In recent years, the world has faced mounting concerns over intensifying global climate change, escalating resource scarcity, and deepening social inequality. The United Nations’2030 Agenda for Sustainable Development has established 17 Sustainable Development Goals (SDGs), providing a comprehensive framework for global action. Concurrently, international treaties such as the Paris Agreement have mandated ambitious carbon emission reduction targets. These developments coincide with a meteoric rise in ESG investing—global assets under management utilising ESG criteria surpassed $30 trillion in 2023—and a shift in sustainability preferences among consumers and employees. Together, these forces have fundamentally reshaped the corporate competitive environment. Against this backdrop, corporations—as major contributors to resource consumption and pollutant emissions—are facing increasingly intense scrutiny regarding their environmental responsibilities. It has become imperative for businesses to urgently incorporate sustainability into their long-term strategic planning, transcend the pursuit of short-term economic gains, and commit to a fundamental transition toward sustainable operations.

Driven by continual advancements in digital technologies—such as big data and cloud computing—China’s digital economy has sustained robust growth. According to official statistics released by the National Bureau of Statistics, the scale of the digital economy exceeded 58 trillion yuan in 2023, representing 46.5% of the nation’s GDP. Digital technologies—such as big data, artificial intelligence, mobile internet, and cloud computing—have emerged as critical drivers of corporate development. Enterprises are increasingly leveraging these digital tools to integrate internal resources, eliminate information silos, and optimise management processes, while actively exploring new models of value creation. Simultaneously, digitalisation enables businesses to tackle the dual challenges of maintaining economic performance and meeting environmental goals. Industry leaders such as Siemens and Toyota have demonstrated that comprehensive digital transformation can yield synergistic gains in both economic and environmental performance.

Existing research presents inconsistent conclusions regarding the relationship between digital transformation and corporate sustainability performance. Although most studies suggest a positive impact of digitalisation on sustainability outcomes1,2, the nature and extent of this relationship remain contested in the academic literature. However, some scholars argue that excessive or misaligned resource allocation during digital transformation could lead to heightened energy consumption and divert investments from other critical initiatives, potentially undermining long-term corporate development3. Simultaneously, digital transformation enhances economic performance by increasing operational and collaborative efficiency, integrating resources, and raising overall productivity4,5,6. In terms of underlying mechanisms, studies have shown that it also indirectly improves corporate sustainability performance through multiple pathways, including strengthening organisational learning and absorptive capacities7,8,9, promoting green technological innovation, and optimising executive team structures10.

This study examines A-share listed companies in China from 2012 to 2022. Grounded in stakeholder theory, the resource-based view, and agency theory, it investigates the mediating roles of internal incentive mechanisms and external governance mechanisms in shaping corporate sustainability outcomes, thereby contributing to the existing literature. Compared with prior studies, this research makes several key contributions. Theoretically, while existing work has largely emphasised direct effects and internal operational factors—such as improving operational efficiency or innovation capacity—or focused on moderating variables, less attention has been paid to the role of external governance mechanisms. This study explicitly analyses the mediating effects of management incentives and analyst attention in the relationship between digital transformation and corporate sustainability performance. This mechanism encourages a transition toward more sustainable development pathways by improving governance effectiveness, realigning stakeholder relationships, and fostering synergy among corporate value, capital market performance, and social benefits.

Fig. 1
figure 1

Conceptual framework of the research.

Full size image

Literature review and research hypotheses

Digital transformation and sustainability performance

Rooted in the resource-based view, a firm’s sustainable competitive advantage derives from its unique resources and capabilities. The impact of digital technology is becoming increasingly important11. Digital transformation can be conceptualised as a strategic process through which enterprises develop and reconfigure digital resources and dynamic capabilities12,13. These newly established assets are the micro-foundation for improving corporate sustainable development performance. Specifically, digital transformation enables enterprises, in the first place, to build digital resource endowments and process optimisation capabilities14,15. By restructuring production processes and organisational architectures, it reduces operational costs, improves efficiency, and enhances supply chain stability and resilience16,17—while also diminishing human error and reducing rates of waste and product returns18. Secondly, digital transformation elevates innovation capacity and capital allocation efficiency19,20, improving overall innovation performance21. Ultimately, it strengthens corporate performance by enabling novel business models—e-commerce and the sharing economy, for example—that open new revenue streams and improve financial outcomes22,23,24.

Furthermore, digital transformation enhances technological innovation capabilities and accelerates R&D cycles and creates broader opportunities for resource integration and market collaboration25, thereby amplifying potential value creation. In the context of environmental governance, the adoption of digital technologies facilitates the optimisation of production processes and techniques. This leads to higher energy efficiency, supports coordinated emission reduction efforts, decreases the consumption of natural resources and the discharge of pollutants, and mitigates the overall environmental footprint of corporate operations26,27. Similarly, the advancement of fintech contributes to stronger corporate governance mechanisms, proving especially impactful in managing carbon emissions28,29,30. Based on the above analysis, this study proposes the following hypotheses:

Hypothesis H1: Digital transformation has a positive impact on corporate sustainability performance.

The mediating effect of management incentives

According to agency theory, the agency problems resulting from separating ownership and management in modern corporations can be alleviated through well-designed monitoring and incentive structures. Corporate digital transformation offers an effective means to enhance corporate governance by mitigating conflicts of interest and reducing information asymmetry. Furthermore, digital transformation facilitates a rational increase in executive compensation and promotes sharing digital dividends between shareholders and management31. Specifically, it enhances corporate governance by refining governance structures and strengthening managerial incentives32. At the same time, digital transformation broadens corporate financing channels, thereby alleviating financing constraints. The resulting improvement in financial flexibility allows firms to allocate capital more effectively—for instance, by implementing equity incentive plans and increasing the proportion of equity-based compensation awarded to management. Simultaneously, executive compensation incentives can serve as a driving force for corporate digital transformation, which mediates such incentives’ effect on curbing real earnings management practices. However, it should be noted that poorly designed cash-based compensation systems—characterised by excessive pay levels, short-term incentive structures that encourage financial manipulation, and insufficient transparency—may undermine the intended motivational effects of compensation contracts. Therefore, digital transformation may improve corporate governance efficiency and promote sustainable development by optimising the design and implementation of executive compensation incentive mechanisms33,34. Based on the above analysis, the following hypotheses are proposed:

Hypothesis H2: Management incentives mediate the relationship between digital transformation and sustainable development performance.

According to stakeholder theory, a company’s survival and development hinge on the support of its various stakeholders, with corporate governance focusing on balancing the expectations and returns of different interest groups. In this process, financial analysts serve as a critical bridge between firms and investors, fulfilling an essential role in transmitting and interpreting corporate information. Digital transformation significantly enhances information transparency and communication efficiency35,36. Regarding information sourcing, data mining and analytics technologies unlock previously dormant information, converting unstructured and unpredictable content into transmissible, standardised digital data37,38, thereby broadening the channels of information acquisition. At the level of information dissemination, digital technologies establish more efficient pathways for the flow of information both within and outside the organisation. Thus, digital transformation fundamentally expands the scope of information companies can disclose. Public and private information constitute the complete data set accessible to analysts. As the volume of information grows, the datasets analysts utilise increasingly converge, enhancing their effectiveness in external oversight. This strengthened monitoring function further incentivises firms to mitigate risks, enhance reputational capital, achieve legitimacy, and create shared value39. Consequently, while maintaining economic performance, companies are driven to evolve into socially recognised, environmentally responsible, and sustainable organisations40. Given the above, we propose the following hypothesis:

Hypothesis H3: Analyst attention mediates the relationship between digital transformation and sustainable development performance.

Based on the analysis provided, we propose the conceptual framework for this paper, illustrated in Fig. 1. In this framework, digital transformation is identified as the core independent variable, while corporate sustainable development performance is the dependent variable. Additionally, analyst attention and executive compensation incentives act as mediating variables.

Methodology

Samples and data collection

We adopted the panel data of A-share listed companies covering 2012–2022 as the research sample for this study. To ensure the data’s objectivity, completeness, and validity, we implemented the following data processing procedures: First, we excluded sample companies with abnormal trading status (i.e., ST- and *ST-designated companies). Second, we removed samples of companies in the financial sector. Third, we excluded samples with severe data missingness. Through this process, we obtained 26,085 valid observations. We sourced data on environmental performance from the Huazheng ESG Rating System, whereas we retrieved other data from the China Social Science Management Research Database (CSMAR). Additionally, we winsorized all continuous variables at the 1% and 99% levels to mitigate the potential impact of extreme values.

Variable measures

Dependent variable

Corporate sustainability performance must encompass both economic viability and environmental sustainability, with their level of synergy directly reflecting an enterprise’s capacity for sustainable development. The long-term stability of return on assets (ROA) is highly correlated with a company’s sustainability capability. The environmental score in Huazheng ESG incorporates quantitative metrics such as greenhouse gas emissions, carbon neutrality pathways, water consumption, waste discharge, and renewable energy usage. It employs a combined positive and negative assessment approach, with negative indicators including industrial emissions, electronic waste, and environmental penalties, providing a comprehensive reflection of corporate environmental performance. Furthermore, the entropy weighting method aligns with the rigorous measurement requirements for sustainability performance. Therefore, the environmental score from Huazheng ESG and the entropy weight of a company’s return on assets (ROA) are selected as proxy indicators for sustainability.

Independent variable

The China Stock Market & Accounting Research Database (CSMAR) employs precise numerical metrics across three dimensions—organisational empowerment, environmental support, and digital outcomes—to measure corporate digital transformation. This approach reduces the impact of discrepancies between annual reports and actual practices on experimental results, offering greater objectivity than existing text analysis methods. This paper utilises the digital transformation index within the database to assess corporate digital transformation. The specific measurement methodology is as follows: Enterprise digital transformation index = 0.3472×strategy-driven score + 0.162×technology-enabled score + 0.0969×organizational-enabled score + 0.0342×environmental-enabled score + 0.2713×digital outcome score + 0.0884×digital application score.

Mediating variable. This study selects compensation incentives, specifically taking the logarithm of management compensation. Higher values indicate more substantial compensation incentives for management.

Mediating variable. This study uses the number of securities analysts tracking a listed company to measure analyst attention. The natural logarithm of the number of teams issuing earnings forecasts for a listed company in each fiscal year, plus one, serves as a proxy variable for analyst attention.

Control variables. In this paper, we have chosen the following control variables for the experiment: firm size(Size), debt-to-equity ratio(Lev), revenue growth rate(Growth), proportion of independent directors(Indep), cash flow ratio(Cashflow), dual-role positions (Dual), firm value(TobinQ), listing tenure(ListAge), audit quality(Big4), and audit opinion(Opinion). Additionally, based on the results of the Hausman test, we employed a fixed-effects model, treating both industry and year as fixed effects.g both industry and year as fixed effects. as fixed effects.

The variables are defined as shown in Table 1.

Table 1 Definition and description of control variables.
Full size table

Model setting

In comparison with the structural equation model, regression analysis is better suited to investigating the specific operational mechanisms within a model. Numerous studies focusing on regression analysis as an empirical approach have been published in mainstream academic journals across the globe. Thus, regression analysis is selected as the empirical method for the present study. Drawing on the theoretical analysis conducted earlier, the following models are estimated to test the research hypotheses put forward above:

First, to verify hypothesis H1, we constructed a regression model examining the relationship between digital transformation and sustainable development performance (1).

$$\:{Dtp}_{it}={\alpha\:}_{0}\:+{\alpha\:}_{1}{Dig}_{it}+{\alpha\:}_{2}Controls+\sum\:Year+\sum\:Industry\:+{\epsilon\:}_{it\:}\:$$
(1)

In Model (1), the subscript i represents firms, and the subscript t represents years. Model (1) employs ordinary least squares regression with fixed effects for year and industry. We control for the effects of time and industry variations on sustainability performance.

Second, to test Hypothesis H2, this study employs managerial compensation incentives (Salary) as the mediating variable. The specific model is as follows(2)༈3༉:

$$\:{Salary}_{it}={\beta\:}_{0}+{\beta\:}_{1}{Dig}_{it}+{\beta\:}_{2}Controls+\sum\:Year+\sum\:Industry\:+{\epsilon\:}_{it}$$
(2)
$$\:{Dtp}_{it}={\gamma\:}_{0}\:+{\gamma\:}_{1}{Dig}_{it}+{\gamma\:}_{2}Salary+{\gamma\:}_{3}Controls+\sum\:Year+\sum\:Industry\:+{\epsilon\:}_{it\:}$$
(3)

Finally, to test hypothesis H3, this paper constructs model (4)(5) with analyst attention as the mediating variable.

$$\:{Analyst}_{it}={\delta\:}_{0}+{\delta\:}_{1}{Dig}_{it}+{\delta\:}_{2}Controls+\sum\:Year+\sum\:Industry\:+{\epsilon\:}_{it\:}\:\:$$
(4)
$$\:{Dtp}_{it}={\eta\:}_{0}\:+{\eta\:}_{1}{Dig}_{it}+{\eta\:}_{2}Analyst+{\eta\:}_{3}Controls+\sum\:Year+\sum\:Industry\:+{\epsilon\:}_{it\:}$$
(5)

Empirical results and discussion

Descriptive statistics

Table 2 displays the results of the descriptive statistics analysis. The mean value of Dtp slightly exceeds the median, indicating a mild right skew in the distribution. This suggests that most firms perform near the average in terms of sustainability. In contrast, the mean value of Dig is below the median, and the distribution also shows a mild right skew, which may indicate that most firms are excelling in digital transformation. There are significant variations in digitisation levels among firms, likely due to differences in the regions and industries in which they operate. The mean value of Analyst Focus is greater than the median, again showing a slight right skew. Meanwhile, the mean value of Salary is close to the median, suggesting that management compensation is relatively evenly distributed among the sample firms. Additionally, there are fewer firms with dual powers than those with separate powers, reflecting the current trend in corporate governance structures. The debt levels of most firms remain within a reasonable range, and the growth data exhibit a normal distribution, indicating a stable growth trend across firms. In auditing, only a small number of firms are audited by the Big Four accounting firms, while the majority receive a standard unqualified audit opinion.

Table 2 Descriptive statistics of variables.
Full size table
Table 3 Correlation analysis of variables.
Full size table

Regression analysis

Basic result

Table 4 presents the results of the primary regression analysis, which examines the relationship between digital transformation and sustainability performance. The findings indicate that larger firm size and greater cash flow are associated with improved sustainability performance. Specifically, the coefficient for digital transformation is 0.001, which is statistically significant at the 1% level (p < 0.01). For each one standard deviation increase in digital transformation level, a firm’s sustainable development performance is expected to rise by about 8% within its variability range. This indicates that digital transformation has a positive impact on sustainability performance, thereby validating hypothesis H1. Additionally, the variance inflation factor (VIF) for all variables is below 10, suggesting that multicollinearity is not a significant issue in the model. These results emphasise that digital transformation can significantly enhance corporate sustainability performance.

Mediating effect analysis

The mediating effect was assessed through stepwise regression, focusing on the role of salary in Models 1, 2, and 3. As shown in the table, the coefficient for Dig in Model 1, the coefficient for Dig in Model 2, and the coefficient for salary in Model 3 were all significant at the 1% level. This finding indicates that salary mediates the relationship between Dig and Dtp. In Model 3, the Dig coefficient remains significant at the 1% level, indicating a direct effect between Dig and Dtp, with wages partially mediating this relationship. Furthermore, based on the regression coefficients from each model, the indirect effect of wages is 0.00001218 (β1γ2), whose positive sign aligns with the 0.0006 direct effect. This result confirms that compensation partially mediates the indirect effect of digitalization on corporate value, accounting for 12.18% of the total effect (β1γ2α1). Essentially, digital transformation optimizes performance measurement systems, enabling the design of more efficient executive compensation schemes. This synergistic effect between managerial incentives and digital transformation ultimately enhances the company’s sustainable development performance. Therefore, H2 is verified.

The mediating effect of Analyst was tested through Models 1, 4, and 5. The coefficients for Dig in Models 1 and 4, as well as the coefficient for Analyst in Model 5, were all significant (at the 1%, 1%, and 5% levels, respectively) in Table 4, indicating that Analyst also mediated this relationship. The coefficient for Dig remained significant in Model 5, confirming its partial mediating effect. The indirect effect mediated by Analyst is 0.0000144 (δ1η2), accounting for 14.4% of the total effect (δ1η2α1). Digital transformation attracts greater analyst attention by generating more structured data and reducing information acquisition costs. Analysts then function as an external governance mechanism, mitigating information asymmetry and easing financing constraints, ultimately enhancing corporate sustainability performance. Therefore, H3 is verified.

Table 4 Digital transformation and corporate sustainability Performance.
Full size table

Robustness tests

Instrumental variables method

This study uses the digital transformation index from the previous period as an instrumental variable and applies two-stage least squares (2SLS) to analyse sustainable development performance. Considering the inherent lag effect in corporate digital transformation, the impact of lagged digital transformation levels on current outcomes validates the instrument. As shown in Column (2) of Table 5, during the first-stage regression, the coefficient for the lagged digital index concerning Dig is significantly positive at the 1% significance level. In the second-stage regression (Column 3), the coefficient of the digital transformation index on sustainability performance is also significantly positive at the 1% significance level. This suggests that, after addressing endogeneity issues, the positive effect of digital transformation on sustainability performance remains strong and consistent.

Propensity score matching method

First, we divided the sample enterprises into two groups based on their level of digital transformation: a high digital transformation group (treatment group) and a low digital transformation group (control group), using the median level of digital transformation as the cutoff point. Next, all control variables were included as covariates, and we employed a 1:1 nearest neighbour matching approach with replacement, utilising a Logit model for the matching process. Finally, regression analysis was conducted using the matched samples. Column (4) of Table 5 results indicate a significantly positive correlation coefficient at the 1% significance level. This confirms that our research findings are robust, even after addressing endogeneity issues, and demonstrates that digital transformation significantly enhances corporate sustainable development performance.

Table 5 Robustness test—Instrumental variables method and propensity score matching method.
Full size table

Leading and lagging explanatory variables

To ensure the robustness of the research findings, this study assessed the digital transformation index for both preceding and lagging periods, with the results shown in Table 6 (Columns 2 and 3). Whether the digital transformation index was measured one period ahead or one period behind, the correlation coefficients remained statistically significant at the 1% level. This confirms the model’s robustness and the reliability of its primary conclusions.

Substituting variables

To further validate the robustness of the research findings, this paper employs the substitute variable method to redefine the dependent variable, Dtp. Specifically, the environmental score from the Bloomberg ESG database is chosen as the metric for environmental performance. We constructed Dtp using return on assets (ROA) and environmental performance (ENVI). The formula used is as follows41: \(\:Dtp=\left[\right(1－|ROA－ENVI\:|)\times\:\:\sqrt{ROA\times\:ENVI}]/1\). Table 6 (Column 4) shows that all correlation coefficients are significantly positive at the 1% level. This finding further supports the credibility of the paper’s main conclusions, demonstrating that the positive impact of digital transformation on corporate sustainable development performance is indeed robust.

Table 6 Robustness test—Leading and lagging explanatory variables and substituting variables.
Full size table

Heterogeneity analysis

Regional heterogeneity of enterprises

The analysis results are shown in Table 7, where columns (2), (3), and (4) present the empirical analysis results for each region, respectively. In the regional grouping analysis, the digitalisation coefficient for the eastern region is significant at the 1% significance level. In comparison, the coefficients for the central and western areas are essential at the 5% significance level. This finding indicates that although digital transformation demonstrates substantial effects across all regions, its impact on enhancing corporate sustainable development performance is more pronounced in the eastern region compared to the central and western areas. From the perspective of institutional theory and complementary assets, the East region benefits from a superior institutional environment and resource endowments. This allows it to maximise the advantages of digital technologies more effectively. Enterprises in the eastern region are more likely to achieve data interconnectivity, enhancing these technologies’ network and scale effects. In contrast, the central and western areas encounter challenges related to market mechanisms, resource availability, and environmental conditions, which limit their ability to fully reap the benefits of digital transformation on corporate sustainable development performance.

Heterogeneity in the Corporate life cycle

From the perspective of the corporate life cycle stage (Columns 5 to 7), the effects of digital transformation are most potent in mature (Coefficient = 0.0005, t = 3.38) and declining (Coefficient = 0.0008, t = 6.05) enterprises, exhibiting high significance at the 1% level. In contrast, these effects are insignificant in growing enterprises (Coefficient = 0.0001, t = 0.48). Considering firms’ core demands and resource constraints at different lifecycle stages: Mature firms typically possess stable cash flows and robust resource bases, enabling them to support large−scale digital investments and withstand transformation risks, thus achieving the most pronounced value enhancement effects. Firms in the decline stage face survival pressures and strategic repositioning, where digital transformation yields higher sustainable development performance. Conversely, growth-stage firms encounter resource constraints that may result in less significant returns on digital investments.

Table 7 Heterogeneity Analysis. ***, ** and * respectively indicate that the parameter Estimation is significant at 0.01, 0.05 and 0.1. The “t” value is in parentheses.
Full size table

Discussion

Unlike the research by Li et al.42., this study finds that digital transformation can effectively enhance corporate sustainability performance. The findings indicate that digital transformation exerts both internal motivational and external oversight effects by increasing management compensation incentives and analyst attention, respectively. These conclusions remain valid after robustness tests employing proxy variables and propensity score matching methods.

Specifically, building on the examination of the digital transformation-sustainability performance relationship, this study proposes two mediating variables. While existing research has focused mainly on analysing moderation effects, exploration of mediating mechanisms remains relatively scarce. digital transformation helps align management interests with non-financial objectives, incentivising greater resource allocation to sustainability initiatives. Concurrently, it significantly enhances corporate transparency, attracting increased analyst scrutiny. Leveraging analysts’ information interpretation and market oversight functions, this heightened attention drives corporate improvements in sustainability practices. The findings indicate that digital transformation can enhance corporate sustainability performance through these two pathways. Furthermore, policies encourage enterprises to adopt new technologies such as artificial intelligence and industrial internet for intelligent upgrades and digital transformation (“Smart Upgrades and Digital Transformation”), while simultaneously promoting green initiatives like energy conservation, carbon reduction, and green manufacturing—as seen in regions like Henan and Anhui. Policies promote the sustainable development of the macroeconomy43. These measures significantly advance sustainability performance. Enterprises should seize this policy window to pursue digital transformation, enhancing sustainable development outcomes.

Conclusions and implications

Conclusions

This study, which is based on the resource-based view, stakeholder theory, and institutional theory, investigates how digital transformation impacts corporate sustainability performance, using a sample of Chinese listed companies. It examines both empirical and theoretical aspects while testing relevant mediating effects. The findings suggest that digital transformation improves corporate sustainability performance by enhancing managerial incentives. Additionally, it can influence external governance through analyst attention. Analysts’ attention and managerial incentives significantly mediate between digital transformation and corporate sustainability performance.

Management implications

From the corporate level

We need to tailor strategies to the stage of development. Growing enterprises should plan a digital approach to expand their market presence rapidly; mature companies must leverage digitalisation to optimise processes and consolidate competitive advantages; declining businesses must explore digital models to pursue business transformation and identify new growth opportunities, fostering sustainable development.

Enterprises should effectively coordinate the internal drivers of management with the external governance role of analysts. Management must build and enhance dynamic capabilities to seize strategic opportunities arising from digital transformation more effectively. Simultaneously, digital transformation can be a key pathway to improve corporate information transparency. Analysts, in turn, should leverage digital technologies to deepen their interpretation and integration of corporate information, thereby conveying insights more efficiently and accurately to relevant market decision-makers. This will ultimately promote sustainable corporate development.

From the government level

The government should strengthen incentives and support for enterprises’ digital transformation by improving the regulatory environment and providing financial subsidies or policy incentives. This facilitates the maximization of benefits derived from digital transformation initiatives.

Limitations and future research

Although this paper examines the mechanisms linking digital transformation and sustainability performance from internal and external perspectives, several limitations remain. First, the research sample is restricted to Chinese A-share listed companies, limiting the scope and temporal relevance of the findings, which may not apply to other enterprises. Second, focusing solely on internal drivers and external governance mechanisms may fail to capture the breadth of sustainability performance fully. Finally, the quantitative metrics used to assess digital transformation and sustainability performance may not fully capture their complexity and multidimensional characteristics. Therefore, future research requires more comprehensive and innovative variables to measure digital transformation and corporate sustainability performance.