Digitalization, Sustainability, and Radical Innovation: A Knowledge-Based Approach

Abstract

This study investigates the interplay between digitalization capability, environmental sustainability perception, and radical innovation performance with a particular focus on the mediating roles of knowledge integration capability and knowledge accumulation. The study utilizes Structural Equation Modeling and the Hayes PROCESS Model to analyze data from 315 firms in technology-driven industries. The findings reveal that digitalization capability significantly enhances radical innovation performance (β = 0.767, p < 0.001, R² = 0.589), while environmental sustainability perception does not directly influence innovation performance nor mediate its relationship with digitalization. However, knowledge integration capability and knowledge accumulation emerge as critical enablers, strengthening the effect of digitalization on innovation outcomes. Moreover, knowledge integration capability positively moderates the digitalization capability–radical innovation performance relationship, demonstrating that firms with higher knowledge integration capability derive greater innovation benefits from digital transformation. In contrast, environmental sustainability perception does not moderate this relationship, suggesting that sustainability perception alone is insufficient to drive radical innovation. The findings provide insights for firms leveraging digitalization to drive innovation and efficiency. Knowledge integration and accumulation are key to sustaining competitive advantage. These results contribute to the literature on digital transformation, innovation management, and sustainability, highlighting the necessity of knowledge-driven mechanisms in leveraging digitalization for innovation success. This study offers valuable managerial insights by highlighting the strategic significance of knowledge integration and accumulation in enhancing the effectiveness of digital transformation on innovation performance. Future research should explore longitudinal dynamics, sectoral variations, and additional moderating factors such as digital leadership and organizational culture to deepen the understanding of this evolving field.

1. Introduction

In today’s dynamic business landscape, digitalization has emerged as an essential strategic factor for organizations seeking to achieve and sustain a competitive advantage [1,2]. The development of digitalization capability (DC) provides firms with significant advantages, such as agile management, big data analytics, and automation-based processes, while also serving as a critical tool for achieving environmental sustainability goals [3]. Recent studies emphasized that digital transformation plays a key role in promoting sustainable business practices and competency frameworks, aligning with global initiatives such as the European Green Deal [4]. As firms undergo digital transformation, they benefit from increased automation, process efficiency, and enhanced customer engagement [5]. The ability to integrate and utilize digital tools allows businesses to adapt to dynamic markets, optimize operational workflows, and create competitive advantages in rapidly evolving industries [6]. Moreover, DC contributes to environmental sustainability by reducing resource consumption and enabling smart, energy-efficient solutions [7]. However, the structure of firms’ DC, the role these capabilities play in sustainable innovation processes, and the methods for evaluating these processes within the context of knowledge management remain subjects of ongoing debate [8].

This study aims to explain the impact of digital transformation on sustainable innovation from a knowledge-based perspective by examining the relationships between DC, environmental sustainability perception (ESP), radical innovation performance (RIP), and knowledge management variables. Specifically, the study addresses the role of knowledge integration capability (KIC) and knowledge accumulation (KA) in firms’ digital transformation processes, the mediating effect of ESP in digital innovation processes, and whether the impact of DC on radical innovation follows a linear or nonlinear trend.

Previous studies have broadly examined the impact of DC on innovation processes; however, they have not sufficiently detailed how this relationship is shaped within the context of environmental sustainability [1,9]. While sustainability perception encourages firms to adopt greener practices, its direct influence on radical innovation remains inconclusive. Some studies suggest that sustainability efforts primarily drive incremental innovation rather than disruptive changes as firms often focus on compliance-driven strategies rather than transformational breakthroughs [10,11]. Additionally, the costs and uncertainties associated with sustainability-oriented innovation may discourage radical transformation unless strong regulatory or market incentives are present [12]. These findings align with our results, which indicate that ESP does not significantly influence RIP, emphasizing the need for firms to integrate sustainability goals with broader knowledge management and digitalization strategies to achieve meaningful innovation outcomes. Furthermore, knowledge management and KIC are considered critical factors shaping the impact of digitalization on innovation processes [13,14]. Nevertheless, the mediating role of knowledge management between environmental sustainability and innovation has not yet been comprehensively examined [8,15].

The primary contribution of this study is to analyze the relationship between DC and radical innovation from a sustainability perspective and to reveal how this relationship is shaped by knowledge integration and ESP. In this context, the study aims to address key questions, including how DC influences ESP, whether ESP acts as a facilitator in enhancing RIP, whether the impact of DC on radical innovation follows a linear pattern or diminishes beyond a certain threshold, and how KIC and KA mediate the relationship between DC and innovation performance.

To develop the theoretical framework for this study, validated scales designed to examine the relationships among DC, innovation, and knowledge management were employed. In particular, the scale introduced in [16] forms the basis of this research, as it specifically assesses the impact of DC on innovation. By integrating knowledge integration and KA variables, this scale offers a robust theoretical structure, making it an appropriate tool for evaluating the role of digital competencies in innovation processes.

This study contributes to the literature by offering a knowledge management perspective on the underexplored relationship between DC, environmental sustainability, and radical innovation. In particular, it provides a theoretical framework for understanding the relationship between digitalization and sustainable innovation and empirically tests how this relationship is shaped by knowledge management components (KIC and KA). The study examines how ESP functions as both a mediator and a moderator in the impact of DC on innovation, and it clarifies the divergent findings in the literature by testing whether the effect of DC on radical innovation follows a linear or an inverted U-shaped pattern [17,18].

Moreover, by utilizing contemporary data collected from technology development centers in Türkiye, the study enhances its applicability and provides valuable managerial insights. This article aims to offer both theoretical and practical contributions by comprehensively examining the relationship between DC, sustainability, and innovation in the literature. It first presents an extensive literature review on the relationships between DC, environmental sustainability, radical innovation, and knowledge management. Subsequently, the research hypotheses and theoretical model are detailed, followed by an explanation of the methodology, measurement scales, data collection process, and analytical techniques used. The empirical analysis results will be used to test the hypotheses, and the theoretical and managerial contributions of the findings will be discussed. Finally, the study concludes with results and recommendations for future research.

2. Conceptual Framework and Theoretical Background

The interaction between digital transformation, sustainability, and innovation has become one of the key factors shaping the competitive advantages of modern enterprises. As businesses enhance their DC, they become more efficient in their operational processes and manage their sustainability strategies more effectively [1]. However, how this process is associated with knowledge management, ESP, and RIP remains a critical research gap that requires further exploration in the academic literature [3].

This section examines the theoretical framework of this study by analyzing the relationships among its core components: DC, ESP, RIP, and knowledge management.

2.1. Digitalization Capability and Digital Transformation in Businesses

DC is defined as a firm’s ability to adopt digital technologies to optimize processes, create new value, and gain a competitive advantage [1]. Digitalization processes encompass the utilization of technologies, such as big data analytics, artificial intelligence-driven business models, and automation, while simultaneously requiring organizations to integrate these technologies with knowledge management and sustainability strategies [9].

DC generally has a positive impact on firms’ innovation performance as they enhance strategic innovation orientation, improve process efficiency, and increase responsiveness to customer needs [19]. Digital transformation enables organizations to integrate more data-driven decision-making mechanisms into their innovation processes and develop new business models [1]. In this context, previous studies found that DC enhances firms’ radical innovation capacity and improves their technological adaptation abilities [8,20].

However, some studies suggested that the effect of DC on innovation performance varies depending on firm size, sectoral competition dynamics, and organizational agility [9]. For instance, small and medium-sized enterprises (SMEs) face greater challenges in enhancing their innovation performance due to limited investment power in digital transformation compared to large firms [18]. Nevertheless, firms with a strong digital infrastructure and a strategic innovation vision can leverage their DC to enhance their RIP [3].

In this regard, our study aims to test the direct impact of DC on radical innovation under H3. The link between firms’ digitalization efforts and their innovation capabilities can differ based on the pace of technological adoption and the effectiveness of their knowledge management practices. While the literature emphasizes that DC is a key driver of innovation, this effect is also contingent on organizational agility and the efficiency of the technology integration process [17]. Therefore, our study will analyze the direct impact of DC on innovation performance and discuss whether sectoral differences play a determining role.

In this study, DC is measured using a multi-item scale adapted from prior research, assessing firms’ ability to integrate and utilize digital technologies in their business processes [16].

2.2. Environmental Sustainability Perception and Green Transformation in Businesses

ESP is defined as the awareness of businesses regarding minimizing their environmental impact and developing sustainable business strategies [8]. The integration of sustainability strategies with digital transformation processes has become increasingly significant and is now recognized as a critical factor shaping firms’ innovation performance [10].

Studies have suggested that firms embracing environmental concerns can foster radical innovation [11]. However, industry-specific factors play a critical role in determining whether sustainability efforts translate into radical or incremental innovation. For instance, in highly regulated industries such as energy and manufacturing, sustainability-driven innovations often focus on efficiency improvements and regulatory compliance rather than disruptive technological breakthroughs [10]. Conversely, in sectors with strong market-driven sustainability incentives, such as electric vehicles and biotechnology, firms may pursue more radical transformations [12]. This distinction suggests that the impact of ESP on RIP is contingent on industry dynamics, regulatory pressures, and market demand for disruptive green technologies. This is because an environmental sustainability approach encourages businesses to develop more efficient production techniques, explore new market opportunities, and promote green innovation [12]. Therefore, within the scope of H2, this study will test the impact of ESP on RIP.

However, the relationship between DC and environmental sustainability remains insufficiently explored [8]. While digital transformation processes can facilitate firms’ achievement of environmental sustainability goals, the effective utilization of digital infrastructure can also promote sustainable innovation. In this context, H1 will examine the impact of DC on ESP. ESP is assessed through survey items that evaluate firms’ commitment to environmentally responsible strategies, aligned with sustainability frameworks [7].

2.3. Knowledge Management: Knowledge Integration and Knowledge Accumulation

Knowledge management is defined as a strategic resource encompassing the processes of data collection, storage, sharing, and analysis within organizations [14]. Extensive research in the literature has examined how knowledge management influences innovation processes, emphasizing the critical role of knowledge integration and KA in these processes [6].

KIC refers to an organization’s ability to combine, analyze, and integrate information from various sources into its innovation processes [21]. Studies indicate that knowledge integration directly impacts innovation performance and enhances firms’ competitive advantage in digital transformation processes [22]. In this context, H5 and H7 will test how knowledge integration shapes the relationship between DC and sustainability/innovation.

KA, in contrast, refers to the strategic reservoir of knowledge that firms develop through internal learning processes [23]. Prior research suggested that KA may serve as a mediating mechanism in the relationship between DC and innovation outcomes [24]. Accordingly, H6 investigates the mediating role of KA in this relationship.

KIC and KA are assessed through established measurement scales that capture firms’ ability to acquire, integrate, and utilize knowledge resources [21,23].

2.4. The Link Between Digitalization Capability, Knowledge Management, and Sustainable Innovation

The impact of DC on innovation performance is shaped through the mediating roles of ESP and knowledge management components. The effective management of firms’ knowledge integration and KA processes can enhance the influence of DC on innovation [6,14]. Additionally, ESP may function as a moderator, strengthening the impact of firms’ digital strategies on innovation processes [8]. In this context, H8 examines how ESP and knowledge management influence the relationship between digitalization and innovation.

Based on the concepts reviewed in the literature, the following hypotheses were developed to systematically investigate the effects of DC on innovation and sustainability processes. Within the scope of this study, the effects of DC on RIP will be analyzed in the context of ESP and knowledge management. In light of the existing literature, the hypotheses were structured to test both direct and indirect effects. The hypotheses of this study are summarized below (

Table 1. Summary of research hypotheses.

Hypothesis	Description
H1	DC positively influences ESP.
H2	ESP positively influences RIP.
H3	DC positively influences RIP.
H4	ESP mediates the relationship between DC and RIP.
H5	KIC mediates the relationship between DC and RIP.
H6	KA mediates the relationship between DC and RIP.
H7	KIC moderates the relationship between DC and ESP.
H8	ESP moderates the relationship between DC and RIP.

).

Beyond firm-specific capabilities, external factors such as industry characteristics and firm size may influence the relationships explored in this study. For instance, large firms often have greater resources to invest in digital transformation, while SMEs may face challenges due to financial and technological constraints. Similarly, industries with high technological dynamism (e.g., IT and biotechnology) may exhibit different patterns in the adoption of DC compared to more traditional sectors.

3. Research Methodology

The interaction between digital transformation and sustainable innovation plays a critical role in enabling businesses to achieve competitive advantage. This study aims to analyze the effects of DC, ESP, and knowledge management variables on RIP. Drawing on the theoretical framework established in the literature review, this study adopted a quantitative research approach, utilizing a structured survey technique for data collection.

To evaluate the proposed hypotheses, Structural Equation Modeling (SEM) and regression analyses were utilized. This section provides a detailed explanation of the methodological framework used in this study, outlining the research design, data collection process, and analytical techniques used to evaluate the relationships among the key variables.

3.1. Research Approach

This study is designed as a cross-sectional quantitative research study aiming to examine causal relationships through hypothesis testing. The preference for a quantitative approach is based on its ability to reveal measurable and statistically analyzable relationships among DC, ESP, and RIP [25]. The cross-sectional design allows for the examination of relationships between DC, ESP, and RIP at a specific point in time, making it suitable for identifying patterns and associations in firms’ behaviors. This method provides a cost-effective and efficient means of gathering data from multiple firms within a short period, enabling robust statistical analysis [26]. However, it does not capture longitudinal changes or causal inferences over time, which future studies could address by employing a longitudinal approach [27].

The research follows a positivist paradigm, employing a deductive approach to test the cause-and-effect relationships between variables [26]. A cross-sectional research design enables data collection within a specific timeframe and provides an appropriate methodology for analyzing relationships between variables [27].

Within the scope of this study, hypotheses were tested using SEM and regression analysis, ensuring a robust examination of the proposed relationships. Compared to previous studies, this research employs an integrated methodological approach combining SEM and the Hayes PROCESS Model to provide a comprehensive examination of DC and innovation outcomes. While prior studies have primarily focused on direct relationships between digital transformation and innovation, this study goes further by incorporating a detailed statistical analysis of multiple mediating and moderating effects, offering a more nuanced understanding of how knowledge management factors shape these relationships. Additionally, the study applies a sector-specific sampling approach by focusing on firms actively undergoing digital transformation, which enhances the relevance and applicability of the findings in contemporary business environments. These methodological advancements contribute to the literature by providing a more holistic perspective on digital transformation and its impact on radical innovation.

3.2. Questionnaire Development

The scales used in this research were adapted from validated and reliable measures in the existing literature. Furthermore, the study presented in [16] significantly contributed to the development of the survey instrument. Additionally, the future research directions proposed in the same study played a crucial role in shaping the research problem.

The survey consists of five main components:

• DC → [1,16,19];
• ESP → [8,10];
• RIP → [16,18,20];
• KIC and KA → [6,11,16].

All items in the survey were measured using a Likert scale ranging from 1 to 5 (1 = strongly disagree; 5 = strongly agree). Additionally, supplementary questions were included to gather demographic information about the participants and the key characteristics of their firms.

The reliability of the scales was tested by calculating Cronbach’s Alpha coefficient, with an expected threshold of α > 0.70 for all variables [25]. Factor analysis was conducted to assess the validity of the measurement scales.

3.3. Sampling and Data Collection

The sample for this research was selected from R&D centers and firms adopting digital transformation strategies in Türkiye. A purposive sampling method was employed [27]. To ensure relevance to the study’s focus, firms were included if they had at least 10 full-time employees, actively engaged in digital transformation initiatives, and operated in technology-driven sectors such as IT, telecommunications, and advanced manufacturing. Firms that did not meet these criteria, such as newly established startups with fewer than 10 employees or businesses that had not yet implemented digitalization strategies, were excluded from the sample. This approach enhances the study’s relevance by ensuring that participating firms have a sufficient level of digital capability and innovation engagement to provide meaningful insights. The sample selection criteria included the following:

• Firms actively implementing digital transformation processes;
• Businesses that adopt innovation strategies and have sustainability goals;
• Companies that invest in R&D and technology.

A total of over 650 firms were contacted via repeated e-mails. The sample size was kept above 200, which is the recommended minimum for SEM analyses [3]. The response rate was calculated at approximately 48%. The study collected 315 usable survey responses through online questionnaires sent via email and through direct phone interviews with selected firms.

This study employed a purposive sampling strategy to ensure that participating firms actively engage in digital transformation and innovation processes. This approach allowed for the selection of firms with relevant experience in leveraging DC and sustainability strategies. While purposive sampling does not enable statistical generalization to all industries, it enhances the contextual relevance of the findings, ensuring that the sample represents firms with strategic involvement in digital transformation [28].

3.4. Data Analysis and Hypothesis Testing

In this study, SPSS 25.0 and AMOS 24 software were used to analyze the collected data and test the research hypotheses. Initially, Cronbach’s Alpha coefficient, Exploratory Factor Analysis (EFA), and Confirmatory Factor Analysis (CFA) were employed to assess the reliability and validity of the measurement scales. Subsequently, SEM was employed to examine the research model and test the hypotheses. Additionally, the Hayes PROCESS Model was applied to test the mediation and moderation effects proposed in the hypotheses.

The measurement structure of the model was evaluated using CFA, and the suitability of the factor analysis was assessed based on the Kaiser–Meyer–Olkin (KMO) value, which was expected to be greater than 0.80. To determine construct validity, the Average Variance Extracted (AVE) and Composite Reliability (CR) values were calculated.

To test the research hypotheses, regression analysis was conducted using SEM, which was chosen because it allows for the analysis of both direct and indirect relationships [29]. The model’s goodness-of-fit criteria were evaluated using key fit indices, including Chi-square/degrees of freedom (χ²/df), Root Mean Square Error of Approximation (RMSEA), Adjusted Goodness-of-Fit Index (AGFI), and Comparative Fit Index (CFI) [25].

While firm size, industry sector, and R&D intensity were collected during the data collection process, these variables were primarily used to define the sample boundaries rather than as control variables in the statistical analysis. The purposive sampling strategy ensured that firms included in the study were actively engaged in digital transformation and innovation processes. Future research could further examine the potential moderating effects of these variables by incorporating them into the structural model.

The mediation effect, which is one of the study’s assumptions, was tested using the Hayes PROCESS Model. In this mediation analysis, the study examined how the effect of DC on RIP was influenced by ESP and knowledge integration [30]. The statistical significance of the mediation effect was assessed using the bootstrap method, and confidence intervals were calculated.

Furthermore, moderation effects proposed in the study were tested using interaction terms and moderated regression analysis. The study analyzed how ESP alters the interaction between DC and RIP and how KIC shapes the effect of DC on sustainability perception. The significance of the moderation effects was evaluated based on the contributions of interaction terms within the regression model.

4. Results

This section presents the results of the analyses conducted to test the research model. First, the sample characteristics and descriptive statistics are provided, followed by the results of the reliability and validity tests. Finally, the hypothesis testing process and results are detailed.

4.1. Descriptive Statistics

The demographic characteristics of the participating firms and managers were analyzed to understand the fundamental structure of the sample group (

Table 2. An overview of the demographic and organizational attributes of the sample.

Variable	Items	%	Variable	Items	%
Age	23–32 years	25.7	Experience	<9 years	24.4
	33–42 years	65.4		10–19 years	59.7
	≥43 years	8.9		≥20 years	15.9
Sex	Female	41.6	Location	İstanbul	32.1
	Male	58.4		Ankara	19.0
Sector	Information Technology	83.2		İzmir	13.0
	Agricultural Technologies	2.2		Other cities	35.9
	Automation	7.3	Duration	Newly Established	1.0
	Textile Technologies	1.9		1–5 years	21.6
	Tourism	2.5		6–10 years	49.8
	E-Commerce	2.9		11–15 years	21.6
Position	Entrepreneur/Owner	25.7		≥16 years	6.0
	Board Member	2.5	Employees	1	-
	Department Manager	36.2		2–9	21.0
	Project Manager	3.8		10–49	71.4
	Software Developer	28.3		50–249	2.5
	Technical Staff	3.5		≥250	5.1

). The distribution of firms was examined based on their industry sectors, firm sizes, the adoption levels of digital transformation strategies, and sustainability policies. Additionally, variables such as the participants’ average age, managerial experience, and education levels were included in the analysis.

Table 2 presents the demographic and organizational characteristics of the sample used in this study. A total of 315 respondents participated, representing firms from various sectors, organizational sizes, and managerial positions.

The respondent sample consists of 58.4% males and 41.6% females. Regarding age distribution, 65.4% of participants are between 33 and 42 years old, 25.7% are in the 23–32 range, and 8.9% are 43 years or older. Regarding work experience, most respondents have 10–19 years of experience (59.7%), while 24.4% have less than 9 years, and 15.9% have 20 or more years of experience. In terms of job positions, 36.2% are department managers, followed by software developers (28.3%), and entrepreneurs/business owners (25.7%), and the rest hold other managerial roles such as project managers (3.8%) and board members (2.5%). The firms in the sample mainly operate in the Information technology sector (83.2%), followed by automation (7.3%), e-commerce (2.9%), and agricultural technologies (2.2%). The sample includes firms with diverse operational histories: 49.8% have been in operation for 6–10 years, followed by 21.6% in both the 1–5 years and 11–15 years categories, while only 6.0% have been in business for over 16 years. Firm size varies across the sample, with 71.4% of firms having between 10 and 49 employees, while 21.0% have between 2 and 9 employees. Larger firms, employing over 250 people, account for only 5.1% of the sample.

These descriptive statistics provide valuable insights into the composition of the sample and help define the scope and generalizability of the findings. The dominance of firms from the information technology sector aligns with the study’s focus on digitalization capabilities. However, the sample also includes firms from diverse industries, which enhances the applicability of the results. The prevalence of mid-sized firms (10–49 employees) suggests that findings will be particularly relevant to SMEs undergoing digital transformation.

In summary, the sample composition reflects a diverse set of industries, experience levels, and firm sizes, allowing for a comprehensive exploration of the relationship between digitalization capabilities, knowledge management, and innovation performance. However, the relatively lower representation of large enterprises (>250 employees) may limit the direct applicability of findings to large-scale corporations.

4.2. Reliability and Validity Tests

The reliability of the measurement scales was assessed using Cronbach’s Alpha coefficients, with a threshold of 0.70 or higher deemed acceptable for each variable [25].

To assess the validity of the scales, both EFA and CFA were conducted (

Table 3. Factor loadings, reliability, and validity of measurement scales.

Dimensions	Survey Items	Factor Loadings
Digital Capability (α = 0.96; AVE = 0.55; CR = 0.94)	Our organization offers structured training programs to enhance employees’ digital competencies.	0.657
	Our digital specialists possess the necessary expertise to perform their tasks efficiently.	0.797
	We collect and analyze large-scale, unstructured, and real-time data for business insights.	0.749
	We consolidate data from various sources into a centralized database for seamless access.	0.736
	We utilize advanced digital tools to process and analyze organizational data.	0.820
	We employ digital infrastructure to swiftly access and retrieve critical internal information.	0.767
	We leverage digital platforms to systematically gather customer feedback.	0.845
	Our organization ensures internal connectivity through digital technologies.	0.808
	We accurately anticipate customer needs by employing data-driven analytics.	0.920
	We facilitate decision-making processes through the visualization of analytical insights.	0.861
	Our executives have a clear understanding of how to leverage digital analytics outcomes.	0.798
	Our leadership comprehends the objectives and requirements of digital transformation for each department.	0.482
	Our executives actively support the adoption of digital technologies to enhance interdepartmental operations.	0.533
	Our managerial team effectively utilizes digital analytics to guide strategic decision-making.	0.375
Knowledge Integration Capability (α = 0.91; AVE = 0.78; CR = 0.93)	We efficiently consolidate and apply our existing knowledge within the organization.	0.862
	We acquire valuable insights from suppliers and integrate them into our operations.	0.892
	We systematically incorporate knowledge obtained from external partners into our business processes.	0.908
	We leverage diverse skill sets to drive product and service development.	0.868
	We efficiently consolidate and apply our existing knowledge within the organization.	0.862
Knowledge Accumulation (α = 0.95; AVE = 0.53; CR = 0.91)	We frequently devise innovative approaches to conducting business.	0.770
	We consistently enhance our managerial skill set to adapt to dynamic environments.	0.878
	We actively engage in the development of groundbreaking technologies.	0.843
	We design and implement novel manufacturing processes.	0.750
	We formulate and execute entirely new marketing strategies.	0.705
	We acquire strategic knowledge from academic institutions and research organizations.	0.683
	We obtain insights from both upstream and downstream supply chain partners.	0.730
	We exchange knowledge with industry peers to foster innovation.	0.411
	We monitor technological advancements by engaging with external sources.	0.672
Radical Innovation Performance (α = 0.87; AVE = 0.49; CR = 0.82)	We have introduced a higher number of new products compared to our competitors.	0.760
	We have incorporated more advanced functionalities into our products than industry counterparts.	0.741
	We have successfully developed and implemented disruptive product and process technologies.	0.772
	We have pioneered novel technologies and methodologies within our sector.	0.612
	We frequently introduce products into emerging markets ahead of competitors.	0.584
Environmental Sustainability Perception (α = 0.88; AVE = 0.81; CR = 0.93)	Our company improves waste management processes by utilizing digital technologies.	0.894
	The digital technologies used in our company support the achievement of sustainability goals.	0.908
	Our company actively employs digital tools to achieve environmental sustainability goals.	0.903

). The KMO test and Bartlett’s Test of Sphericity were carried out to determine whether the data were suitable for factor analysis.

4.3. Structural Equation Modeling Results

SEM was used to test the research model (Figure 1). First, the measurement model was validated, followed by the testing of the structural model. The model fit indices (χ²/df, RMSEA, CFI) were examined to assess the statistical adequacy of the model.

The internal consistency of the measurement model was assessed using Cronbach’s Alpha and CR coefficients. The results indicate that both values exceeded the recommended threshold of 0.70, demonstrating good internal reliability [31].

The goodness-of-fit statistics for the CFA model indicated an acceptable fit: χ² (517) = 1535.695, χ²/df = 2.970, RMSEA = 0.079, CFI = 0.895, IFI = 0.895, TLI = 0.886, RMR = 0.059, and GFI = 0.772. The acceptability of these indices is supported by widely used criteria in the SEM literature. Specifically, RMSEA values below 0.08 indicate reasonable model fit [32], CFI and IFI values close to or above 0.90 suggest good fit [29], and χ²/df values below 3 are considered acceptable [33]. These results confirm that the proposed model exhibits an adequate level of fit. According to commonly accepted thresholds [32,33], a χ²/df ratio below 3 suggests an acceptable fit, while RMSEA values below 0.08 indicate a reasonable approximation error. Additionally, CFI and IFI values close to 0.90 suggest an acceptable fit, though the GFI value of 0.772 falls slightly below the recommended threshold of 0.80, indicating room for model refinement [34].

The standardized factor loadings for all latent variables were above the recommended threshold of 0.55 [35], confirming sufficient indicator reliability. Additionally, the AVE for each construct was above 0.50 [36], indicating good convergent validity. The square root of the AVE for each construct was also greater than the correlation coefficients with other constructs, demonstrating strong discriminant validity [36].

Figure 1 illustrates the structural relationships among DC, ESP, RIP, and knowledge management components (KIC and KA). The arrows indicate the hypothesized paths between these variables, while the numerical values on the paths represent standardized coefficients obtained through SEM. The model was tested to examine how DC influences innovation outcomes and whether ESP plays a role in this process. The findings indicate that DC exhibits a direct positive effect on RIP, confirming the fundamental role of digital transformation in fostering innovation. Additionally, the model demonstrates strong goodness-of-fit indices, ensuring the robustness of the hypothesized relationships. These aspects will be further analyzed in the subsequent sections.

These findings suggest that the measurement model exhibits acceptable reliability, adequate model fit, and strong construct validity, supporting its suitability for further hypothesis testing. However, minor model adjustments could improve the GFI value for an even stronger overall fit.

4.3.1. Direct Effects

The direct effect of DC on RIP was analyzed. Additionally, the impact of ESP and knowledge management variables on innovation processes was examined.

H1 posits that DC positively influences ESP. This hypothesis was examined using a simple linear regression analysis. The regression model was not statistically significant: F(1, 313) = 0.65, p = 0.422. The model explained only 0.2% of the variance in ESP (R² = 0.002, adjusted R² = −0.001), indicating that the predictor did not make a meaningful contribution to the model. Examining the regression coefficients, the unstandardized regression coefficient (B) for DC was 0.048, SE = 0.060, t(313) = 0.80, and p = 0.422. These results indicate that DC does not have a significant effect on ESP. Therefore, H1 is not supported.

H2 suggests that ESP positively influences RIP. This hypothesis was examined using a simple linear regression analysis. The regression model was not statistically significant: F(1, 313) = 0.054, p = 0.816. The model explained only 0.0% of the variance in RIP (R² = 0.000, adjusted R² = −0.003), indicating that ESP does not significantly predict RIP. Examining the regression coefficients, the unstandardized regression coefficient (B) for ESP was −0.013, SE = 0.054, t(313) = −0.233, and p = 0.816. These findings indicate that ESP does not have a significant effect on RIP. Therefore, H2 is not supported.

H3 posits that DC positively influences RIP. This hypothesis was examined using a simple linear regression analysis. The regression model was statistically significant (F(1, 313) = 448.24, p < 0.001), indicating that DC significantly predicts RIP. The model explained 58.9% of the variance in RIP (R² = 0.589, adjusted R² = 0.588), suggesting that DC has a strong predictive power. Examining the regression coefficients, the unstandardized regression coefficient (B) for DC was 0.783, SE = 0.037, t(313) = 21.17, and p < 0.001, indicating a strong and statistically significant effect of DC on RIP. The standardized beta coefficient (β = 0.767) further confirms that DC has a substantial positive effect on RIP. Based on these results, H3 is supported.

The analysis of direct effects provides critical insights into the relationships between DC, ESP, and RIP. The findings indicate that H1 and H2 were not supported, suggesting that DC does not have a significant effect on ESP, and ESP does not significantly predict RIP. This result implies that ESP alone may not be a direct determinant of radical innovation in the examined context. In contrast, H3 was strongly supported, demonstrating that DC has a significant and positive impact on RIP (β = 0.767, p < 0.001). The model explained 58.9% of the variance in RIP (R² = 0.589), confirming that digital transformation plays a crucial role in fostering radical innovation. These findings highlight the importance of DC in enhancing innovation performance while suggesting that additional variables may mediate or moderate the relationship between digitalization and sustainability-driven innovation.

Therefore, in the next section, mediation and moderation effects will be examined to gain a deeper understanding of the role of digital transformation in sustainable innovation.

4.3.2. Mediation Analysis

The mediation effects proposed in this study were analyzed using the Hayes PROCESS Model. The study tested how the effect of DC on RIP is shaped by ESP and knowledge integration variables.

H4 suggests that ESP mediates the relationship between DC and RIP.

To examine the mediating role of environmental sustainability perception (ESP) in the relationship between DC and RIP, Hayes PROCESS Model 4 was utilized. The analysis was conducted with 5000 bootstrap samples to ensure robust statistical inference.

The direct effect of DC on RIP was significant (B = 0.785, SE = 0.037, p < 0.001), indicating that DC positively influences RIP (

Table 4. Mediation analysis results for H4.

Effect	B	SE	t	p	LLCI	ULCI
Direct Effect (DC → RIP)	0.785	0.037	21.24	<0.001	0.7124	0.8579
Indirect Effect (DC → ESP → RIP)	−0.0022	0.0040	—	—	−0.0120	0.0047
Effect of DC on ESP	0.048	0.060	0.80	0.422	−0.0696	0.1657
Effect of ESP on RIP	−0.046	0.035	−1.33	0.186	−0.1150	0.0224

). The indirect effect of DC on RIP through ESP was not statistically significant (B = −0.0022, BootSE = 0.0040, BootLLCI = −0.0120, BootULCI = 0.0047) as the confidence interval includes zero. Additionally, DC did not significantly predict ESP (B = 0.048, p = 0.422), and ESP did not significantly predict RIP (B = −0.046, p = 0.186). Since the indirect effect was not significant, the mediation hypothesis H4 is not supported. This suggests that ESP does not act as a mediator between DC and RIP, and the observed relationship between DC and radical innovation remains direct rather than indirect.

H5 suggests that KIC mediates the relationship between DC and RIP.

Hayes PROCESS Model 4 was employed to investigate the potential mediating role of KIC in the relationship between DC and RIP. The mediation analysis was carried out using 5000 bootstrap samples to enhance statistical reliability.

The direct effect of DC on RIP was significant (B = 0.4095, SE = 0.0434, t(313) = 9.43, p < 0.001), indicating that DC positively influences RIP (

Table 5. Mediation analysis results for H5.

Effect	B	SE	t	p	LLCI	ULCI
Direct Effect (DC → RIP)	0.4095	0.0434	9.43	<0.001	0.3241	0.4949
Indirect Effect (DC → KIC → RIP)	0.3734	0.0529	—	—	0.2686	0.4726
Effect of DC on KIC	0.7850	0.0440	17.86	<0.001	0.6985	0.8715
Effect of KIC on RIP	0.4757	0.0393	12.11	<0.001	0.3984	0.5530

). The indirect effect of DC on RIP through KIC was statistically significant (B = 0.3734, BootSE = 0.0529, BootLLCI = 0.2686, BootULCI = 0.4726) as the confidence interval did not include zero. Additionally, DC significantly predicted KIC (B = 0.7850, p < 0.001), and KIC significantly predicted RIP (B = 0.4757, p < 0.001). Since the indirect effect was significant, the mediation hypothesis H5 is supported. This suggests that KIC acts as a mediator between DC and RIP, indicating that firms with stronger DC enhance their RIP through knowledge integration mechanisms.

H6 suggests that KIC mediates the relationship between DC and RIP.

Hayes PROCESS Model 4 was utilized to assess whether KA acts as a mediator in the relationship between DC and RIP. The mediation analysis was performed with 5000 bootstrap samples to ensure statistical robustness.

The direct effect of DC on RIP was significant (B = 0.6144, SE = 0.0762, t(313) = 8.06, p < 0.001), indicating that DC positively influences RIP (

Table 6. Mediation analysis results for H6.

Effect	B	SE	t	p	LLCI	ULCI
Direct Effect (DC → RIP)	0.6144	0.0762	8.06	<0.001	0.4644	0.7644
Indirect Effect (DC → KA → RIP)	0.1685	0.0865	—	—	0.0032	0.3422
Effect of DC on KA	0.9292	0.0288	32.26	<0.001	0.8726	0.9859
Effect of KA on RIP	0.1814	0.0719	2.52	0.0122	0.0398	0.3229

). The indirect effect of DC on RIP through KA was statistically significant (B = 0.1685, BootSE = 0.0865, BootLLCI = 0.0032, BootULCI = 0.3422) as the confidence interval did not include zero. Additionally, DC significantly predicted KA (B = 0.9292, p < 0.001), and KA significantly predicted RIP (B = 0.1814, p = 0.0122). Since the indirect effect was significant, the mediation hypothesis H6 is supported. This suggests that KA acts as a mediator between DC and RIP, indicating that firms with strong DC enhance their RIP through accumulated knowledge.

4.3.3. Moderation Analysis

To test the moderation effects of KIC and ESP, interaction terms were added to the regression models. The significance of the moderation analysis was evaluated based on the interaction variables within the model.

Hayes PROCESS Model 1 was applied to assess the moderating role of KIC in the relationship between DC and RIP. The analysis was conducted using 5000 bootstrap samples to ensure statistical reliability.

The interaction term (DC × KIC) was statistically significant (B = 0.0530, SE = 0.0222, t(311) = 2.38, p = 0.0178), indicating that KIC moderates the effect of DC on RIP (

Table 7. Moderation analysis results for H7.

Effect	B	SE	t	p	LLCI	ULCI
DC (X) → RIP (Y)	0.2666	0.0738	3.61	<0.001	0.1213	0.4119
KIC (W) → RIP (Y)	0.2956	0.0851	3.47	<0.001	0.1282	0.4630
Interaction (DC × KIC → RIP)	0.0530	0.0222	2.38	0.0178	0.0092	0.0968

). The main effect of DC on RIP was significant (B = 0.2666, p < 0.001), showing that higher DC leads to higher RIP. The main effect of KIC on RIP was also significant (B = 0.2956, p < 0.001), suggesting that firms with stronger knowledge integration capabilities achieve higher RIP.

The effect of DC on RIP varies at different levels of KIC, demonstrating a moderation effect.

The results show that as KIC increases, the effect of DC on RIP strengthens, confirming the presence of a positive moderating effect (

Table 8. Effect size analysis results for KIC.

KIC Level	Effect of DC on RIP	SE	t	p	LLCI	ULCI
Low (KIC = 2.00)	0.3726	0.0458	8.14	<0.001	0.2825	0.4627
Medium (KIC = 3.00)	0.4256	0.0436	9.76	<0.001	0.3398	0.5114
High (KIC = 4.00)	0.4786	0.0519	9.21	<0.001	0.3764	0.5808

). Since the interaction effect is statistically significant, the moderation hypothesis H7 is supported. This suggests that firms with higher KIC can leverage DC more effectively to enhance RIP.

Hayes PROCESS Model 1 was employed to evaluate whether environmental sustainability perception (ESP) moderates the relationship between DC and RIP. The analysis utilized 5000 bootstrap samples to enhance statistical robustness.

The interaction term (DC × ESP) was not statistically significant (B = −0.0338, SE = 0.0368, t(311) = −0.92, p = 0.3589), indicating that ESP does not significantly moderate the effect of DC on RIP (

Table 9. Moderation analysis results for H8.

Effect	B	SE	t	p	LLCI	ULCI
DC (X) → RIP (Y)	0.8976	0.1278	7.02	<0.001	0.6461	11.490
ESP (W) → RIP (Y)	0.0624	0.1233	0.51	0.6135	−0.1803	0.3051
Interaction (DC × ESP → RIP)	−0.0338	0.0368	−0.92	0.3589	−0.1063	0.0386

). The main effect of DC on RIP was significant (B = 0.8976, p < 0.001), showing that higher DC leads to higher RIP. However, the main effect of ESP on RIP was not significant (B = 0.0624, p = 0.6135), suggesting that ESP alone does not significantly predict RIP.

Since the interaction effect is not statistically significant, the moderation hypothesis H8 is not supported. This suggests that ESP does not influence the strength of the relationship between DC and RIP, and DC remains a strong predictor of RIP independent of sustainability perceptions.

4.4. Hypothesis Testing Results

This section presents a table summarizing the results of the tested hypotheses (

Table 10. Summary of hypothesis testing results.

Hypothesis	Path	Effect (B)	SE	t	p	Support
H1	DC → ESP	0.0481	0.0598	0.80	0.422	Not Supported
H2	ESP → RIP	−0.013	0.054	−0.23	0.816	Not Supported
H3	DC → RIP	0.7829	0.0370	21.17	<0.001	Supported
H4	DC → ESP → RIP (Mediation)	−0.0022	0.0040	—	—	Not Supported
H5	DC → KIC → RIP (Mediation)	0.3734	0.0529	—	—	Supported
H6	DC → KA → RIP (Mediation)	0.1685	0.0865	—	—	Supported
H7	DC × KIC → RIP (Moderation)	0.0530	0.0222	2.38	0.0178	Supported
H8	DC × ESP → RIP (Moderation)	−0.0338	0.0368	−0.92	0.3589	Not Supported

). The acceptance or rejection status of the hypotheses is provided, along with regression coefficients and p-values to offer a clear overview of the findings.

5. Discussion

This study aimed to examine the impact of DC on RIP while considering the roles of ESP and knowledge management components (KIC and KA). The research findings contribute to the existing literature on digital transformation, innovation, and sustainability by clarifying how DC drives innovation and whether sustainability perception influences this relationship.

5.1. Key Findings and Theoretical Implications

The findings suggest that DC plays a crucial role in driving radical innovation, supporting the argument that firms investing in digital transformation gain significant advantages in their innovation performance. While digital transformation enhances radical innovation by enabling firms to adopt disruptive technologies and novel business models, it also fosters incremental innovation by improving operational efficiencies and optimizing existing processes [9]. Firms implementing digitalization strategies often first experience incremental innovation through automation, data-driven decision-making, and gradual process improvements before achieving radical innovation breakthroughs [37]. However, for digital transformation to translate into radical innovation, firms must develop strong knowledge management frameworks that encourage experimentation and cross-functional collaboration. These findings suggest that firms should balance digital investments between incremental process improvements and transformative innovation efforts to maximize long-term competitive advantage.

These findings align with previous research emphasizing the transformative power of digitalization in fostering innovation [1,19]. However, contrary to some prior studies suggesting that sustainability-oriented firms achieve higher innovation performance [11,15], this study finds that ESP alone does not significantly drive radical innovation. One possible explanation is that while sustainability efforts contribute to corporate social responsibility and operational efficiency, they may not directly translate into the disruptive processes required for radical innovation. Instead, firms may need to complement their sustainability strategies with structured knowledge management frameworks to enhance innovation outcomes. However, contrary to expectations, ESP does not significantly influence RIP, nor does it mediate or moderate the relationship between DC and innovation.

The direct effect of DC on RIP was strongly supported (H3), indicating that firms with higher levels of digitalization achieve superior innovation outcomes (B = 0.7829, p < 0.001). This finding aligns with prior studies that highlight the transformative power of digital technologies in fostering innovation [1,19].

ESP did not significantly predict RIP (H2) nor act as a mediator (H4) or moderator (H8) in this relationship, suggesting that firms’ sustainability awareness alone is insufficient to drive radical innovation. This result contrasts with prior research suggesting that sustainability-oriented firms engage in more innovative activities [11].

One possible explanation for this finding is that sustainability perception primarily influences firms’ compliance-based and efficiency-driven innovation efforts rather than driving disruptive or radical change. While sustainability initiatives encourage firms to optimize resources and reduce environmental impact, they do not inherently create the dynamic capabilities required for breakthrough innovations [10,12]. In contrast, knowledge-driven mechanisms, such as KIC and KA, directly enhance firms’ absorptive capacity and their ability to synthesize new technological knowledge [24]. These mechanisms facilitate cross-functional collaboration, enabling firms to better leverage digitalization for innovation. Therefore, firms prioritizing digital transformation should invest in knowledge management strategies to maximize their innovation outcomes.

Knowledge management components—KIC (H5) and KA (H6)—were found to mediate the relationship between DC and RIP. The significant mediation effects of KIC and KA highlight the central role of knowledge management in digital transformation processes. This finding is consistent with the knowledge-based view [13] and dynamic capabilities framework [38], which suggest that firms must actively develop mechanisms to integrate, store, and utilize knowledge for sustained innovation performance. The results suggest that digitalization alone is not sufficient for achieving radical innovation; rather, organizations must leverage their digital investments through effective knowledge-sharing and integration strategies. This underscores the importance of interdepartmental collaboration and cross-functional knowledge flows as key enablers of digital innovation. This suggests that firms leveraging their DC through effective knowledge integration and accumulation processes are more likely to enhance their innovation performance. These findings align with prior research emphasizing the role of knowledge in digital transformation and innovation [13,14].

The moderating role of KIC (H7) was supported, indicating that firms with higher KIC levels experience a stronger impact of DC on radical innovation. This finding suggests that firms with well-established knowledge management structures can maximize the innovation benefits of digital transformation. Previous studies have indicated that firms with high absorptive capacity [20] are better positioned to translate digital investments into innovative outputs. Our study extends this argument by demonstrating that firms with high KIC not only benefit more from digitalization but also sustain their competitive advantage through effective knowledge utilization. This finding holds important managerial implications, emphasizing the need for firms to cultivate knowledge-sharing cultures and invest in digital platforms that facilitate organizational learning and knowledge exchange.

5.2. Managerial Implications

These findings offer significant managerial implications for firms aiming to enhance their innovation performance through digital transformation.

Investing in DC is crucial for fostering radical innovation. Firms should prioritize digital transformation strategies, focusing on technologies such as AI, big data analytics, and automation to enhance their innovation potential.

The findings of this study have direct application in industries undergoing digital transformation, such as manufacturing, healthcare, finance, and logistics. For example, firms in the manufacturing sector can use DC to optimize production efficiency and improve innovation in product design. In healthcare, digital transformation enables better management of patient data and the development of diagnostic tools based on artificial intelligence. The financial sector benefits from DC through improved risk management, fraud detection, and data-driven decision-making. Similarly, logistics companies can use digitalization to streamline supply chain operations, reduce inefficiencies, and improve customer experience. These applications illustrate how firms in various industries can leverage DC and knowledge management to drive innovation and competitiveness.

Knowledge management processes play a critical role in maximizing the impact of DC on innovation. Organizations should develop strong mechanisms for knowledge integration and accumulation to effectively utilize digital transformation initiatives. To enhance knowledge management in digital transformation, firms can take several actionable steps. First, implementing enterprise knowledge-sharing platforms (e.g., cloud-based knowledge repositories, and AI-driven decision support systems) can facilitate seamless collaboration and knowledge exchange across departments. Second, companies should invest in training programs focused on digital literacy and knowledge management to build an innovation-driven workforce. Third, fostering cross-functional teams that integrate IT, R&D, and business strategy units can accelerate the transformation of knowledge into actionable innovation. Finally, leveraging AI and big data analytics to extract actionable insights from organizational knowledge can enhance decision-making and innovation performance.

Sustainability awareness alone does not guarantee innovation success. While environmental sustainability is an important aspect of corporate responsibility, firms should combine their sustainability efforts with knowledge-driven strategies to enhance their innovation outcomes.

KIC enhances the effectiveness of digital transformation. Managers should foster environments that encourage cross-functional knowledge sharing and collaboration, ensuring that digital transformation efforts translate into meaningful innovation improvements.

5.3. Limitations and Future Research Directions

Despite its contributions, this study has some limitations that should be acknowledged.

Sample Composition: The study primarily focuses on firms operating in technology development centers and innovation-driven sectors, limiting the generalizability of the findings to more traditional industries. Future research should examine a broader range of industries to enhance external validity.

Cross-Sectional Design: The research employs a cross-sectional approach, which limits the ability to establish causality. Longitudinal studies could provide deeper insights into the evolving impact of DC on innovation. Future research could benefit from a longitudinal approach to examine how knowledge management processes influence digital innovation over time. The dynamic nature of digital transformation suggests that the integration and accumulation of knowledge may have delayed or cumulative effects on innovation performance [38]. By tracking firms over extended periods, researchers can better assess how organizations adapt their knowledge management strategies in response to evolving digital trends and market conditions. Such studies could offer more precise insights into causal relationships and the long-term sustainability of digital transformation initiatives.

Additional Moderators and Mediators: Future studies could explore additional factors, such as organizational culture, leadership styles, or firm agility, to further understand the dynamics between digitalization, sustainability, and innovation.

6. Conclusions

This study provides empirical evidence on the roles of DC, ESP, and knowledge management in shaping RIP. The findings emphasize that DC is a key driver of innovation, and its impact is significantly enhanced when firms effectively manage their knowledge resources.

While the direct relationship between digitalization and radical innovation was confirmed, ESP does not play a significant role in shaping this relationship. Instead, knowledge integration and accumulation emerge as critical mechanisms that enable firms to maximize the benefits of digital transformation.

These insights contribute to the literature on digital transformation, innovation management, and sustainability by providing a nuanced understanding of how firms can leverage DC for innovation success. This study also contributes to the broader discourse on the intersection of digital transformation and sustainability by questioning the direct role of ESP in driving radical innovation. While sustainability remains a key strategic priority for firms, our findings suggest that digital innovation success depends more on knowledge-driven processes than on sustainability awareness alone. These insights align with the evolving view that firms must go beyond compliance-driven sustainability efforts and focus on integrating digital and knowledge management capabilities to create meaningful innovation impact [5]. Future research should further explore how firms can align their digital strategies with sustainability goals in ways that lead to measurable innovation outcomes. Practically, the study underscores the importance of integrating digital strategies with robust knowledge management frameworks to enhance innovation outcomes.

Despite its contributions, this study has several limitations that should be acknowledged. First, the use of a cross-sectional research design limits the ability to establish causal relationships. Future research could employ longitudinal studies to examine the dynamic evolution of digitalization, sustainability, and innovation performance over time. Second, the reliance on self-reported data introduces potential biases, such as common method variance and social desirability effects. Incorporating objective performance measures or multi-source data could help mitigate these concerns. Third, while the purposive sampling strategy ensures relevance to the research focus, it may limit the generalizability of the findings to other industries or non-digitalized firms.

Future research should build upon these findings by examining longitudinal effects, exploring industry-specific variations, and incorporating additional contextual factors that may influence the digitalization–innovation relationship. A potential avenue for future research is to examine the role of digital leadership and organizational culture in shaping the impact of DC on innovation. Leadership styles that promote digital openness, experimentation, and knowledge sharing may serve as critical enablers in leveraging digital transformation for radical innovation. Additionally, longitudinal studies could provide deeper insights into how firms dynamically adjust their digital and knowledge management strategies over time to sustain innovation performance. Future studies could also explore how firms in non-technology sectors navigate digital transformation and whether similar knowledge-based mechanisms apply in more traditional industries. As digital transformation continues to reshape competitive landscapes, organizations that effectively align their DC with knowledge-based strategies will be better positioned to drive radical innovation and maintain sustainable competitive advantages.