information acquisition struggles to keep pace with technological change) or knowledge indigestion (information overload coupled with weak absorption capacity).

According to existing scholarly research, leveraging the flexibility and efficiency provided by technology enables enterprises to achieve better external connectivity, effective communication, and automated information processing [26,27]. Corporate innovation capabilities are enhanced through the absorption of new knowledge [16,23]. Therefore, this study proposes that DP capabilities and AC (DP&AC) jointly influence the enhancement of SI.

Based on this, the research employs an adaptive perspective to conceptualize the combined effects of DP&AC and demonstrates the mechanism through which this fit mechanism impacts corporate SI. Adaptation represents how multiple factors within an organization align so they can work together to produce beneficial outcomes [28]. Drawing on the organizational management adaptation studies by Benitez et al. [29] and Lin et al. [28] on organizational management, the study proposes two fit mechanisms for DP&AC: complementary fit and balanced fit. The fit mechanisms of DP&AC are positioned as drivers of corporate SI, and the study analyzes the mediating role of the knowledge ambidexterity between these fit mechanisms and corporate SI.

This study integrates two research perspectives on SI in enterprises—information technology and AC—thereby broadening the research framework for understanding the mechanisms of SI implementation and contributing to the field of sustainable development. The remainder of the study is structured as follows: Section 2 reviews existing theoretical frameworks and the current state of research in related fields. Section 3 presents the conceptual model and associated hypotheses. Section 4 details questionnaire design, survey implementation, sample selection, and data analysis methods. Section 5 reports the results of data analysis, while Section 6 discussions and implications of the findings. Finally, Section 7 summarizes the research and proposes directions for future research.

2. Theoretical Background and Literature Review

The concept of SI originates from eco-innovation, environmental/green innovation, and social innovation, and is grounded in all dimensions of sustainable development [2,30]. SI refers to a sustainability-oriented innovation structure designed to drive changes in processes and products based on explicit goals of creating social and environmental value, while simultaneously generating economic returns [8,31]. Consequently, Adams et al. [2] introduced and proposed a systematic framework: “Beyond operational optimization and organizational transformation lies highly radical, game-changing systemic innovation”.

Subsequently, researchers argue that the systemic perspective adopted by companies is central to achieving SI [7]. Tebaldi et al. [32] thematically described identified domains in the literature related to SI, namely barriers and motivations for implementing scientific innovation, studied innovation stages, degrees and types of innovation, and dimensions for measuring the sustainability of innovation. Other studies focus more on supply chain management systems and innovations related to products and processes [3,4,8].

Therefore, SI constitutes a core competitive advantage for enterprises [2]. For an organization, the more relevant the market development environment is to performance, the more it will strive for SI. Existing research indicates that enterprises with high levels of SI place greater emphasis on external knowledge acquisition and engage in R&D activities and collaborative agreements to seek optimal benefits [3,4,33].

2.1. Resource Orchestration Theory and Fit Theory

The resource orchestration theory posits that resources must be accumulated, integrated, and leveraged to unlock their potential for value creation [34,35]. The primary challenge in achieving resource orchestration lies in understanding how managers mobilize and organize resources [18]. The mobilized resources are integrated into an effective structure to ensure greater coherence, coordination, and purposeful direction [36]. Resources and capabilities only influence performance when structured, bundled, and utilized in ways appropriate to specific objectives [19]. This study views the pursuit of SI within organizations as a potential pathway for managers to mobilize resources [34]. More specifically, we regard SI as a channel through which an organization’s strategic sustainable knowledge resources can be effectively mobilized to create genuine value for the company [23,30].

“Fit” serves as a key viewpoint for illustrating and quantifying resource orchestration [28,37]. The concept of fit posits that suggests that to attain desired results, there must be a particular level of congruence among various elements [37]. It is equivalent to the notions of alignment and harmony [38]. Fundamentally, fit indicates the way in which numerous factors within an organization are coordinated to work in tandem and produce advantages [39]. Current studies in information systems and supply chain management utilize fit mechanisms to investigate how a stronger alignment between digital technologies and organizational objectives, characteristics, and management practices promotes sustainable performance, positing that fit mechanisms can be considered as a factor driving competitive advantage [40].

According to research on fit mechanisms in organizational management [38,40,41], the causal processes underlying different fit mechanisms may vary [41]. Therefore, we apply two fit mechanisms to coordinate and adjust DP&AC: complementary fit and balanced fit.

Complementary fit suggests that DP&AC are not in conflict but are highly compatible. It demonstrates that these capabilities are mutually compatible and even mutually reinforcing [40]. This approach focuses on enhancing the absolute scale of both DP&AC, maximizing corporate benefits by developing and leveraging complementary resources between them [42].

Balanced not only to the equilibrium or trade-off between DP&AC, but also to the proactive consideration of both [38,40]. It involves the extent of a company’s efforts to align its DP capabilities with AC, thereby mitigating risks arising from over-commitment to either and enhancing SI [36]. Balanced fit differs from complementary fit. For instance, a company’s decision to prioritize AC over DP capabilities may result in an imbalanced fit despite complementarity between the two [43].

Although complementary fit and balanced fit differ conceptually and operationally, involving distinct causal processes, they are compatible and mutually supportive [28,43]. This implies that enterprises can not only employ these mechanisms independently but also simultaneously leverage both to enhance SI performance. For instance, when firms achieve high levels of both complementary fit and balanced fit, they can secure greater benefits with limited risks.

These two types of fit mechanisms measure the combined effect of DP&AC on SI in enterprises (Figure 1). According to resource orchestration theory, DP capabilities represent the information and knowledge resources an enterprise can access, while AC signifies the information and knowledge resources the enterprise possesses and utilizes. Based on the logic of fit, the study emphasizes the complementary nature of these two and their joint impact on SI. DPs enable enterprises to achieve better information exchange and sharing, allowing real-time monitoring and acquisition of critical information. Research indicates that enterprises with higher AC respond rapidly to external environmental changes based on information [27,44], thereby enhancing SI [45].

Sustainable development necessitates adopting fresh perspectives, and a prevalent approach to driving innovation and enhancement involves generating novel knowledge and integrating it into innovative practices [8]. SI focuses on leveraging corporate resources and capabilities to identify existing internal and external knowledge. Better exploration and exploitation of knowledge can help enterprises achieve the strategic goals of SI [2,6]. Consequently, this study incorporates the knowledge ambidexterity into the research model to construct a mechanism revealing the relationship between DP capabilities and SI. To achieve this research objective, the study employs resource orchestration theory to explain the DP&AC fit mechanism and the mediating role of knowledge ambidexterity between the fit mechanism and SI.

2.2. Digital Platform

DPs themselves represent the fundamental infrastructure for e-commerce, online communication, and digital social interactions [46]. Researchers often use terms like “interaction, value creation, infrastructure, transaction, structure, exchange, and mediation” to highlight the characteristics of DPs. The influence of DPs is expanding to connect and facilitate various types of businesses within society [47], from fin-tech to e-commerce, with their value being undeniable [23]. It dissolves boundaries within the digital economy, functioning as an integrated operating system for business operations and management [23,47]. According to Parker et al. [48], DPs are described as information-based platforms that facilitate business interactions between external producers and consumers.

DPs serve as an open, participatory infrastructure that mediate interactions and exchanges, creating value for all participants [23,47]. In the B2B context, DPs provide an infrastructure to facilitate interaction between buyers and sellers in two-sided markets [17]. Through extensive participation by business partners, DPs enable buyers to gather rich market information. This allows firms to efficiently search for and identify suitable partners [23,49], gaining access to more business opportunities than physical markets at relatively lower costs [26]. Consequently, as enterprises increasingly rely on these DPs, many now utilize them as sources for connecting and collaborating with business partners to develop new products, rather than merely selling goods [19,50].

With the widespread adoption of DPs, the focus of value creation has shifted from traditional linear value chains to interwoven networks [1,16]. The significance of DPs has risen from the functional IT level to the strategic and managerial levels [27]. Engaging the resources and capabilities of the entire organization, DPs have progressively transformed into virtual venues for innovation activities such as new product development [17,50]. DPs streamline corporate connections with external entities, enabling access to diverse external knowledge resources [26,51]. They also allow organizations to rapidly encode, store, formalize, and distribute an increasing volume of knowledge, leading to greater knowledge diversity throughout the system [17,46]. However, the sheer volume of knowledge information can also severely disrupt existing business models. How to leverage acquired information resources to achieve business objectives remains a topic worthy of in-depth exploration [16,35]. Therefore, investigating the fit mechanism between DP&AC to achieve SI through knowledge development and utilization is essential.

2.3. Absorptive Capability

The ability to recognize the value of external information is a crucial component in developing innovative capabilities. AC refers to an organization’s ability to acquire, assimilate, transform, and utilize knowledge [12,45,52]. Zahra and George [53] argue that AC manifests through the simultaneous presence of four core elements: two representing latent AC (knowledge acquisition and knowledge assimilation), and two representing realized AC (knowledge transformation and knowledge exploitation). Potential AC may be particularly strong in companies possessing organizational mechanisms such as job rotation, cross-functional interfaces, and participatory decision-making [10,54].

Once an enterprise acquires and absorbs external knowledge (potential AC), it must transform and utilize this knowledge. This involves integrating existing knowledge with new knowledge and incorporating the new knowledge into operations to generate new insights and outcomes [12,53,55]. Realized AC enables companies to innovate and create value [54]. Recent studies have found that realized AC is a prerequisite for developmental and exploratory innovation, green product and process innovation [44], as well as new business creation and self-renewal (e.g., strategic repositioning, business redefinition) [45,56]. As the ability to recognize environmental changes, AC directly influences organizational performance by facilitating the deployment of necessary capabilities to renew knowledge and skill base, and by enabling the most flexible use of resources and competencies [45].

The ability of firms to acquire new knowledge depends on robust search practices and corresponding knowledge conversion functions [57]. Therefore, they need to combine resources and capabilities to maintain a strong AC [58]. However, developing dynamic capabilities proves particularly difficult and challenging for small enterprises [59]. This also presents a challenge for managers. They must be capable of integrating diverse technologies and tools to adapt to new social developments [10,47]. In other words, only companies possessing robust capabilities to acquire and absorb knowledge based on digital technologies can achieve high levels of strategic flexibility. This enables them to respond strategically to business risks and opportunities [45,55]. These findings underscore the significant value of researching how the fit mechanisms of DP&AC enhance SI levels in enterprises.

3. Hypothesis Development and Model Construction

3.1. The Fit Mechanism Between DP&AC and SI

3.1.1. Complementary Fit and SI

The mutual compatibility of DP&AC, which generates a multiplier effect for core enterprises, is referred to as a complementary alignment mechanism and is termed complementary adaptation. This adaptation mechanism emphasizes the impact of DP–AC synergies on SI. Firstly, DPs can capture and absorb external information within tools such as enterprise resource planning, supply chain management, or customer relationship management [1], supporting the utilization and absorption of systematic knowledge for innovation purposes [35,46]. Second, firms’ AC enables the development and exploration of external knowledge, facilitating the transformation from knowledge to innovation [42]. This fit mechanism highlights how DPs can help leverage and maximize the effectiveness of AC, thereby generating a joint effect on SI [33,59]. Thus, DPs create conditions that enhance AC and its effectiveness, amplifying the role of AC in SI.

H1. 

The complementary fit (degree) between an enterprise’s DP&AC has a positive impact on its SI.

3.1.2. Balanced Fit and SI

Unlike complementary fit between DP&AC, balanced fit mechanisms place greater emphasis on maintaining a close relative match in their scale. Specifically, a higher degree of balanced fit—or a tighter equilibrium between DP&AC—enhances SI performance by mitigating the potential risks arising from over-commitment to or neglect of either capability [28]. In other words, when firms overemphasize AC while neglecting DP capabilities, the external knowledge and resources delivered through DPs diminish. This may result in limited access to broad external knowledge channels [42], hindering effective knowledge acquisition and assimilation, thereby impeding SI [27]. Conversely, when a firm’s DP capabilities far exceed its AC, the lack of AC may result in the firm acquiring a vast amount of knowledge and resources through DPs yet failing to transform, assimilate, and utilize useful knowledge [35,48]. In such cases, the firm not only misses out on reaping the advantages of the synergistic interaction between DP&AC but also faces the danger of substantial resource inefficiency caused by a severe disparity between them. On the other hand, when firms achieve a well-balanced alignment (i.e., a close equilibrium between DP&AC), they can avoid this risk while simultaneously enhancing their SI performance. Given the preceding discussion, we put forward the following proposition:

H2. 

The balanced fit (degree) between DP&AC has a positive impact on its SI.

3.2. The Mediating Role of Knowledge Ambidexterity

In today’s context, knowledge stands as a crucial resource that enables enterprises to continuously transform acquired knowledge through dynamic capabilities, thereby creating economic, social, and environmental value [55]. The role of DP capabilities in facilitating exploration and exploitation closely mirrors the dual-process perspective on learning and innovation [41]. This ambidexterity integrates exploration and exploitation, though the former requires openness and divergence while the latter demands closure and convergence [9]. Technological capabilities that foster exploratory and exploitation learning promote ambidexterity, thereby helping enterprises identify, evaluate, and select external information and technologies, that is, AC. Simultaneously, technological capabilities may also be associated with firm performance [52]. For instance, within complex innovation contexts, knowledge ambidexterity was found to enhance complex innovation benefits alongside AC while mitigating costs associated with increased complexity [57,58]. Technological capabilities that foster duality positively contribute to both innovation performance and financial performance [10]. Innovation capability can indeed be regarded as a significant contribution to knowledge management [12]. Dualistic enterprises can continuously refine existing processes while acquiring new alternatives. Thus, firms that both explore and exploit can maximize innovation [9]. Research has examined how exploration and exploitation jointly influence firm performance in the context of technological innovation.

The alignment between a firm’s DP&AC facilitates knowledge ambidexterity, thereby enhancing SI performance [12]. As firms develop their technological capabilities, they become more likely to absorb new external information [18]. AC further enhances the firm’s ability to identify emerging technological developments and trends, thereby strengthening knowledge exploration and exploitation capabilities. It also accelerates a firm’s ability to discover new opportunities and respond to technological changes [55]. Therefore, the higher the complementarity fit between DP&AC, the stronger the firm’s knowledge ambidexterity [52].

Furthermore, exploration enables firms to enter diverse technological domains, increasing the diversity and heterogeneity that foster the creation of new knowledge. This newly generated knowledge can then be exploited to develop more effective innovations [42]. Companies that better reuse, apply, and leverage existing/new knowledge (i.e., knowledge exploitation) can outperform competitors by more effectively transforming existing products/processes, thereby enhancing the firm’s SI [10].

Simultaneously, the development of corporate DP capabilities involves the accumulation and storage of knowledge. This not only increases its absorption capacity but also increases its participation in knowledge and experience through the evaluation, utilization, and implementation of new technologies. Consequently, enterprises become more effective in deploying and leveraging existing knowledge [41,46]. In other words, the greater the fit between an enterprise’s DP&AC, the more it facilitates knowledge exploration and exploitation. Higher DP capabilities in a specific domain strengthen an enterprise’s knowledge accumulation, reinforcing its self-learning capacity and promoting SI through knowledge exploitation [12,52]. Therefore, we propose the following hypotheses:

H3. 

Knowledge ambidexterity mediates the relationship between complementary fit (degree) and supply chain sustainable innovation.

H4. 

Knowledge ambidexterity mediates the relationship between balancing fit (degree) and supply chain sustainable innovation.

Based on the above four hypotheses, the study constructed a comprehensive model (Figure 2). In addition to the primary outcome variables and mechanisms, this research also included firm size, firm age, size of the firm’s technology department, and average annual number of technical training sessions. These were introduced to mitigate the impact of extraneous factors on the model’s forecasting accuracy. Firstly, we took into account firm size, given that larger enterprises are perceived to possess redundant resources for innovation. An equally reasonable expectation is that more mature firms command greater resources and experience. Similarly, we also control for the frequency of IT skills training within the company, as such training may translate into unique capabilities for leveraging IT tools to foster innovation. The size of the IT department can reflect variations in the extent to which new technologies enhance innovation. Detailed descriptive statistics for these control variables are outlined in

Table 1. Measurement.

Construct	Items	Loading
Digital platform capability	Our platform facilitates easy access to data from our partners’ IT systems.	0.957
	It ensures seamless integration between our partners’ IT systems and ours, covering areas such as forecasting, production, manufacturing, and shipping.	0.855
	The platform is capable of exchanging real-time information with our partners.	0.825
	It aggregates relevant data from our partners’ databases, such as operational data, customer performance, and cost details.	0.858
	The platform is flexible, allowing easy adaptation to include new partners.	0.842
	It can be effortlessly extended to incorporate new IT applications or features.	0.869
	The platform adheres to widely accepted standards recognized by most current and potential partners.	0.854
	It is built with modular software components, many of which are reusable across other business applications.	0.853
Absorptive capability	In our enterprise, management supports the development of prototypes.	0.956
	In our enterprise, management regularly reconsiders technology and adjusts it based on new knowledge.	0.878
	In our enterprise, management has the ability to work more efficiently by adopting new technologies.	0.863
	In our enterprise, employees have the ability to organize and utilize the knowledge they have gathered.	0.849
	In our enterprise, employees are accustomed to absorbing new knowledge, preparing it for further use, and making it available.	0.827
	In our enterprise, employees successfully link existing knowledge with new insights.	0.863
	In our enterprise, employees are able to apply new knowledge to practical work.	0.846
Knowledge ambidexterity	Employees apply their knowledge and skills in activities focused on incremental improvements and problem-solving.	0.953
	Employees leverage their expertise to resolve problems.	0.854
	The company encourages incremental process improvements through a program that gathers employee ideas and suggestions.	0.830
	The company gains easy access to new technologies through partnerships with other companies, universities, consulting firms, etc.	0.838
	The company invests in R&D to enhance or develop new products and processes.	0.864
	The company can introduce new technologies into its processes or products with minimal resistance to change.	0.864
Sustainable innovation	Our company selects product materials that generate the least pollution during product development, design, and manufacturing.	0.971
	Our company chooses product materials that consume the least energy and resources during product development, design, and manufacturing.	0.872
	Our company uses the minimum amount of materials to compose products during product development, design, and manufacturing.	0.842
	Our company carefully considers whether products are easy to recycle, reuse, and decompose during product development, design, and manufacturing.	0.877
	Our company effectively reduces the emission of harmful substances or waste during product development, design, and manufacturing.	0.863
	Our company recycles waste and emissions during product development, design, and manufacturing so that they can be treated and reused.	0.844
	Our company reduces water and energy consumption during product development, design, and manufacturing.	0.840
	Our company reduces the use of raw materials during product development, design, and manufacturing.	0.853

.

4. Methodology

4.1. Composite Construct and Measurement

The measurement tools in this study comprise two main components. First, we employed construct-specific measurement scales to assess DP capacities, AC, knowledge ambidexterity, and SI (Table 1). Subsequently, we calculated complementary fit and balanced fit for DP&AC using an artificially constructed method. All structural measurement items incorporated in this study were adapted from the existing literature to align with the theoretical framework of this research, thereby ensuring content validity.

In this study, DP refers to digital information technologies that support information exchange activities with partners. DP capability reflects an enterprise’s ability to leverage DPs for acquiring external resources to create competitive advantages. The measurement of DP capability is based on an enhanced version of the measurement tool developed by Cenamor et al. [17], utilizing 8 items to reflect the level of capability in applying DPs in management practices. AC emphasizes the internalization and exploitation of information within an organization. To measure AC, we adopted the scale developed by Flatten et al. [54]. This measurement comprises 14 items covering four dimensions: knowledge acquisition, assimilation, utilization, and transformation.

We drew upon the measurement tools used in Gonzalez and de Melo’s [9] study under similar circumstances to reflect knowledge ambidexterity. SI is measured through 8 items based on Du and Wang [56]. These items require participants to evaluate how extensively environmental criteria are incorporated into the development of new products or in designing products that support recycling, reuse, and decomposition, thereby yielding ecological advantages. Additionally, respondents assess the degree of material recovery, energy conservation, and environmental pollution reduction during production processes

The study employs the ‘artifact’ construct type of variables to measure the two types of DP&AC fit outcomes. While latent variables are generally conceptualized as behavioral constructs inferred from observable indicators, artifacts represent designed structures composed of multiple components [29]. We used a sum-of-logarithms approach to test the complementary fit between DP&AC, using the mathematical formula “DP&AC complementary fit = |ln (dp + ac)|”. For measuring balanced fit, we adopted “DP&AC balanced fit = |ln 1/|dp − ac||”. Absolute difference is a prevalent method for achieving a matching fit in scholarly works [38,40]. Specifically, the absolute difference |dp − ac| quantifies the proximity between DP&AC, thus serving as a fit metric. According to Venkatraman [37], a value closer to zero indicates a better match between DP&AC. Following prior research by Chatterjee et al. [38], we considered that low absolute differences between latent variable scores would denote a high level of fit. Consequently, we inverted the absolute difference between scores (1/|dp − ac|); however, this adjustment gave disproportionately large weights to scores differing by less than one standard deviation. Due to the non-normal distribution of these values, we subsequently applied a normalized logarithmic transformation, leading to variables that closely resembled a normal distribution [28].

4.2. Data Collection

To validate the above research hypotheses, the study selected Chinese enterprises with relatively high digital platform application maturity for investigation, ensuring that corporate respondents possessed the necessary knowledge to answer the questions [43]. The choice of Chinese enterprises as the sample is not only due to the global representativeness of digital platforms in terms of their scale of adoption and development speed within these companies, but also because China, as the world’s largest renewable energy market, requires enterprises to simultaneously integrate domestic technologies (e.g., photovoltaic silicon materials) with global knowledge (e.g., European energy storage technologies) to achieve co-creation of value [22]. This creates fertile ground for testing and developing dual-source knowledge and sustainable innovation. The findings of this study will provide a Chinese solution for sustainable innovation development in enterprises from developing countries.

The study was conducted between August and September 2023. The survey yielded a total of 328 complete responses. During balanced fit model transformation, |dp − ac| = 0 precluded its inclusion as a denominator in calculations. Additionally, four samples yielding zero values after balanced fit transformation were excluded, leaving 324 valid samples for analysis. The sample size in this study fully complies with the recommended empirical rule: 10 times the number of indicators in the structure with the most indicators [60].

We employed a survey questionnaire as the instrument for variable measurement. The questionnaire was structured into two parts: the first part collected fundamental details about the respondent enterprises (including their experience with digital technology adoption, company size, and company age), and another containing indicators measuring each variable in our research model. The descriptive statistics of sample are shown in

Table 2. Descriptive statistics of sample.

Sample Characteristics		Number	Percent (%)	Sample Characteristics		Number	Percent (%)
Enterprise size	<50	29	8.95	Enterprise technical training (times/year)	0	30	9.26
	50–99	109	33.64		1–2	112	34.57
	100–199	106	32.72		3–5	106	32.72
	200–500	59	18.21		6–10	60	18.52
	>500	21	6.48		>10	16	4.93
Years in business	<1	39	12.04	Technical personnel in the enterprise	<5	32	9.88
	1–5	151	46.60		5–20	155	47.84
	5–10	103	31.79		21–50	131	40.43
	>10	31	9.57		>50	6	1.85

. These indicators were evaluated using a five-point Likert scale. The questionnaire was translated into Chinese by a doctoral candidate and subsequently back-translated into English by multiple researchers. The translated English scales were compared against the original English versions, with minor adjustments made based on these comparisons to minimize linguistic differences and ensure semantic accuracy. An expert and several company managers were invited to pretest the questionnaire. Based on their feedback, the scales were revised to enhance precision and clarity of expression.

4.3. Common Method Bias

Considering the study’s use of a cross-sectional survey approach with a single respondent per case, there is a potential risk of common method bias (CMB). To mitigate this, we started the questionnaire by offering a thorough description of digital technologies and interspersed demographic questions randomly among the main items. These strategies were effective in reducing CMB throughout the data collection process. Second, we examined model fit indices for both sets of models using confirmatory factor analysis (

Table 3. Result of CMB test.

	χ2	df	χ2/df	CFI	TLI	SRMR	RSEAM
Model 1	468.057	371	1.262	0.990	0.989	0.028	0.031
Model 2	468.983	373	1.257	0.990	0.989	0.028	0.032

). In the first model, each item loaded onto its respective factor. In the second model, a common factor connecting all items was added as a homogenous bias factor. Compared to the first model, the RMSEA and SRMR decreased below 0.05, while CFI and TLI changed by less than 0.02. Thus, there is sufficient justification to conclude that CMB is not an issue in our data or estimations.

4.4. Measurement Model Assessment

Table 4. Test of validity and reliability.

Test of Data Validity and Reliability								HTMT Ratio
	α	CR	AVE	1	2	3	4	1	2	3	4
1. DP	0.959	0.980	0.748	0.865
2. AC	0.956	0.981	0.756	0.510	0.869			0.516
3. KA	0.947	0.980	0.754	0.461	0.514	0.868		0.465	0.513
4. SI	0.961	0.981	0.759	0.451	0.450	0.439	0.871	0.458	0.456	0.440

Abbreviations: KA, knowledge ambidexterity.

presents the extracted average variance explained (AVE), Cronbach’s alpha (α), and composite reliability (CR). According to the criteria defined by Hair et al. [60], these values demonstrate satisfactory validity and reliability.

Discriminant validity (DV) refers to the extent to which a set of items can distinguish one variable from others. In Table 3, diagonal values represent DV. According to the Fornell-Larcker criterion, the square root of the AVE for each construct should exceed the square of the correlation between any pair of latent constructs. As an additional measure, HTMT test was further conducted to ensure their values did not exceed 0.85 (Table 3). These results support the adequacy of the items used to measure the constructs in our analysis [61]. They also confirm the relevance of the constructs employed in the model.

5. Hypothesis Testing

After validating the model’s validity, the proposed hypotheses were examined. Results are presented in

Table 5. Hypotheses testing results.

Model	Coef.	S.E.	z	p
Direct effect
SI ← cf	1.298	0.199	6.51	0.000
SI ← bf	0.158	0.066	2.38	0.017
SEM test
KA ← cf	1.836	0.159	11.57	0.000
SI ← KA	0.199	0.061	3.25	0.001
SI ← cf	1.021	0.214	4.77	0.000
KA ← bf	0.151	0.068	2.21	0.027
SI ← KA	0.303	0.058	5.23	0.000
SI ← bf	0.122	0.064	1.91	0.056

Abbreviations: KA, knowledge ambidexterity; cf, complementary fit; bf, balanced fit.

using statistical representations. Findings indicate that both fit modes between DP&AC effectively enhance SI (SI ← cf, Coef. = 1.298, p < 0.005; SI ← bf, Coef. = 0.158, p < 0.050). H1 and H2 are supported. Furthermore, the structural equation model results analysis indicates that knowledge ambidexterity plays a significant mediating role in the relationship between the two DP&AC fit mechanisms and SI. To further examine the mediating mechanism of knowledge ambidexterity, the study conducted a bootstrapping analysis with 2000 resamples to assess the indirect effect. Results indicate (

Table 6. Bootstrap test.

	Coef.	S.E.	z	p
cf—KA—SI
_bs_1(ind_eff)	0.277	0.095	2.91	0.004
_bs_2(dir_eff)	1.021	0.255	4.01	0.000
bf—KA—SI
_bs_3(ind_eff)	0.035	0.019	1.84	0.065
_bs_4(dir_eff)	0.123	0.058	2.12	0.034

Abbreviations: KA, knowledge ambidexterity; cf, complementary fit; bf, balanced fit.

) that knowledge ambidexterity significantly mediates the relationship between the complementary fit mechanism of DP&AC and SI (bs_1, Coef. = 0.277, p < 0.005). Therefore, H3 holds. Conversely, H4 is not supported. This is because knowledge ambidexterity’s mediating effect in the relationship between the DP&AC balanced fit mechanism and SI is not significant at the 95% confidence level (bs_3, Coef. = 0.035, p = 0.065).

6. Result Discussion and Implications

6.1. Result Discussion

H1 and H2 are supported by the data. This finding validates the core theoretical proposition of this study: that a firm’s DP&AC are mutually embedded and jointly contribute to SI growth. Unlike existing studies that treat organizational dynamic capabilities as mediators between technological capabilities and sustainable performance [16,21] or as contextual conditions [5,27], this study posits the DP&AC fit mechanisms as a unified representation of a firm’s information processing capability. This perspective aligns strongly with Barney et al.’s [62] recent call for recognizing that a company’s resources and capabilities generate value more efficiently when linked through a series of contracts. The study demonstrates through data analysis that both fit mechanisms exert a positive and consistent impact on corporate sustainable innovation. This also provides a positive dialectic on the causal relationship between technological resources and corporate sustainable performance [21]. The research will contribute to a deeper understanding of the alignment between digital resources and organizational dynamic capabilities, as well as their underlying mechanisms in driving corporate to SI.

Specifically, inconsistent conclusions emerged regarding the mediating role of the knowledge ambidexterity between the two fit mechanisms of DP&AC and SI. H3 supports the complementary relationship between DP&AC. Their interaction enhances the overall knowledge management capabilities of enterprises, providing fertile ground and robust tools for knowledge ambidexterity activities. On one hand, DP capabilities enhance enterprises’ knowledge exploration capabilities through information collection and analysis [24]. On the other hand, AC strengthens the firm’s ability to internalize and exploit new information and knowledge [42]. Consequently, complementary fit naturally drives SI by elevating the firm’s level of knowledge ambidexterity [18]. This conclusion bridges the gap between the DP&AC fit mechanism and corporate sustainable innovation, while also providing reference evidence for research on the relationship between the DP&AC fit mechanism and organizational capabilities (such as organizational flexibility) and competitive advantages (such as corporate resilience and sustainable performance).

However, H4 did not provide significant support for the mediating role of knowledge ambidexterity between the DP&AC balanced fit mechanism and SI. This result was unexpected yet plausible. Balanced adaptation in DP&AC may represent a higher-order organizational state. When both DP capabilities and AC reach advanced levels, knowledge ambidexterity can proceed with low losses and high efficiency. At this stage, the balanced fit of DP&AC functions more as a contextual condition, yet its impact on SI may not be direct. Therefore, future research should validate the mediating role of the DP&AC balanced fit mechanism in the process of achieving sustainable innovation within enterprises. Conversely, DP capabilities—with their inherent network and information processing capacities—exert direct effects on enhancing production processes and corporate innovation [15,16]. In other words, the influence mechanism of DP&AC balanced fit may bypass knowledge itself, acting upon corporate SI performance through other more critical pathways. Therefore, a potentially stronger mediating mechanism than the knowledge ambidexterity may exist between the DP&AC balanced fit model and SI, awaiting further exploration.

Comparing the findings from H3 and H4, DP&AC’s complementary fit mechanism focuses on addressing weaknesses and activating the knowledge ambidexterity base of enterprises to achieve SI. The balanced adaptation mechanism emphasizes the construction of a top-level infrastructure to indirectly support the low-consumption and high-efficiency realization mechanism of enterprise sustainable innovation. It even acts on corporate sustainable innovation by activating other unknown mechanisms.

6.2. Theoretical Implications

First, this study employs resource orchestration theory and the fit mechanisms to theoretically elucidate the joint impact of the fit mechanism between DP&AC on SI. Previous research has predominantly focused on the unilateral effects of single factors on SI, overlooking the actual composition and configuration of multiple resources that may drive SI [30]. Consequently, this study diverges from the prior literature examining either the direct effects or simple interaction effects of DP&AC. It offers new insights into multidimensional mechanisms (such as complementary fit and balanced fit) that combine diverse resources to enhance SI.

Second, the study demonstrates that the impact of the fit mechanisms between DP&AC on SI is not limited to a single pathway. It challenges the myopic view of fit (i.e., simplistically interpreting fit as merely mediating or moderating) [39], offering new insights into understanding the formation of SI. The findings indicate that complementary fit and balanced fit influence outcome variables through distinct mechanisms, contributing to the advancement of fit theory.

Third, this study delimits the mediating boundary conditions of knowledge ambidexterity by uncovering its critical mediating role within the context of DP&AC complementary fit. Simultaneously, it demarcates the scope of its explanatory power by demonstrating that this mediating effect exhibits no significant association in scenarios of DP&AC balanced fit and SI. These findings offer a more nuanced theoretical framework for future research in selecting appropriate mediating variables. Furthermore, the conclusion suggests that when organizational capabilities reach an exceptionally advanced level, they may transcend the conventional ambidexterity paradox and attain a novel state of hyper-ambidexterity. This observation provides intriguing material for discussions on dualistic organizational theory research and will also inspire studies on hyper-dualistic organizational capabilities.

6.3. Management Implications

This study primarily encompasses the following key management implications. Enterprises can build SI by implementing strategies that complement and balance DP capability with AC.

First, companies should regularly assess their levels of DP capability and AC. Where significant gaps exist, targeted strategies, such as introducing new technology platforms or training employees’ skills, should be adopted to rapidly stimulate SI through complementary fit. Simultaneously, enterprises should treat both capabilities as long-term strategic investments, pursuing their synergistic advancement. Cultivating a balanced, robust capability portfolio establishes the foundational environment and core competitiveness for SI that is difficult to replicate.

Second, enterprises should select distinct capability-building paths based on their developmental stage and resource endowments. Regarding the prioritization and selection of adaptation strategies, our research indicates that the complementary fit strategy exerts a stronger influence on SI than the balanced fit strategy. This implies that under resource and capability constraints, enterprises should prioritize the complementary fit strategy to enhance SI, placing balanced adaptation strategies in secondary importance. Large enterprises with robust resources should adopt a long-term perspective, investing in the balanced development of DP&AC to build enduring competitive advantages.

Third, to fully leverage their existing resource endowments, enterprises must first establish structural enablers that promote bidirectional communication and organic integration between DP&AC. This can be achieved through institutionalizing cross-functional digital teams or developing enterprise-wide data governance frameworks that enable seamless data interoperability across internal operational silos and external ecosystem partners.

7. Conclusions and Limitations

Based on resource orchestration theory, this study examines the impact of DP capability and AC on SI through two fit mechanisms. Findings indicate that both fit mechanisms enhance knowledge ambidexterity capabilities and SI. Knowledge ambidexterity exerts a significant mediating effect only under the complementary fit of DP and AC, while its mediating role is insignificant under the balanced fit model. This research holds theoretical value and practical significance for advancing studies on digital transformation and SI. However, the study has certain limitations and indicates topics worthy of further research.

First, given the prevalence and maturity of digital platforms, this study employs Chinese enterprises as the data sample for validating the theoretical model. Our findings may thus be region-specific. Future research may need to examine dynamic panel data from different regions or data samples from other emerging economies to broaden the applicability boundaries of the relationship between the fit mechanism and corporate sustainable innovation.

Second, the study developed a theoretical model based on resource orchestration theory. This model does not encompass all pathways in the relationship between the DP&AC fit mechanisms and SI. Based on different research theories, future studies may uncover additional mechanism models within this relationship. These could include the mediating role of organizational resilience and agility in the DP&AC balanced fit mechanism and SI, as well as the moderating effects of contextual factors like environmental dynamics and industry competition on the relationship between different fit mechanisms and SI. Furthermore, based on the findings from the study on the DP-AC balanced fit model, this research explores the stronger mediating mechanisms between the DP-AC balanced fit model and corporate sustainable innovation.

Finally, using cross-sectional data to characterize the fit between DP&AC has limitations. Employing longitudinal case studies to track one or more firms could provide deeper insights into the relational mechanisms between DP&AC fit mechanisms and SI. Furthermore, future research could categorize samples into four groups based on high and low scores across the two dimensions of DP&AC. Comparative analysis of these four sample types would deepen the understanding of the relational logic between digital technology and SI growth, thereby enhancing the comprehensive comprehension of the DP-AC fit mechanism. Research into the fit mechanisms of four sample categories will further assist corporate managers and researchers in understanding the alignment between digital platform usage and organizational dynamic capabilities in creating corporate value.