A Study on the Impact of Digital Transformation on Enterprise Performance: The Mediating Role of Dual Innovation and the Moderating Role of Management Power

Abstract

Digital transformation and technological innovation have an essential impact on enterprise performance, and clarifying the relationship between the two has become a real problem that enterprise management needs to solve. This study utilizes data from China’s listed enterprises between 2012 and 2022 to comprehensively analyze the influence and mechanism of digital transformation on enterprise performance. The findings indicate the following: (1) Digital transformation can effectively promote improving enterprise performance and dual innovation levels. (2) Digital transformation positively impacts enterprise performance primarily through enhanced exploitative innovation, exploratory innovation, dual innovation balance, and complementarity. (3) Managerial power positively moderates the promotion effect of digital transformation on enterprise performance. (4) The impact of digital transformation on enterprise performance is more significant in the eastern and central regions, high-tech industries, and enterprises with good profitability. Based on these results, enterprises are suggested to accelerate digital transformation, fully tap the synergistic driving effect of dual innovation, strengthen the management’s cognition and leadership of digital transformation, and implement dynamic and differentiated digital transformation strategy based on region, industry, and their own characteristics.

1. Introduction

The 17 United Nations Sustainable Development Goals (SDGs) are the international community’s common vision and framework for action to build a more just, inclusive, and sustainable future. As one of the world’s largest developing countries, China participates in setting the SDGs and is an active promoter. With the continuous improvement of digital infrastructure construction, the digital economy, with its advantages of key data resources, efficient communication media, and the integration of multiple scenarios, has become the core driving force to promote the transformation of old and new factors of production and drive the optimization of the industrial chain, which will not only provide “new kinetic energy” for reshaping the industrial structure and promoting the high-quality development of the economy but also further intensifies the competition among enterprises in the field of value creation and supply. In the face of a complex and changing environment and limited resources, enterprises must urgently accelerate the process of digital transformation to achieve long-term sustainable development. In this process, enterprise performance, as a comprehensive indicator of profitability and market competitiveness over some time, provides management with decision support and direction for improvement. At the same time, in order to maintain competitive advantages and realize sustainable development, enterprises are not only required to carry out technological upgrading and product innovation in core or related fields based on the existing technological foundation in order to maintain short-term gains [1], but also to emphasize the in-depth exploration and development of new technologies and knowledge for long-term development [2], and to balance short-term gains by dual innovation and long-term development needs, and the application of emerging digital concepts and digital technologies also has an essential impact on the dual innovation capabilities of enterprises. Therefore, as the pace of digital transformation accelerates, it is necessary to explore the intrinsic links and interaction mechanisms between digital transformation, dual innovation, management power, and enterprise performance.

To achieve sustainable development, enterprises should rely on digital technology for innovation and reform in various fields, including product development. Regarding the economic benefits of digital transformation, the macro level mainly focuses on the impact of digital technology on decarbonization, the traditional economy, and the development of innovative industries. The digital economy realizes carbon emission reduction through three paths: adjusting the level of trade openness [3], promoting financial development [4], and enhancing government efficiency [5]. Digital technology through industrial digitization and digital industrialization of the dual-wheel drive path to empower the transformation and development of the real economy [6], and through the combination of iteration, integration and innovation, and mutual empowerment with the traditional factors of production to promote the growth of total factor productivity [7]. The region can improve the level of coordination between the production and consumption links in the supply chain by innovating the manufacturing production and service methods, contributing to the reconstruction of the traditional industrial organization form [8], and strengthening industrial governance. The enterprise level focuses on the impact of digital transformation on business sustainability through production processes, organizational structure, and capital markets. Digital technology penetrates all aspects of the enterprise life cycle with data as the key production factor, reduces costs, increases efficiency through business datatization [9], and effectively alleviates the external financing dilemma [10]. At the same time, enterprises invest in information technology to improve production methods [11], which facilitates work automation, reshapes production relations [12], and develops new profitable growth models [13]. In addition, digital technology has introduced advanced management concepts and internal control methods, making enterprises’ financial and internal control management more transparent [14] and promoting dynamic adaptability and scalable changes in organizational structure [15]. In this process, the demand for excellent managers and highly skilled personnel increases, which promotes the adjustment of the composition of human capital and the share of labor income [16], which in turn changes the human capital composition of enterprises. Moreover, digital technology improves the ability of market players to deal with non-standardized and unstructured data, reduces the uncertainty faced by external investors in investment decision-making [17], and effectively mitigates the damaging impact enterprises suffer due to market volatility. Studies on the relationship between digitization and enterprise innovation have focused on enterprise innovation performance [18] and efficiency [19]. The digital economy can not only effectively compensate for enterprises’ natural resource dependence to carry out green innovation and develop a green economy [20], but the new ways of value capture associated with digitization also contribute to process innovation and product innovation in manufacturing enterprises [21].

However, little literature has focused on the impact of management power and dual innovation behavior on enterprise performance in digital transformation. First, the strength of management power directly affects managers’ strategic orientation and path of resource allocation, which affects the rate and effectiveness of digital transformation and influences the firm’s leading position in the frontiers of products, services, and technologies. Secondly, embedding digital concepts and the wide application of digital technology further enhance the innovation ability of enterprises. However, a single innovation mode will lead to problems such as innovation curing, and exploitative innovation and exploratory innovation are enterprise innovations of different natures, and it is necessary to analyze further the influence mechanism of dual innovation between digital transformation and enterprise performance. Thirdly, as the digital transformation of enterprises deepens, the mode of choice between exploiting and exploring innovation also shifts, and this dynamic process is related to the profitability and future development of enterprises. The possible marginal contributions of this paper are the following: (1) Given the long-term strategic value of digital transformation, it profoundly explores how management power affects the relationship between digital transformation and enterprise performance. (2) Based on the dual innovation research framework, it explores the impact pathways of exploitative innovation and exploratory innovation between digital transformation and enterprise performance, which enriches the economic outcomes of enterprises’ digital technology application. (3) It further studies the role of dual Innovation, dual innovation balance and dual innovation complementarity in digital transformation and enterprise performance, to clarify how dual innovation can break through resource constraints and improve enterprise performance through digital transformation, aiming to provide theoretical support for China’s enterprises to digitally transform, solve innovation dilemmas, and achieve sustainable economic development.

2. Theoretical Analysis and Research Hypotheses

2.1. Digital Transformation and Enterprise Performance

Enterprise digital transformation is a transformational process of using digital technologies to create new business processes or customer experiences to meet changing business and market needs. The factors influencing enterprise digital transformation are mainly categorized into internal and external factors. Internal factors mainly include resources, technology, cognition, talent, and organizational practices [22], while external factors mainly include financing constraints, policy support, technology application, and industrial synergism. Enterprises’ perception of the long-cycle characteristics of digital transformation has led some SMEs to fall into the transformation decision-making dilemma, and the lack of digital talent and knowledge on how to effectively use digital tools affect the digital transformation of SMEs so that enterprises have to consider the high cost of digital transformation and the uncertainty of the expected future benefits. The incremental profits of the “Schumpeter effect” amplify the production efficiency of the same labor input. In Schumpeter’s theoretical framework, innovation is the source of economic growth and the main way enterprises obtain incremental profits. Schumpeterian incremental profit is the amplification of the production benefits of the same labor input through innovation, which is not only technological innovation but also includes the introduction of new production methods, new markets, new sources of raw material supply, and new forms of organization. Based on the Schumpeterian innovation mechanism of “profit increase—innovation growth” [23], by optimizing access to financing and expanding the influence of the financial system, digital technology promotes the innovative behavior of manufacturing firms, significantly cuts down on the variable costs of production, and thus significantly improves the profit level of incumbent firms under constant conditions. At the same time, by enabling high-quality development and data sharing, firms that are the first to develop new technologies can capitalize on the technology gap to reap additional profits (i.e., Schumpeterian rents) [24] and realize the multiplier effect of economic efficiency by reducing the price markups of firms and improving the efficiency of resource allocation [25].

Digital transformation achieves the goal of reducing costs and increasing efficiency by enhancing the effectiveness of enterprise management and increasing the comprehensive value of enterprise products and services. To the internal enterprise, digital transformation accelerates the internal and external circulation and sharing of enterprise knowledge and technology resources [26], optimizes the existing technology system, and improves business performance. At the same time, the construction of digital information systems makes data the core driving force of production management and improves the efficiency of value creation. In addition, digital transformation prompts enterprises to re-examine and optimize the existing resource allocation, promotes the transformation of the organizational structure networked and flattened, and promotes the comprehensive innovation of the business strategy and mode of thinking, improving the overall performance level of the enterprise. To the outside of the enterprise, digital transformation can reduce the difficulty of obtaining market information, enhance the sensitivity of the perception of changes in the external environment, prompt enterprises to establish a more refined consumer operation network, maintain market competitiveness, and improve corporate efficiency. With the improvement of digital infrastructure and the enhancement of supply chain coordination, the industrial structure and the nature of competition have changed [27], the reshaping of industry boundaries provides unprecedented opportunities for enterprises to cross traditional areas and explore emerging markets. Therefore, the following hypotheses are proposed:

H₁. 

Digital transformation positively drives enterprise performance.

2.2. Impact of Digital Transformation on Dual Innovation

Technological innovation can be classified as exploitative and exploratory innovation based on its magnitude and nature. Raisch et al. [28] argue that when organizations implement these two types of innovation in parallel or alternately, it constitutes a dual innovation strategy. Exploitative innovation focuses on iterating and optimizing products, services, and markets within the framework of existing knowledge and resources, pursuing the predictability and stability of goals and efficiency, and is a form of the adaptive development of firms. Exploratory innovation, on the other hand, focuses on the creation of new knowledge, the development of new technologies and products, the exploration of new production processes and the development of new markets, and is characterized by a high degree of embeddedness, high risk, and high return, and seeks to realize the overall innovation of the enterprise and a fundamental leap in technology.

Dual innovation can enhance the long-term competitiveness of enterprises while continuing the existing business model and maintaining the stability of operation and revenue. It has become an essential way to improve enterprises’ sustainable development and competitive advantage significantly [29,30]. Integrating data elements into the whole chain of enterprise operation realizes digital empowerment, revolutionizes the way of value co-creation among innovation subjects, and becomes a powerful engine for enterprises to cross the technological bottleneck and comprehensively promote innovation capability and enterprise effectiveness. For enterprise exploitative innovation, data elements embedded in the enterprise are conducive to the exploration of the potential innovation value of existing resources, and promote technological upgrading and intelligent transformation to expand the new functions of products and services, which strengthens the ability of exploitative innovation in the established development path. Then, with the characteristics of large-scale dissemination and the replication of data elements, it penetrates different industrial fields, broadens the ways for enterprises to acquire, share, and reorganize data technology innovation resources, effectively reduces innovation costs and shortens the research and development cycle. In addition, with the help of big data and other advanced technologies, enterprises can accurately capture and analyze market dynamics information and flexibly adjust their strategies accordingly to respond to diversified consumer demands, as well as improve the efficiency of product supply and demand matching and effectively reduce the risks that may be faced in the process of utilizing technological innovation. The deep penetration and embedding of data elements for enterprise exploratory innovation have broadened the market players involved in innovation [31]. Diversifying innovation subjects helps enterprises quickly grasp the dynamics of emerging fields and technological advances, accelerating the integration of external innovation resources and linking to the speed of the external environment, enhancing the enterprise’s willingness to explore exploratory innovation and motivation. At the same time, digital transformation also helps to break established cognition and thinking inertia, as well as acceptance and in-depth exploration of diversified knowledge areas, paving the way for breakthroughs in key core technologies. In addition, the massive information, knowledge, and other innovation elements provided by digital technologies such as artificial intelligence and cloud computing in the transformation process will help enterprises build high-level innovation capabilities that are difficult to imitate and uniquely competitive to maintain long-term competitive advantages.

H_2a. 

Digital transformation is positively correlated with exploitative innovation.

H_2b. 

Digital transformation is positively correlated with exploratory innovation.

2.3. The Mediating Role of Dual Innovation Between Digital Transformation and Enterprise Performance

It is difficult for any enterprise to maintain a leading position in all fields consistently, and only by continuously accumulating experience and gathering resources can technological innovation create more value [32]. Based on the dual innovation analysis framework, it is sorted out how digital transformation affects enterprise performance through exploitative innovation and exploratory innovation: for exploitative innovation, through digital transformation, enterprises can improve the utilization efficiency of the existing innovation elements and technology level, promote the increase of technology stock [33], improve the production process and management mode, enrich the form of the products to meet the current market demand better, and transform the digital achievements into actual productivity and improve enterprise performance. For exploratory innovation, the digital transformation strategy can realize the innovation and upgrading of production and operation processes and the in-depth optimization of product processes, promote the output of new technological achievements, open up entirely new production areas and product markets, or even reshape the existing market pattern and trigger industry changes [34], accelerating the improvement of overall enterprise performance.

Exploitative and exploratory innovation are considered two different innovation strategies; enterprises’ survival and development must choose different innovation behaviors according to the uncertainty of the external environment and the limited internal resources. Firms with access to external resources can benefit from the complementarity of dual innovation, while firms with limited resources must focus on the balance of dual innovation [35,36]. There is a need for balance between the two types of innovation, and there will also be a degree of inclusiveness and complementarity. Dual innovation balance is the degree to which the two exploitative and exploratory innovation approaches are balanced, finding the best fit between maintaining existing strengths (exploitative innovation) and exploring future potential (exploratory innovation). Dual innovation complementarity refers to the complementary and promotional relationship between the two types of innovation. The balanced development of dual innovation can effectively activate the idle resources of the enterprise, alleviate the internal conflict brought about by the competition for resources, and realize the optimal allocation and efficient use of resources [37], and at the same time, avoid the over-reliance on exploitative innovation leading to the enterprise falling into the “innovation trap,” and avoid the over-pursuit of exploratory innovation leading to the enterprise facing the expensive R&D costs, market uncertainty, and the high costs of the future. The mutual complementation and promotion of exploitative and exploratory innovation enable enterprises to broaden their business scope, increase market share and economic benefits [38] and lay a solid foundation to obtain and consolidate their competitive advantages. Exploitative innovation can broaden the scope of the enterprise’s search for new knowledge, enhance its knowledge absorption and internalization capabilities, accumulate rich intellectual capital and capability base for exploratory innovation, and provide technical support for the reshaping of organizational structure, operational paradigm, and business processes, which helps to improve the enterprise’s operational efficiency within the existing framework. The new knowledge and technologies generated by exploratory innovation activities constitute the prerequisites for exploitative innovation activities [39]. The promotion of the innovative application of knowledge by enterprises not only contributes to the further improvement of the level of exploitative innovation but also opens up a path to satisfy the highly differentiated needs of the market, which is conducive to improving consumer stickiness and enabling enterprises to benefit from the continuous interaction and value co-creation with consumers. The deepening of digital transformation greatly facilitates the parallelism and mutual reinforcement of these two types of innovation activities, solidifying an enterprise’s position in the existing market and helping it seize the first opportunity to win a competitive advantage in the emerging market. Based on the above, the following hypotheses are proposed:

H_3a. 

Digital transformation enhances enterprise performance through the promotion of exploitative innovation.

H_3b. 

Digital transformation enhances enterprise performance through the promotion of exploratory innovation.

H_3c. 

Digital transformation improves business performance by increasing the balance of dual innovation.

H_3d. 

Digital Transformation Improves Business Performance by Increasing Dual Innovation complementarity.

2.4. The Moderating Role of Management Power Between Digital Transformation and Enterprise Performance

Finkelstein [40] argues that management power is the ability of managers to realize their managerial vision, that digital transformation and operational activities are part of the overall corporate strategy, and that the influence of top managers on corporate strategy depends mainly on whether they have the necessary power [41]. At the same time, enhanced management power directly improves managers’ control over the enterprise and resource allocation efficiency [42], so managers are more willing to customize the development strategy that fits the enterprise according to the team’s capabilities. Management is a crucial factor influencing the efficiency of enterprise investment decision-making; improving enterprise performance benefits from optimizing efficient investment decision-making, financing activities, and operation management. Digital transformation effectively promotes the docking and cooperation between the upstream and downstream enterprises of the industrial chain, thus realizing the enhancement and optimization of the efficiency of the entire industrial system, reducing the subjective opinions and decision-making bias of the management [43], enabling managers to make more scientific and reasonable resource deployment decisions, ensuring that enterprises can accurately grasp the market dynamics, and enhancing the operational efficiency and competitiveness of enterprises. In addition, appropriate management power provides solid support for market development and new project investment, which contributes to the implementation of the enterprise’s digital transformation strategy, enhances market competitiveness, and is ultimately reflected in the significant improvement of enterprise performance. Based on this, the following hypotheses are proposed:

H₄. 

Management power can positively moderate the relationship between digital transformation and enterprise performance.

The conceptual model constructed based on the analyzed content of the above theoretical mechanisms is shown in Figure 1.

3. Study Design

3.1. Sample Selection and Data Sources

This paper selects the data related to Chinese listed companies from 2012 to 2022 as the research sample, and the data are mainly derived from the Cathay Pacific database, the Vantage database, the China Research Data Service Platform, and the China Economic and Financial Research Database. Meanwhile, in order to improve the data quality: (1) the original samples are screened to exclude the samples with incomplete or abnormal ST categories and related indicators; (2) the continuous variables are subjected to a bilateral 1% shrinking tail treatment to avoid the influence of extreme values; and (3) the regression results are adjusted by the robust standard error to make the statistical inference results more robust.

3.2. Variable Measurement and Description

(1)

Enterprise performance (EP). Most of the existing studies choose indicators such as net interest rate on total assets, return on net assets, and Tobin’s Q value to measure enterprise performance, in which net interest rate on total assets can reflect the ability of enterprises to utilize their assets to obtain profits, and can more accurately reflect the profitability of enterprises and the efficiency of asset operation in a certain period. Therefore, this paper refers to Sun C. et al.’s [44] approach to measuring corporate performance using the net interest rate on total assets.

(2)

Digital transformation (DT). Referring to the research of scholars such as Zhen H.L. [45], 139 digitization-related word frequencies under the categories of technology classification, organizational empowerment, and digital application are counted. By crawling the annual reports of listed companies from 1999–2023, the original report text is organized into panel data, the text length of the full text of the annual report is further counted, the text length of the full text in the Chinese and English parts is counted, and then the dictionary of digitization terms is constructed. The number of exact words is counted after removing the stop words, and the computed word frequency of digitization transformation and the word frequency of the level of each dimension is used as the digital transformation degree index. The higher the digitization index, the higher the level of enterprise digital transformation.

(3)

Mediating variables. The number of patents obtained more directly reflects the enterprise’s achievements and strength in technological innovation; therefore, referring to the research of Zhong C.B. et al. [46], this is the total number of utility model patents and design patents obtained in the year plus 1 to take the logarithm to measure exploitative innovation (I), and the number of invention patents obtained in the year plus 1 to take the logarithm to measure exploratory innovation (R). Also, concerning existing studies [35,47], the balanced metric is used to measure binary innovation balance (BA), and the product term of binary innovations is used to measure binary innovation complementarity (CP), which is given in the following formulas:

$${\mathrm{BA}}_{i,t}=1-{\displaystyle \frac{|I_{i,t}-R_{i,t}|}{I_{i,t}+R_{i,t}}}$$

(1)

$${\mathrm{CP}}_{i,t}=I_{i,t}\times R_{i,t}$$

(2)

(4)

Managerial power (MP). To comprehensively reflect the size and distribution of the power of corporate management, it is necessary to consider the position and influence of the management within the company, as well as the management’s economic interests and other factors, so concerning the study of Liu J.M. et al. [48], the general manager’s years of service, the two positions in the board of directors, the size of the board of directors, the proportion of internal directors, and the proportion of management shareholding is included in the management power measurement system, and principal component analysis is used to obtain a composite score to measure it.

(5)

Control variables. These refer to the existing research to control the factors that may affect enterprise performance [44,45,46]: asset–liability ratio (DAR), capital intensity (CI), retained earnings ratio (RE), cash ratio (CR), Tobin’s Q value (TQ), and environmental competitiveness (HHI).

3.3. Model Construction

In order to test the relationship between digital transformation and business performance, the article constructs the following model:

$${\mathrm{EP}}_{i,t}={\beta }_0+{\beta }_1{\mathrm{DT}}_{i,t}+{\beta }_i{\mathrm{Controls}}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{it}$$

(3)

In Equation (3), EP_i,t represents enterprise performance; DT_i,t represents digital transformation; Controls represents control variables; ${\lambda }_i$ represents individual fixed effect; ${\mu }_t$ represents time fixed effect; and ${\epsilon }_{it}$ represents the random disturbance term.

In order to verify whether exploitative innovation (I), exploratory innovation (R), balanced binary innovation (BA), and complementary binary innovation (CP) play a mediating role in the process of digital transformation affecting enterprise performance, fixed-effects hierarchical regression analysis is used to verify the hypotheses proposed in the theoretical analysis. The basic model is constructed according to the specific needs of the analysis, and the mediation effect measurement model is constructed as follows:

$$I_{i,t}={\delta }_0+{\delta }_1{\mathrm{DT}}_{i,t}+{\delta }_i{\mathrm{Controls}}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{it}$$

(4)

$${\mathrm{EP}}_{i,t}={\alpha }_0+{\alpha }_1{\mathrm{DT}}_{i,t}+{\alpha }_2{I}_{i,t}+{\alpha }_i{\mathrm{Controls}}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{it}$$

(5)

To assess the moderating role of managerial power (MP) in the relationship between digital transformation and enterprise performance, we construct a measurement model for this moderating effect as follows:

$${\mathrm{EP}}_{i,t}={\theta }_0+{\theta }_1{\mathrm{DT}}_{i,t}+\theta 2{\mathrm{MP}}_{i,t}+{\theta }_i{\mathrm{Controls}}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{it}$$

(6)

$${\mathrm{EP}}_{i,t}={\iota }_0+{\iota }_1{\mathrm{DT}}_{i,t}+{\iota }_2{\mathrm{PW}}_{i,t}+{\iota }_3\mathrm{DT}\times {\mathrm{MP}}_{i,t}+{\iota }_i{\mathrm{Controls}}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{it}$$

(7)

In Equation (7), DT × MP represents the interaction term between the explanatory variable digital transformation (DT) and the moderating variable management power (MP). The significance of this interaction term provides evidence for a moderating effect while all other variables are maintained as specified in the foundational regression model.

4. Empirical Analysis

4.1. Correlation Analysis and Descriptive Statistics

Pearson correlation coefficients were calculated, and none of the correlation coefficients between the variables exceeded 0.8. The mean value of the variance inflation factor VIF was 1.26, indicating no severe multicollinearity problem. Descriptive statistics of all variables are shown in

Table 1. Definition of variables and descriptive statistics.

Variable Types	Symbols	Method of Measurement	Sample Size/One	Average	Standard Deviation	Minimum	Maximum
Independent variables	DT	Combining text analytics to calculate the Enterprise Digital Transformation Index	20,630	2.065	1.506	0.000	7.444
Dependent variable	EP	Net interest rate on total assets	20,630	0.038	0.125	−14.302	0.493
Mediating variables	I	The sum of the number of utility models and designs obtained plus one takes the natural logarithm	20,630	1.189	1.438	0.000	7.799
	R	Number of patents for inventions granted in the year plus one in natural logarithms	20,630	0.721	1.018	0.000	7.195
	BA	1 − |I − R|(I + R)	12,069	0.391	0.378	0.000	1.000
	CP	I × R	20,630	1.597	3.668	0.000	45.352
Moderating variables	MP	Principal component analysis	20,630	−0.176	0.988	−2.514	1.984
Control variables	DAR	Total enterprise liabilities/total assets	20,630	0.417	0.232	0.008	10.495
	CI	Net fixed assets at end of year/total assets at end of year	20,630	2.841	3.035	0.424	22.937
	RE	Percentage of assets with retained earnings	20,630	0.136	0.725	−53.417	1.046
	CR	Sum of money funds and trading financial assets/current liabilities	20,630	0.999	1.946	1 × 10	70.449
	TQ	Tobin’s Q value	20,630	2.138	1.975	−4.192	92.299
	HHI	Herfindahl–Hirschman Index	20,630	0.142	0.143	−0.412	2.817

. Among them, the dual innovation balance caused some samples to be invalid during the calculation process due to decimal calculations, and only 12,069 valid data were used in the subsequent test of the mediating effect of the dual innovation balance. The minimum value of enterprise performance is −14.302, the maximum value is 0.493, and the standard deviation is 0.125, indicating that the overall business performance of China’s listed enterprises is more balanced, but the level is still on the low side overall. The maximum value of enterprise digital transformation (DT) is 7.444, the minimum value is 0, the average value is 2.065, and the standard deviation is 4.506, which, on the one hand, indicates that the overall level of digital transformation in China’s listed enterprises is yet to be further improved, and on the other hand, there is a significant gap in the degree of digital transformation among enterprises.

4.2. Base Return

The results of the benchmark regression on digital transformation and firm performance are shown in

Table 2. Benchmark regression of digital transformation on enterprise performance.

Variables of Interest	Base Regression	Introducing Interaction Terms		Base Regression
	(1)	(2)	(3)	(1)
	EP	EP	EP	EP
DT	0.006 *
	(1.740)
DT × I		0.001 ***
		(3.386)
DT × R				0.001 ***
				(3.614)
Controls	Yes	Yes		Yes
Constants	0.103 ***	0.107 ***		0.107 ***
	(4.169)	(4.663)		(4.680)
Year	Yes	Yes		Yes
Firm	Yes	Yes		Yes
N	20,630	20,630		20,630
R2	0.243	0.265		0.263

Note: (1) *, *** indicate significant at the10% and 1% levels of statistical significance, respectively. Values in parentheses are t-values. (2) <0.000 means that the data have an actual value but the three effective digits after the decimal point are zero.

. As shown in column (1), the regression coefficients are positive and significant, indicating that digital transformation positively contributes to firm performance; thus, hypothesis H₁ is verified. Column (2) introduces the interaction term between digital transformation and exploitative innovation (DT × I), and column (3) introduces the interaction term between digital transformation and exploratory innovation (DT × R); the regression coefficients are positive and significant, indicating that digitization and dual innovation jointly promote enterprise performance, which provides the basis for the subsequent mediation effect test.

4.3. Mediating Effect Test

To further clarify the impact of dual innovation between digital transformation and enterprise performance, the estimation results of the mediation effect model are shown in

Table 3. A test of the mediating role of dual innovation.

Variables	Base Regression	Exploitative Innovation Mediation Regression Test		Exploratory Innovation Mediating Regression Test		Mediation Regression Tests for Dual Innovation Equilibrium		Dual Innovation Complementarity Mediation Regression Test
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	EP	I	EP	R	EP	BA	EP	CP	EP
DT	0.006 *	0.023 **	0.006 *	0.016 **	0.006 *	0.011 **	0.001 **	0.072 ***	0.006 *
	(1.740)	(2.075)	(1.734)	(2.057)	(1.730)	(2.082)	(2.022)	(2.648)	(1.729)
I			0.001 *
			(1.778)
R					0.002 ***
					(2.607)
BA							0.002 *
							(1.848)
CP									4.4 × 10−4 **
									(2.332)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constants	0.103 ***	1.009 ***	0.102 ***	0.613 ***	0.102 ***	0.310 ***	0.070 ***	1.013 ***	0.103 ***
	(4.169)	(20.765)	(4.141)	(19.028)	(4.155)	(10.689)	(9.483)	(9.287)	(4.163)
N	20,630	20,630	20,630	20,630	20,630	12,069	12,069	20,630	20,630
R2	0.243	0.061	0.247	0.032	0.247	0.017	0.305	0.031	0.246

Note: *, **, *** indicate significant at the10%, 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

. The coefficients in columns (2) and (3) are positive and pass the significance test, indicating that digital transformation actively promotes exploitative innovation and exploitative innovation plays a mediating role in the digital transformation process driving enterprise performance, and hypotheses H_2a and H_3a hold. The coefficients in columns (4) and (5) are positive and pass the significance test, indicating that digital transformation actively promotes exploratory innovation, and exploratory innovation plays a mediating role in the digital transformation process driving enterprise performance, hypothesizing that H_2b and H_3b are valid. The regression coefficients of digital transformation on the regression coefficients of binary innovation balance and binary innovation balance on enterprise performance in columns (6) and (7) are positive and pass the significance test, indicating that binary innovation balance mediates the relationship between digital transformation and enterprise performance, and the hypothesis H_3c holds. The regression coefficients of digital transformation on binary innovation complementarity and binary innovation complementarity on enterprise performance in columns (8) and (9) are positive and pass the significance test, indicating that binary innovation complementarity mediates between digital transformation and enterprise performance, and the hypothesis H_3d holds. Through observation, it can be found that digital transformation promotes enterprise exploitative innovation more than exploratory innovation, and the significant index of digital transformation on enterprise performance decreases after the introduction of binary innovation, indicating that binary innovation plays a partially mediating role between digital transformation and enterprise performance.

4.4. Moderating Effect Test

The explanatory variable digital transformation and the moderating variable management power are centralized to avoid their cross-terms generating multicollinearity in the regression, and the results of the moderating effect test are shown in

Table 4. Testing for the moderating effect of management power.

Variables	(1)	(2)
	EP	EP
DT	6.21 × 10−3 ***	6.08 × 10−3 ***
	(4.761)	(4.657)
MP	0.003 *	−0.003
	(−1.704)	(−1.632)
DT × MP		0.00169 *
		(1.881)
Controls	Yes	Yes
Constants	0.103 ***	0.104 ***
	(18.341)	(18.438)
N	20,630	20,630
R2	0.243	0.243

Note: *, *** indicate significant at the10% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

. The regression coefficients of the cross terms in column (2) are positive and pass the significance test, which indicates that management power has a positive moderating effect between digital transformation and enterprise performance, and thus hypothesis H₄ is established.

4.5. Heterogeneity Analysis

Influenced by regional variability, high-tech industries, and enterprises’ business status, the enhancement effect of digital transformation on enterprise performance may be characterized by significant heterogeneity due to enterprise characteristics. The results of the heterogeneity discussion are shown in

Table 5. Heterogeneous regression results of digital transformation on enterprise performance.

Variables	EP
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Eastern	Central	Western	High-Tech Industry	Non-High-Tech Industries	High-Loss Enterprises	Low-Loss Enterprise
DT	0.002 ***	0.020 ***	0.002	0.002 ***	0.009	−0.001 **	0.004 *
	(3.450)	(3.073)	(1.065)	(2.812)	(1.352)	(−2.512)	(1.792)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constants	0.123 ***	−0.003	0.113 ***	0.115 ***	0.016	0.104 ***	0.104 ***
	(40.725)	(−0.104)	(11.106)	(14.622)	(0.162)	(13.351)	(5.168)
N	14,889	3497	2244	12,291	8339	10,363	10,267
R2	0.343	0.046	0.375	0.419	0.124	0.452	0.112

Note: *, **, *** indicate significant at the10%, 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

.

4.5.1. Heterogeneity Test Based on Region

Considering the differences in the policy environment of digital transformation in different regions, the total sample is split into three regional subsamples, namely, the eastern region, central region, and western region, according to the location of the enterprises’ offices and the results n columns (1)–(3) of Table 5 show that the promotion effect of digital transformation on enterprise performance is more evident in the eastern and central regions, which may be because, compared with the western region, the eastern and central regions are more economically developed, and enterprises in this region have more capital and technological base to support digital transformation, which has a more noticeable effect on enterprise performance improvement.

4.5.2. Heterogeneity Test Based on High-Tech Industries

Considering that high-tech industries usually receive more policy support and market attention from the government, the enterprise samples are divided into high-tech and non-high-tech industry groups according to whether they are high-tech industries or not, and the results in columns (4) and (5) of Table 5 show that the enhancement effect of digital transformation on enterprise performance is more evident in the high-tech industries, which may be because the high-tech industries are more likely to acquire and apply new technologies such as big data, cloud computing, etc., which provide an excellent technological foundation for digital transformation to enhance its contribution to enterprise performance.

4.5.3. Heterogeneity Test Based on Firms’ Profitability Status

Due to differences in resources and capabilities, loss-making and profitable firms may also differ in the implementation path of digital transformation. The sample is divided into a high loss-making enterprise group and a low loss-making enterprise group based on the median enterprise loss, and the results in columns (6) and (7) of Table 5 show that compared with high loss-making enterprises, digital transformation has a more significant effect on promoting the performance of enterprises with good operating status. The possible reason for this is that poorer business results may lead to increased financing costs and wider funding gaps, exacerbating the difficulty of enterprise digital transformation and making it difficult to exert the facilitating effect of digital transformation on enterprise performance or even producing a negative effect.

4.6. Robustness Check

4.6.1. Endogeneity Test

The instrumental variable method was used to test the endogeneity problem. Drawing on the study of Kong L. et al. [49], the mean value of digital transformation in the same industry in the same year is used as the instrumental variable (IV), and the instrumental variable is subjected to the over-identification test and the weak instrumental variable test. The results of the two-stage regression are shown in columns (1) and (2) of

Table 6. Robustness test of basic regression of digital transformation on enterprise performance.

Variable	Test for Endogeneity		Exclusion of Special Economic Zones	Exclude Special Event Shocks	Replace the Core Variable Measure
	(1)	(2)	(3)	(4)	(5)	(4)
	Stage 1	Stage 2	EP	EP	Replacing the Measurement of Explanatory Variables	Replacement of Explanatory Variables Measurement
	DT	EP			EP1	EP
IV	0.912 ***
	(36.4771)
DT		0.050 ***	0.007 ***	0.002 ***	0.001 *
		(9.963)	(4.036)	(3.770)	(1.899)
DT1						0.005 *
						(1.712)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Constants	−0.077 **	−0.036 ***	0.094 ***	0.124 ***	0.123 ***	0.105 ***
	(−2.035)	(−2.680)	(13.117)	(44.583)	(13.695)	(4.301)
N	20,630	20,630	15,575	18,156	20,630	20,630
R2	0.501	0.062	0.233	0.356	0.243	0.246

Note: *, **, *** indicate significant at the10%, 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

: The first stage’s regression coefficient is significant, indicating that the instrumental variable can explain the endogeneity variables better, and the selected instrumental variable is reasonable and valid; in the second stage, the regression coefficients are still significant, indicating that the regression results are still reliable after considering the endogeneity issue.

4.6.2. Exclusion of Interference from Special Economic Regions

Municipalities directly under the central government have certain special characteristics in terms of policy autonomy and the attractiveness of attracting investment, which may lead to differences in the digital transformation efforts of firms located in these cities compared to those in general provinces. In this regard, this paper excludes the sample of enterprises located in municipalities and re-runs the regression, and the regression results, as shown in column (3) of Table 6, are consistent with the previous conclusions.

4.6.3. Deletion of Sample Data During Stock Market Crashes and Epidemics

Given the background of economic globalization, domestic and international shocks of economic events affect the process of enterprise digitization. In the time series of the sample data, two crucial events affecting economic fluctuations occurred: the Chinese stock market crash in 2015 and the COVID-19 pandemic in 2020. Taking into account the characteristics of the crisis of the posteriority of the crisis, this paper excludes the samples of the enterprises of the years 2015–2016 and 2020–2021. The regression results, as shown in the regression results in column (4) of Table 6, are consistent with the previous findings.

4.6.4. Replacement of Core Explanatory Variable Measures

(1)

Replacing the measurement of explanatory variables. Total return on assets (ROA) takes into account all of the assets of the enterprise, including net assets and liabilities, which can comprehensively reflect the ability of the enterprise to obtain corporate benefits through the use of total assets, including net assets and liabilities, and adopt ROA to re-measure the performance of the enterprise, and the regression data are shown in column (5) of Table 6.

(2)

The replacement of explanatory variables measurement referred to the study of Wu F. et al. [17] to replace the object of digitized word frequency statistics. Focusing on specific digital business scenarios application, the level of digital transformation was re-measured, with regression data as shown in column (6) of Table 6.

(3)

Replacing the measurement of mediating variables. The number of patent applications can reflect the activity and investment of enterprises in technological innovation, so the number of invention patents applied for by enterprises in the same year is used to re-measure the dual innovation for the mediation effect test, and the regression data are shown in columns (1)–(8) of

Table 7. Robustness test of mediating effect and moderating effect.

Variables	Using Innovative Intermediary Role Robust Regression Test		Exploratory Innovation Intermediary Role Robust Regression Test		Equilibrium Robustness Tests		Complementarity Robustness Test		Management Power Moderation Test
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
	I	EP	R	EP	BA	EP	CP	EP	EP	EP
DT	0.020 *	0.006 *	0.033 ***	0.006 *	0.009 *	0.001 *	0.087 **	0.006 *	6.17 × 10−3 ***	6.07 × 10−3 ***
	(1.819)	(1.734)	(3.141)	(1.727)	(1.772)	(1.950)	(2.159)	(1.729)	(4.733)	(4.653)
I		0.001 **
		(2.203)
R				0.001 **
				(2.365)
BA						0.003 **
						(2.192)
CP								4.89 × 10−4 ***
								(3.188)
MP2									−0.002	−0.002
									(−0.923)	(−0.831)
DT × MP2										1.62 × 10−3 *
										(1.730)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constants	1.140 ***	0.102 ***	1.008 ***	0.102 ***	0.452 ***	0.063 ***	2.031 ***	0.103 ***	2.031 ***	0.103 ***
	(24.016)	(4.098)	(25.259)	(4.125)	(14.382)	(9.567)	(13.876)	(4.130)	(13.876)	(4.130)
N	20,630	20,630	20,630	20,630	12,293	12,293	20,630	20,630	20,630	20,630
R2	0.048	0.248	0.099	0.248	0.008	0.299	0.049	0.249	0.243	0.243

Note: *, **, *** indicate significant at the10%, 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

.

(4)

The replacement of moderating variable measurement. Replacing the management power composite indicator system to re-measure the moderating variables in order to avoid chance results [50], the regressions are shown in columns (9) and (10) of Table 7. The regression results are consistent with the previous findings, indicating that the analysis obtained in the previous section is robust.

5. Discussion

The growth of the digital economy has highlighted the need for environmentally resilient and forward-looking technological strategies, and enterprise performance is directly linked to sustainable economic development. To date, numerous studies have delved into the value of digital technologies in driving sustainable economic development from different countries, industries, and perspectives across the globe. Pigola et al.’s study on Latin American countries shows that digital entrepreneurial ecosystems can significantly benefit sustainable economic development by strategically developing a part of digitization [51]. While in Pakistan, Soomro studied local SMEs using Structural Equation Modeling (SEM) and Artificial Neural Networks (ANN) techniques and found that social media applications, big data analytics, IoT applications, and blockchain applications play a key role in creating economic and social value for SMEs [52]. British academic Jibril argues that owner-managers who prioritize the SDGs are more likely to drive innovation in new products and processes and that digital technology further strengthens the link between the SDGs and product and process innovation [53]. In addition, digital infrastructure/products (hardware) and farmers’ digital capabilities (software) are seen as the two pillars that drive digital sustainability transformation, regardless of the country’s stage of development or policy context [54]. Lin et al. developed a moderating intermediary model based on the new economic growth theory and information asymmetry theory to test the positive impact of digital transformation on sustainable innovation performance through the alignment of digital technology and business [55]. Meanwhile, other scholars have taken Chinese state-owned enterprises (SOEs) as the object of their research and affirmed the role of digital transformation in technological innovation in terms of “improving quality and increasing quantity” [56]. In the quasi-natural experiment of the “Belt and Road” initiative [57], the promotion of the dual innovation equilibrium around the competition effect and the degree of competition in the product market was studied in depth.

Based on the theory of dual innovation, this paper constructs a relationship model between digital transformation and enterprise performance. It explores the mediating role of dual innovation between digital transformation and enterprise performance and the moderating role of management power between digital transformation and enterprise performance to provide theoretical insights for enterprises attempting to improve economic efficiency through digital transformation. Through empirical testing, firstly, the use of digital technology provides strong data support and analytical capabilities, and even at a lower level of digital development, the use of digital technology on the efficiency and feasibility of research and development of technology still helps to improve enterprise efficiency (Hypothesis H₁), emphasizing that attention should be paid to the promotion of digital strategy for sustainable development, in line with the conclusions of existing research. Secondly, this paper further pairs the analysis of the impact of digital transformation on exploitative innovation and exploratory innovation behaviors (Hypotheses H_2a and H_2b) and finds that digital transformation promotes both exploitative innovation and exploratory innovation, but exploratory innovation itself has higher risk and uncertainty, so companies may prefer to use digital transformation to optimize and enhance existing knowledge and skills. It is also found that dual innovation differentially improves firms’ product quality and competitiveness (Hypotheses H_3a and H_3b), with exploratory innovation bringing new knowledge and technology to firms to satisfy potential future market demand and exploitative innovation transforming this new knowledge and technology into actual productivity and market competitiveness to improve short-term market performance. Meanwhile, in the context of how firms gain benefits from continuous technological innovation in competitive markets, this paper proposes that this process tends to focus more on the synergistic effects of innovation (Hypotheses H_3c and H_3d) and that the balanced and complementary nature of binary innovation enables firms to respond to market demands and innovation directions promptly, significantly enhancing enterprise performance. In addition, digital transformation often requires strategic decisions and strong execution at the enterprise level. Through its power, the management tilts more resources to the key areas of digital transformation. This concentrated investment of resources helps to accelerate the pace of digital transformation and enhance the overall competitiveness of the enterprise, which ultimately manifests itself in a positive impact on enterprise performance (Hypotheses H₄).

6. Conclusions and Recommendations

This paper selects the relevant data of Chinese listed companies from 2012 to 2022 as the initial research sample, empirically examines the impact of digital transformation on dual innovation and enterprise performance as well as the role mechanism of dual innovation and management power between the two, and further analyzes the heterogeneous impacts of different regions, whether it is a high-tech industry, and the loss status of the enterprise, and concludes that the research is as follows: (1) Digital transformation can effectively promote enterprise performance and the level of dual innovation, and the effect of digital transformation on the enhancement of exploitative innovation is more potent than that of exploratory innovation, but at present, there are still phenomena such as the low degree of the overall digitalization level of the enterprise and the big difference, which indicates that there is a large room for improvement in the promotion of the role of digital transformation on the performance of the enterprise. (2) Exploitative innovation, exploratory innovation, dual innovation equilibrium, and dual innovation complementarity play a mediating role in the digital transformation enhancement of enterprise performance. (3) Managerial power can play a positive moderating role between digital transformation and enterprise performance. (4) The results of heterogeneity analysis show that the effect of digital transformation on firm performance is more significant in the east and central regions, high-tech industries and firms with good profitability status, and the above conclusions still hold after the robustness test.

6.1. Policy Suggestions

(1)

Build an environment for digital transformation and extend the value of digital technology applications. The government can introduce a series of incentives, including tax breaks, capital subsidies, and other preferential policies, to reduce the economic burden of enterprises at the initial stage of digital transformation and increase their enthusiasm for participating in the transformation. Enterprises can consider increasing their R&D investment in cutting-edge digital technologies such as 5G Internet and artificial intelligence, deeply integrating digital technologies into their daily business processes, management modes, and operation strategies, realizing the reshaping and optimization of business processes, enhancing production efficiency and product quality, improving enterprise competitiveness and market adaptability, and giving full play to the prying effect of digital transformation on enterprise performance.

(2)

In a relatively dynamic environment, focus on the formation of the synergistic interaction of dual innovation. Enterprises should reasonably distinguish and utilize dual innovation according to their innovation level, market demand, and technological development trend, realize the balance of dual innovation, and actively utilize the complementarity of the two to promote each other. Enterprises with strong innovation ability can focus on the research and development of new technologies and new products to obtain long-term competitive advantages. Enterprises with weak innovation ability can focus on the optimization and upgrading of existing technologies and products to quickly respond to market changes while giving full play to the rapid response and cost-effectiveness advantages of exploitative innovation in the short term and also focusing on the leading role of exploratory innovation in long-term development, and, through the reasonable allocation of resources and organizational arrangements, realize the balance and complementarity of the two modes of innovation.

(3)

Strengthen management’s knowledge of digital transformation and optimize management’s power allocation and decision-making mechanism. Enterprises can conduct digital transformation training for management to enhance their knowledge of digital technology, data-driven decision-making, and innovative management models, ensure that management can deeply understand the strategic significance of digital transformation, and at the same time, rationally allocate management power to ensure that they have sufficient power in key decision-making areas, to be able to quickly and effectively respond to the challenges of the digital transformation process.

(4)

Based on enterprises’ characteristics, implement dynamic and differentiated digital transformation strategies.

In driving enterprise digital transformation to improve performance, the eastern and central regions and the west must define their transformation goals and choose the right digitalization tools. Enterprises in the eastern and central regions can focus on strengthening data governance, optimizing organizational structures and processes, and cultivating digital talent to unleash the full potential of digital transformation. Meanwhile, western regions must increase infrastructure development, learn from advanced experiences, focus on specialty industries, and strengthen talent training and introduction. In addition, enterprises from East and Central China should be encouraged to invest in the western region to promote the development of the digital economy in the west.

For high-tech industry enterprises, it is recommended that they deepen and expand based on existing technology, strengthen the key core technology research and development, and construct independent innovation capacity in order to enhance their competitive advantage. For example, through intelligent transformation, Shandong Weiqiao Textile successfully built a set of textile, printing and dyeing, clothing, and home textiles in one of the complete chains of high-end, intelligent, green production systems. Traditional industry enterprises must be based on their own business characteristics and market demand, as well as clear transformation goals, paths, and measures to achieve a transition from the traditional business model to the digital, intelligent direction of the transformation. For example, China Fifteen Metallurgical Construction Group Co., Ltd. has formed an enterprise operation management system with its characteristics through digital transformation, realizing all-around and whole-process fine supervision and management of engineering projects.

Enterprises with good profitability should further use digital technology to expand their business, optimize their operations and management, and innovate their products and services through big data, artificial intelligence, and other technologies to meet changes in market demand. On the other hand, loss-making enterprises must first clarify their needs and pain points in digital transformation, avoid unthinkingly following the trend or over-investment, and identify the key areas and priorities of digital transformation through internal diagnosis and external consulting to realize precise transformation.

6.2. Points for Improvement

The study has some limitations. First, this study is limited to exploring data from listed companies in China, which is restricted by the available data sources and fails to include data on digital transformation in other countries or regions. Although we are committed to improving the data quality, this limitation undermines the generalizability of the study’s conclusions. Second, the study explores the impact of digital transformation on business effects at a broader level, and further depth is needed in the future, either through specific case studies or by focusing on examples from specific regions or industries, to reveal more detailed impact paths and provide more operationalized recommendations. In addition, while this study addressed how digital transformation affects firms’ operating costs, organizational structure, and market performance through innovation and management in its analysis, it only briefly elaborated on these mechanisms and lacked an in-depth exploration of the balance between the costs and benefits of digital transformation. Therefore, future research should construct a comprehensive cost–benefit analysis model that includes the costs of digital transformation to present a more comprehensive and balanced picture of the multidimensional impact of digital transformation.