A Study on the Impact of Different Organizational Levels on Digital Transformation in Enterprises

Abstract

Based on data from A-share listed companies in Shanghai and Shenzhen between 2007 and 2020, this study categorizes enterprises into decision-making and execution levels. We found that a high degree of digital transformation at the decision-making level enhances enterprise performance, while excessive transformation at the execution level diminishes performance. We tested the influence of the digital transformation degree of the decision-making layer and the executive layer on enterprise performance by constructing an econometric model. The mechanism behind the results was tested, and the influence mechanism between digital transformation and enterprise performance was explored from the perspective of TOE theory. Mechanism analysis showed that digital transformation impacts performance by increasing R&D investment, driving revenue growth, and enabling the efficient use of government subsidies. Managing decision-level expenses also strengthens performance and supports digital transformation. This research offers fresh insights into how technology drives digital transformation in a changing external environment, guiding management at different hierarchical levels in enterprises.

1. Introduction and Literature Review

Since the launch of ‘Industry 4.0’ in 2013, China’s digital landscape has evolved significantly, employing technologies and approaches like AI, IoT, blockchain, cloud computing, and big data. These advancements are integral to China’s modernization. The diverse industrial sectors in China have served as fertile ground for exploring next-gen tech, seamlessly integrating it with the real economy, a pivotal move for supply-side reform and industrial optimization. To promote the digital economy, China has implemented various policies and streamlined top-level designs, driving digital transformation and elevating the manufacturing industry. Traditional industries are steadily digitizing, presenting opportunities for business model innovation, collaboration within the industrial chain, and addressing evolving challenges.

Scholars have varied views on enterprise digital transformation. This article adopts the TOE (Technology–Organization–Environment) framework, considering internal and external factors, along with technology’s impact on organizations. We used this framework to analyze digital transformation architecture.

Technologically, prior research mostly focused on using digital tech to boost efficiency and performance. Information technologies, with those relating to big data serving as a representative example (Zhang Yeqing et al., 2021) [1], offer opportunities for companies to enhance their data capital. Building on existing definitions, Vial (2019) [2] defines digital transformation as using technology to optimize physical objects by connecting data and information. Domestic and international research has found that factors such as information technology infrastructure (Powell T C, 1997) [3], digital talents within an organization (Kroh J et al., 2018) [4], and the technological expertise levels of employees (Gurbaxani V and D Dunkle, 2019) [5] all influence the development capabilities of enterprise digital transformation. The integration of digital technology into enterprises fosters resource integration, process construction, organizational restructuring, and the development of transformative operational models (Sun Xinbo et al., 2019) [6]. Blending IT with traditional production factors elevates product value and enhances business agility (Qi Yudong, Xiao Xu, 2020) [7]. However, many enterprises still have weak digital capabilities due to issues like incomplete planning, unclear vision, and inadequate talent training (Li Lan et al., 2022) [8]. Thus, exploring digital transformation solely from a technological perspective is inadequate.

Digital transformation research reveals that the relevant challenges encompass not just technology but also organizational restructuring (Ilvonen et al., 2018) [9]. Moreover, digital tech is linked to production environments and company size (SV B et al., 2021) [10], and organizational culture and readiness are vital aspects (Zhang Z et al., 2021) [11]. Also, digital transformation fundamentally restructures organizations, impacting products, processes, structures, and management philosophies (Verhoef P.C. et al., 2021) [12], leading to new business models. Focusing on organizational management and employees’ digital theory learning is vital. Successful transformation requires collaboration across all organizational levels.

Besides internal factors, digital transformation is influenced by factors like supply chain dynamics, external environmental uncertainty, and competitive pressures (Li Yunhe et al., 2022; Du Yong et al., 2023; Fan Hejun et al., 2023) [8,13,14]. In uncertain decision-making environments, businesses often rely on behavioral cues from their social networks (Chen Qingjiang et al., 2021) [15]. Policy uncertainty affects company development. In such situations, digital transformation bolsters information processing and risk management (Fang Mingyue et al., 2023) [16]. Digital policies significantly influence digital development in specific regions, and the combined effects of these policies are often more effective than individual policy implementations (Wang Xueyuan and Li Xueqi, 2022) [17]. Governments play a pivotal role in resource allocation, using various policy measures such as fiscal subsidies and loan support to empower business development actively (Ni Kejin and Liu Xiuyan, 2021) [18]. Government financial support can directly provide businesses with the resources they need for digital transformation, reducing barriers to entry for enterprises looking to engage in this process.

The existing research underscores that enterprise digital transformation encompasses various organizational aspects. Hence, examining digital transformation from a singular perspective, whether technology-centric or management-focused, is inadequate for unveiling its mechanisms. This paper adopts the TOE (Technology–Organization–Environment) framework, considering the interplay between technology, internal organization dynamics, and the external context. In many enterprises, digital transformation strategies usually originate from the decision-making level. They involve employing digital technology to analyze various organizational aspects, such as activities, processes, business models, and employee capabilities, with the goal of their reconstruction. However, it is important to note that the intentions behind decision-makers’ digital transformation efforts may not always align with the outcomes at the execution level.

Here, one wonders if different levels of digital transformation in an organization affect performance. Do these effects remain consistent, and what drives them?

To address these questions, we used the TOE framework. Python programs were used to extract ‘digital transformation’ words from annual reports. A comprehensive index reflecting decision-making-level transformation was developed, while the execution level was gauged according to the proportion of digital assets to total assets. As enterprise digitalization depends on multiple factors, solely assessing technological progress is insufficient. This paper, incorporating the TOE framework, examines digital transformation’s impact on organizational hierarchy, focusing on technology, organization, and the environment. The empirical results reveal diverse effects on different levels within an organization. The results obtained via mechanism testing show that digital transformation influences performance by boosting R&D spending, staffing, revenue growth, and the utilization of government subsidies. Managing expenses at the decision-making level enhances performance and facilitates digital transformation.

The marginal contribution of this paper lies in the following:

Currently, most studies predominantly explore the causes or effects of enterprise digital transformation from a single dimension. However, in all its complexity, digital transformation is often subject to simul-taneous internal and external influences. Internal factors such as organizational structure, the investment and development levels of internal digital technologies, and external macro-policy environments pro-foundly impact enterprise digital transformation. Clearly delineating the effects of each dimension on en-terprise digital transformation can assist organizations in better leveraging internal and external resources for digital transformation decision making. Therefore, this paper, integrating the TOE (Technology–Organization–Environment) theory framework, comprehensively reveals the relationship between enter-prise digital transformation and organizational levels from the dimensions of technology, organization, and the environment. This work enriches the research framework of digital impact mechanisms. In today’s unpredictable environment, our research provides a new perspective from which enterprises can understand how to efficiently utilize technology to propel digital transformation amidst the constantly changing macro external environment and micro internal organization.

The present research generally provides management recommendations for actors at the decision-making level of enterprises based on research conclusions. However, the sustainable development of a company relies not only on the strategic guidance of affairs at the decision-making level but also on actors at the execution level’s implementation of these strategies. Clearly understanding the impact of the decision-making and execution levels on enterprise digital transformation can help companies formulate more pertinent digital transformation strategies at different levels. Therefore, this paper explores the impact of enterprise digital transformation from the micro perspective of different organizational levels within an enterprise, specifically focusing on the “decision-making level—execution level.” This aids companies in clarifying the relationship between digital transformation and different organizational levels and subsequently provides guidance for management practices at different levels of an enterprise.

2. Theoretical Analysis and Hypothesis Formulation

(1)

Digital Transformation at the Decision-Making Level and Enterprise Performance

The digital transformation has infused new vitality into the development of enterprises, offering opportunities for self-renewal and market expansion. However, traditional brick-and-mortar enterprises, particularly those engaged in extensive equipment operations such as manufacturing firms, exhibit lower levels of digitization. They face high opportunity costs, formidable challenges, and prolonged timelines in the process of digital transformation and upgrading (Xu et al., 2020) [19]. To sustain and deepen digital transformation, enterprises need to develop a strong awareness of transformation, to shift their strategic planning direction (Qi and Xiao, 2020) [20], to acquire the ability to overcome transformation pressures, and to continually search for transformation solutions tailored to their needs.

Key decisions such as digital transformation involve significant roles played by executives and decision-makers within organizations (Ni and Liu, 2021) [18]. These actors bear the responsibility for the comprehensive management of the entire institution, necessitating the establishment of organizational goals to facilitate the realization of targeted solutions. From an environmental perspective, enterprise digital transformation is a long-term process. As helmsmen of enterprise digital transformation, decision-makers need to promptly assess the current environmental challenges facing a company, clarify the existing transformation foundation, and subsequently formulate targeted digital transformation strategies to guide the direction of enterprise digital transformation (Li et al., 2022) [8]. Furthermore, high-level digital transformation decisions are not only influenced by external environmental factors but also shaped by internal organizational structures. Moreover, these decisions reciprocally influence enterprise organizational structure and technological change.

From an organizational perspective, high-level decision-makers often facilitate the deep integration of digital technology and the physical economy during organizational change (Xiao and Qi, 2019) [20]. Additionally, considering technological change, the degree of acceptance and the perception of new technology among decision-makers are critical determinants of corresponding transformation behaviors. In the process of technological innovation, the level of involvement of decision-makers in technological innovation activities determines their attitudes and decision-making behavior. The proactive promotion of specific new technologies by decision-makers not only influences the innovation and research directions of an enterprise but also drives the allocation of resources for a given new technology, aiding in overcoming organizational inertia and promoting innovative behavior within the enterprise (Raffaelli et al., 2019) [21]. Furthermore, decision-makers play a crucial role in the development and allocation of an enterprise’s internal resources and capabilities, contributing to the rapid combination and adjustment of resources in the context of technological change (Helfat and Peteraf, 2015) [22]. Therefore, this paper posits the following hypotheses:

H1: 

The level of digital transformation implemented by decision-makers positively influences enterprise performance.

(2)

The degree of digital transformation at the execution level and firm performance.

The decision-making layer serves as the initiator of and an advocate for digital transformation, while the execution layer is the concrete executor responsible for carrying out specific operational tasks, thereby creating more economic benefits for the enterprise.

From a technological perspective, the digital transformation at the execution layer is reflected in the investment in information equipment and employees’ adoption of new technologies. When new information technologies are integrated into production-related processes, employees undergo relevant training to acquire new operational skills, thereby allowing them to apply these skills to work and enhancing production efficiency. However, the introduction of digital technology also implies that enterprises need to bear additional management costs related to equipment maintenance, employee training, and consultation with specialized institutions for digital transformation, leading to higher learning costs and greater uncertainty (Liu et al., 2021) [23]. This can, to a certain extent, cause systemic disruptions within an enterprise as a whole, increase managerial complexity, and potentially weaken the overall performance of an enterprise (Yu et al., 2017) [24].

From an organizational perspective, insufficient willingness of the operational layer to adopt digital transformation is a crucial influencing factor hindering an enterprise’s ability to drive digital transformation. Employees with a high level of inertia tend to prefer maintaining the status quo, demonstrating lower willingness to adopt changes. Even in the presence of viable alternative choices, if there are no incentive measures in place, employees may be resistant to changing their existing practices, leading to a tendency to resist, particularly when it comes to showcasing higher resistance to new technologies (Hsieh P and W Lin, 2018) [25]. Apart from a lack of willingness to adopt changes, employees may also actively resist the transformation process or engage in passive resistance due to a lack of understanding of the value of digital transformation. From an external environmental perspective, in a complex and dynamic market, excessive and blind development of transformation may yield minimal short-term financial performance improvement for an enterprise, or it may even lead to unintended consequences, resulting in a decline in overall business performance. Some studies have suggested that despite numerous enterprises actively advancing digital transformation, the effects remain inconspicuous (Chen et al., 2020) [26]. Based on these considerations, this paper proposes the following hypothesis:

H2: 

The digital transformation at the execution layer will weaken the effects of driving enterprise performance.

3. Research Design

3.1. Data Source

This study investigates the impacts of different organizational levels on digital transformation using A-share listed companies in the Shanghai and Shenzhen stock markets from 2007 to 2020 as its sample. It examines the digital transformation levels of the decision-making and execution levels and their effects on firm performance.

Data, except for those concerning organizational structure, were sourced from Wind and CSMAR databases and underwent quality screening, wherein (1) financial institutions were excluded; (2) companies with ST, PT, *ST, or similar conditions were removed; and (3) entities with incomplete or abnormal information and those without controlling shareholders were eliminated. In this study, we analyzed 14,000 valid observations from 2007 to 2020 meeting these criteria, with a winsorization process applied to continuous variables to address extreme values.

3.2. Variable Setup

(1)

Dependent Variable.

Digital transformation is linked to intangible assets, long-term value, stock market information, and capital costs, making stock market valuation indicators essential. This paper uses Tobin’s Q, a commonly applied metric, to measure the performance of both decision-makers and executives in determining corporate performance.

(2)

Independent Variable.

Measuring the degree of digital transformation can be challenging. Typically, scholars focus on publicly traded companies and extract keywords related to “digital transformation” from their annual reports using Python programs. However, these reports mainly reflect managerial-level willingness to implement digital transformation and its outcomes, overlooking executive-level progress. To better measure digital transformation at various organizational levels, in this study, we used text analysis applied to the CSMAR database, characterizing managerial-level willingness as the intensity of digital transformation (lnDT).

Table 1. Keywords related to “Degree of Digital Transformation at the Managerial Level”.

Keyword	Definition	Example
Strategic Level	Refers to digital tech in digital transformation and its development trends.	Blockchain, AI, Business Intelligence, IoT, Mobile Internet, Industrial Internet, Mobile Connectivity, Digital Finance, B2B, B2C, C2B, C2C, O2O, and so on.
Operational Level	The primary concern is how technology and equipment deployment support flexible digital transformation expansion.	Smart Financial Contracts, Data Visualization, Investment Decision Support Systems, Machine Learning, Natural Language Processing, Cloud Computing, Stream Computing, Converged Architecture, Digital Marketing, Unmanned Retail, and so on.
Application Level	Refers to the tech and equipment underlying a company’s digital transformation.	Digital Currency, Distributed Computing, Differential Privacy Technology, Big Data, Data Mining, Text Mining, Image Understanding, Intelligent Data Analysis, Semantic Search, and so on.

outlines the keywords extracted when constructing these variables, providing detailed definitions of each keyword and explaining their relationships with the managerial level’s digital transformation. Additionally, the proportion of digital assets to total assets (lndtasset) was calculated and selected to represent the degree of digital transformation at the executive level.

(3)

Control Variables.

Building upon previous research findings, we incorporated a series of control variables into the model to mitigate the impact of other research variables on digital transformation in companies. These control variables specifically include company size (Size), company age (Firm Age), financial leverage (Lev), equity concentration (Top1), profitability (ROA), dual roles (Dual), monthly average excess turnover rate of stocks (Dturn), institutional investor ownership percentage (INST), and audit opinion type (Audittype).

Table 2. Variable definitions.

Variable Type	Variable Name	Variable Symbol	Variable Definition
Dependent Variable	Corporate Performance	TobinQ	Enterprise market price (stock price)/enterprise replacement cost
Independent Variables	Degree of Digital Transformation at the Managerial Level	lnDT	Ln (word frequency related to digital transformation)
	Degree of Digital Transformation at the Executive Level	lndtasset	Net digital assets of the enterprise/total assets
Mediating Variables	R&D Expenditure Ratio	RDSpendSumRatio	(R&D expenditure/sales revenue) × 100%
	R&D Personnel Ratio	RDPersonRatio	(R&D personnel/the total number of employees in the current year) × 100%
	Revenue Growth Rate	Growth	(Revenue growth amount/total revenue of the previous year) × 100%
	Management Expense Ratio	Mfee	(Management expenses/main business income) × 100%
	Government Subsidy	lnsubsidy	Government subsidies received by the enterprise for digital transformation in the current year
Control Variables	Company Size	Size	Ln (total assets for the year)
	Company Age	FirmAge	Ln (current year—the year of a company’s establishment + 1)
	Financial Leverage	Lev	Total liabilities at year-end/total assets at year-end
	Equity Concentration	Top1	Quantity of shares held by the largest shareholder/the total number of shares
	Profitability	ROA	Net profit/the average total assets
	Dual Role	Dual	1 if the Chairman of the Board and the CEO are the same person, and 0 otherwise
	Average Monthly Excess Turnover Rate of Stocks	Dturn	Average monthly turnover rate of stocks in the current year-the average monthly turnover rate of stocks in the previous year
	Audit Opinion Type	Audittype	1 if the audit unit issues a standard unqualified opinion, and 0 otherwise
	Institutional Investor Ownership Ratio	INST	Total shares held by institutional investors/the outstanding shares

provides detailed explanations of the aforementioned indicators.

4. Model Construction

In this study, we constructed Model (1) and Model (2) to separately examine the impact of the digital transformation degree of a company’s managerial and executive levels on the company’s performance. In these models, the dependent variable is the company’s performance level (TobinQ), the key explanatory variables are the strength of digital transformation at the managerial level (lnDT) and the executive level (lndtasset), and the control variables (CVs) include the set of control variables. In this study, we further controlled for for the impact of time (Year) and industry (Industry) factors on a company’s digital transformation to mitigate endogeneity issues. ε_(i,t) represents the random error term.

$${TobinQ}_{i,t}=\alpha +\beta {}_1{lnDT}_{i,t}+\sum \beta {}_i{CVs}_{i,t}+\sum \beta {}_{j}Year+\sum \beta {}_{k}Industry+{\epsilon }_{i,t}$$

(1)

$${TobinQ}_{i,t}=\alpha +\beta {}_{2}ln{dtasset}_{i,t}+\sum \beta {}_i{CV}_{i,t}+\sum \beta {}_{j}Year+\sum \beta {}_{k}Industry+{\epsilon }_{i,t}$$

(2)

The descriptive statistics of the main variables can be found in Appendix A. The average level of digital transformation at the managerial level for the sampled public companies is 2.02, with a standard deviation of 1.13, indicating significant variation in the digital transformation degree among the companies in the sample. On the other hand, the average level of digital transformation at the executive level is 0.19, with a standard deviation of 0.12, indicating that the sampled companies have equipped themselves with various technologies and facilities at the foundational technological level for digital transformation.

5. The Impact of Digital Transformation on Core Business Performance in Enterprises

(1)

Baseline Regression

In this study, we began by exploring the fundamental link between ‘digital transformation degree’ and ‘corporate performance’ and progressively conducted preliminary baseline tests. The empirical results regarding the impact of the digital transformation degree at a company’s managerial and executive levels on corporate performance are presented in

Table 3. Baseline regression: digital transformation and corporate performance.

Variables	(1)	(2)	(3)	(4)
	Managerial Level		Executive Level
	TobinQ	TobinQ	TobinQ	TobinQ
lnDT	0.1133 ***	0.0881 ***
	(0.0192)	(0.0202)
lndtasset			−0.9196 ***	−0.4294 **
			(0.1554)	(0.1683)
Control variables	NO	YES	NO	YES
Fixed effects	NO	YES	NO	YES
N	13,807	10,170	13,807	10,170
$R^2$	0.0077	0.4354	0.0060	0.4331

Note: ** and *** indicate significance at the 5% and 1% levels, respectively; t-values adjusted for robust standard errors are shown in parentheses; factors at the time and industry levels have been controlled for. The same applies to the table below.

.

Regression (1) and Regression (2) in Table 3, based on Model (1), assess the effect of the managerial level’s digital transformation degree on corporate performance. In Regression (1), when including only the core explanatory variable, the results demonstrate a positive association between the managerial level’s digital transformation degree and corporate performance (β = 0.11, p < 0.01). This suggests that the managerial level’s digital transformation significantly promotes digital transformation within a company. In Regression (2), after incorporating enterprise-level and control factors like industry and year, the effect of the managerial level’s digital transformation degree on digital transformation within a company remains significantly positive (β = 0.09, p < 0.01). Economically, a 1% increase in the managerial level’s digital transformation inclination corresponds to a 0.09 increase in corporate performance, equivalent to a 4.0% improvement compared to the sample period mean of 2.17 (i.e., 0.09/2.17 × 100%). This underscores the substantial impact of a stronger managerial-level inclination towards digital transformation on corporate performance both statistically and economically.

When the managerial level, which oversees strategic decisions in a company, displays a stronger inclination towards digital transformation, this leads to the adoption of measures such as increased investment in research and development, the integration of intelligent equipment, and efforts to secure government funding through related initiatives. These actions, over the long term, drive enhancements in corporate performance. Thus, this study provides empirical evidence supporting Hypothesis H1.

Conversely, Regression (3) and Regression (4) in Table 3, based on Model (2), scrutinize the impact of the executive level’s digital transformation degree on corporate performance. In Regression (3), when only the core explanatory variable is considered, the results reveal a negative connection between the executive level’s digital transformation degree and corporate performance (β = −0.92, p < 0.01). This initial finding suggests that the digital transformation degree at the executive level significantly hinders digital transformation within a company. In Regression (4), after introducing enterprise-level and control factors such as industry and year, the influence of the executive level’s digital transformation degree on digital transformation within a company remains significantly negative (β = −0.43, p < 0.05). Economically, a 1% increase in basic digital assets corresponds to a 0.43 decrease in corporate performance, equivalent to a 19.7% reduction compared to the sample period mean of 2.17 (i.e., 0.43/2.17 × 100%). This implies that companies with a higher proportion of executive-level digital assets will experience a decline in performance both statistically and economically.

The research results given above suggest that digital technology, in the realm of management activities, engenders internal workflow changes within a company while simultaneously precipitating overall disruptions and increased management complexity. This includes the introduction of new operational methods for digital technology research, entirely novel workflows, the resolution of unforeseen technical issues, and other costs related to training, consulting, and specialized technology management positions. Managers must also address issues stemming from unfamiliarity with the value of digital technology, such as resistance or passive resistance to technology adoption. Hence, this study provides empirical evidence supporting Hypothesis H2. Subsequently, the mechanism-testing section of this paper delves into the reasons for this negative effect.

(2)

Robustness Checks

To enhance the validity of the hypotheses, this section investigates variable specifications. First, we imposed a lag on the explanatory variables in both Model (1) and Model (2) equal to two to three years and incorporated them back into the two econometric models for regression. The results of the robustness checks (

Table 4. Robustness checks: explanatory variable alterations.

Variables	(1)	(2)	(3)	(4)
	Managerial Level		Executive Level
	TobinQ	TobinQ	TobinQ	TobinQ
L2.lnDT	0.1020 ***
	(0.0270)
L3.lnDT		0.0940 ***
		(0.0318)
L2.lndtasset			−0.4265 **
			(0.1840)
L3.lndtasset				−0.6429 ***
				(0.2133)
Control Variables	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES
N	6046	4520	6046	4520
$R^2$	0.4577	0.4719	0.4543	0.4703

** and *** indicate significance at the 5% and 1% levels, respectively.

) confirm that the digital transformation degree at various levels of a company significantly impacts corporate performance but with varying effects. The driving effect of digital transformation at the managerial level on corporate performance is notably pronounced in the short term, suggesting a top-down transformation process led by managerial decision-makers. Conversely, the impact of digital transformation at the executive level on corporate performance remains significantly negative and becomes more pronounced over time. This underlines the robustness and reliability of the research findings, which are not fundamentally affected by changes in external conditions. This result aligns with the implications of the earlier baseline regression and provides additional corroborative evidence.

Secondly, we substituted the dependent variable. Tobin’s Q measures corporate performance from a financial perspective, while corporate performance is a relatively comprehensive indicator. Therefore, in this article, we changed our measurement perspective and, starting from the comprehensive production efficiency of various elements within a company, replaced the previously measured corporate performance with Total Factor Productivity (TFP_LP) calculated using the LP method. The results are shown in

Table 5. Robustness checks: dependent variable substitution.

Variables	(1)	(2)	(3)	(4)
	TFP_LP	TFP_LP	TFP_LP	TFP_LP
lnDT	0.0642 ***	0.0539 ***
	(0.0182)	(0.0117)
lndtasset			−1.2019 ***	−1.8564 ***
			(0.1913)	(0.1054)
Control Variables	NO	YES	NO	YES
Fixed Effects	NO	YES	NO	YES
N	12,870	10,398	12,870	10,398
R2	0.0045	0.7572	0.0182	0.7824

*** indicates significance at the 1% level.

, and the empirical results remain consistent with the original model.

Furthermore, following the research methodology employed by He Fan et al. (2019) [27], data on the digital transformation of various levels of a company from the previous period (F.lnDT) and (F.lndtasset) were selected as instrumental variables for a two-stage regression test to mitigate potential endogeneity issues in the baseline regression. The results in

Table 6. Endogeneity Test: instrumental variable approach results.

Variables	(1)	(2)	(3)	(4)
	TobinQ	TobinQ	TobinQ	TobinQ
F.lnDT	0.1403 ***	0.0783 ***
	(0.0221)	(0.0209)
F.lndtasset			−1.477 ***	−0.8062 ***
			(0.180)	(0.167)
Control Variables	NO	YES	NO	YES
Fixed Effects	NO	YES	NO	YES
N	9073	8415	9073	8415
$R^2$	0.0128	0.4437	0.0148	0.4439

*** indicates significance at the 1% level.

demonstrate that introducing instrumental variables into the model for regression once again led to statistically significant results, ensuring the robustness of the research findings.

6. Mechanism Testing

The preceding analysis highlights the varying impacts of different organizational levels on the relationship between the digital transformation degree and corporate performance. To elucidate the underlying mechanisms and understand why an increase in executive-level digital assets suppresses corporate performance, this section introduces the TOE theoretical framework. It considers the mechanisms of digital transformation and its impact on corporate performance within the three dimensions of a company’s dynamic external environment, internal organization, and digital technology adoption. The mediation effect model is primarily employed for mechanism identification and testing, as detailed in Models (3)–(8), following the approach described by Wen Zhonglin et al. (2004) [28].

$${TobinQ}_{i,t}=\alpha +\beta {}_1{lnDT}_{i,t}+\sum \beta {}_i{CVs}_{i,t}+\sum \beta {}_{j}Year+\sum \beta {}_{k}Industry+{\epsilon }_{i,t}$$

(3)

$${Mediator}_{i,t}={\alpha }^{\prime }+{\vartheta }_1{lnDT}_{i,t}+\sum {\vartheta }_{i}CV{s}_{i,t}+\sum {\vartheta }_{j}Year+\sum {\vartheta }_{k}Industry+{\epsilon }_{i,t}$$

(4)

$${TobinQ}_{i,t}=\alpha +{\delta }_1{Mediator}_{i,t}+{\delta }_2{lnDT}_{i,t}+\sum {\delta }_{i}CV{s}_{i,t}+\sum {\delta }_{j}Year+\sum {\delta }_{k}Industry+{\epsilon }_{i,t}$$

(5)

$${TobinQ}_{i,t}=\alpha +\beta {}_{2}ln{dtasset}_{i,t}+\sum \beta {}_i{CVs}_{i,t}+\sum \beta {}_{j}Year+\sum \beta {}_{k}Industry+{\epsilon }_{i,t}$$

(6)

$${Mediator}_{i,t}={\alpha }^{\prime }+{\vartheta }_1{}_{2}ln{dtasset}_{i,t}+\sum {\vartheta }_{i}CV{s}_{i,t}+\sum {\vartheta }_{j}Year+\sum {\vartheta }_{k}Industry+{\epsilon }_{i,t}$$

(7)

$${TobinQ}_{i,t}=\alpha +{\delta }_1{Mediator}_{i,t}+{\delta }_{2}ln{dtasset}_{i,t}+\sum {\delta }_{i}CV{s}_{i,t}+\sum {\delta }_{j}Year+\sum {\delta }_{k}Industry+{\epsilon }_{i,t}$$

(8)

6.1. Technological Dimension

Companies, at the application level of technology, develop proficiency in utilizing next-generation information technologies, enhancing infrastructure, and seamlessly integrating business applications. These applications span various facets of a company’s production and operational processes, encompassing areas such as predicting consumer demand and market development in product or service R&D, intelligent equipment management and batch optimization in production, the optimization of operational modes in management, feedback on consumer opinions in sales and service, and the integration of a company’s logistics system in the supply chain.

The effects of technology investment and application extend beyond their immediate implementation, fostering structural adjustments in production factors and stimulating innovation within companies (Liu Weigang, 2022) [29]. These technological factors play a pivotal role in propelling the implementation of digital transformation. Yet, realizing the full potential of technology necessitates not only substantial research and development (R&D) investment but also a focus on employee training and the attraction of technical talent. Thus, assessing the R&D expenditure ratio (RDSpendSumRatio) and the R&D personnel ratio (RDPersonRatio) as intermediate variables within the technological dimension becomes essential. Through the pathway ‘Decision Maker/Executor Digitalization Level → R&D Investment → Firm Performance’, this paper delves into the mechanism linking a company’s R&D investment with its level of digitalization. As demonstrated by (Zhang Yeqing et al., 2021) [1], the digitalization level of decision-makers exhibits a significant positive correlation with the R&D investment ratio, and this ratio, in turn, correlates positively with firm performance. Furthermore, the digitalization level of decision-makers retains a significant positive correlation with firm performance. This suggests that the impact of decision-makers’ digitalization levels on firm performance operates as a partial mediation effect through R&D investment, implying that it influences firm performance both directly and indirectly through R&D investment (

Table 7. Mechanism test for R&D investment.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	Managerial Level			Executive Level
	TobinQ	RDSpendSumRatio	TobinQ	TobinQ	RDSpendSumRatio	TobinQ
lnDT	0.1343 ***	0.9111 ***	0.1001 ***
	(0.0394)	(0.2468)	(0.0342)
RDSpendSumRatio			0.0323 ***			0.0343 ***
			(0.0082)			(0.0084)
lndtasset				0.0908	−4.2802 *	0.3188
				(0.3302)	(2.4954)	(0.3371)
Control Variables	YES	YES	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES	YES	YES
N	3268	3247	3142	3268	3247	3142
$R^2$	0.4558	0.2594	0.4765	0.4493	0.2492	0.4732

* and *** indicate significance at the 10% and 1% levels, respectively.

).

On the other hand, the digitalization level of executives is significantly positively correlated with the technology investment ratio, and the technology investment ratio is significantly positively correlated with firm performance. However, the digitalization level of executives does not have a significant direct effect on firm performance. This indicates that R&D investment in companies plays a fully mediating role in the pathway ‘Executor Digitalization Level→R&D Investment→Firm Performance’. In other words, the impact of the digitalization level of executives on firm performance is entirely mediated through R&D investment as an intermediary variable (

Table 8. Mechanism test for R&D personnel.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	Managerial Level			Executive Level
	TobinQ	RDPersonRatio	TobinQ	TobinQ	RDPersonRatio	TobinQ
lnDT	0.1343 ***	5.5141 *	0.1262 ***
	(0.0394)	(2.9041)	(0.0419)
RDPersonRatio			0.0002 ***			0.0002 ***
			(0.0001)			(0.0001)
lndtasset				0.0908	−59.5952 ***	0.0993
				(0.3302)	(18.3646)	(0.4091)
Control Variables	YES	YES	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES	YES	YES
N	3268	2328	2255	3268	2328	2255
$R^2$	0.4558	0.0129	0.4631	0.4493	0.0128	0.4582

* and *** indicate significance at the 10% and 1% levels, respectively.

).

Furthermore, we delve into the relationship between enterprise R&D personnel input and the digitalization level. As evident in Table 8, there is a significant positive correlation between the digitalization level of decision-makers and the percentage of R&D personnel input. This percentage of R&D personnel correlates significantly and positively with enterprise performance. Interestingly, the digitalization level of decision-makers maintains a significant positive correlation with enterprise performance. The intentions of decision-makers in their digitalization endeavors are a gauge of their commitment to digitalization. For decision-makers, a positive pathway emerged: “Digitalization intentions of decision-makers→Technology investment (promotion)→(enhanced) enterprise performance.” Decision-makers, who shape enterprise strategies, take a range of relevant measures as their digital transformation intentions intensify. Beyond increasing R&D investment, they may introduce intelligent equipment, enhance their organization’s internal information systems, and/or boost digital transformation initiatives to attract a broader consumer base and seek government funding support. These proactive measures positively impact enterprise performance.

The digitalization level of the execution layer is significantly negatively correlated with technology investment. Technology investment is significantly positively correlated with enterprise performance, and the digitalization level of the execution layer is not significantly correlated with enterprise performance. It can be seen that enterprise R&D input plays a complete mediating role in the “digitalization level of the execution layer—R&D personnel input—enterprise performance” path. Since the level of the execution layer’s digital assets is used to measure the digitalization level of the execution layer, it can be seen that a negative path of “execution layer digital assets—technology investment (inhibition)—(reduction) enterprise performance” has formed. This is because when a company has more digital assets, this conveys the message to decision-makers that “digital investment is sufficient”, which provokes decision-makers to reduce technology investment. The weakening of technology investment will inhibit basic work efficiency and negatively affect enterprise performance.

6.2. Organizational Dimension

In the contemporary Chinese business landscape, the quality of decisions made by a company’s leadership dictates the priority of specific initiatives. Grounded in a resource-based perspective, the sufficiency of internal and external resources within an organization becomes crucial during the digital transformation process. Integrating this capability into digital business operations emerges as a critical focus when implementing transformation strategies. Therefore, in the organizational dimension, this paper employs two key indicators—a company’s revenue growth rate (Growth) and management expense ratio (Mfee)—to evaluate a company’s decision-making prowess in terms of production, operations, and governance. This assessment aims to ascertain whether companies are effectively executing their digital transformation efforts.

The empirical results (

Table 9. Mechanism test for revenue growth rate.

Variables	(1)	(2)
	Managerial Level	Executive Level
	Growth	Growth
lnDT	0.0225 ***
	(0.0063)
lndtasset		−0.4451 ***
		(0.1706)
Control Variables	YES	YES
Fixed Effects	YES	YES
N	10,398	10,114
$R^2$	0.0961	0.4332

*** indicates significance at the 1% level.

) indicate that the degree of digital transformation in a company’s decision-making layer is significantly positively correlated with revenue growth rate and overall firm performance. This is because when actors in the decision-making layer have a higher degree of willingness to engage in digital transformation, they are more likely to take actions that promote digitalization, including increased promotion, which, in turn, stimulates revenue growth. On the other hand, the basic level of digital transformation is negatively correlated with revenue growth rate. This is because when basic digital assets are increased, the personnel in the execution layer need to invest more time and effort in training for digital workflow, which has a negative impact on the revenue growth rate.

Table 10. Mechanism test for management expense ratio.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	Managerial Level			Executive Level
	TobinQ	Mfee	TobinQ	TobinQ	Mfee	TobinQ
lnDT	0.0887 ***	0.0031 *	0.0761 ***
	(0.0202)	(0.0016)	(0.0175)
Mfee			3.5670 ***			3.6333 ***
			(0.4795)			(0.4901)
lndtasset				−0.4298 **	0.0215	−0.5120 ***
				(0.1699)	(0.0129)	(0.1653)
Control Variables	YES	YES	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES	YES	YES
N	10,114	10,398	10,114	10,114	10,398	10,114
$R^2$	0.4350	0.3510	0.4630	0.4328	0.3505	0.4619

*, ** and *** indicate significance at the 10%, 5% and 1% levels, respectively.

illustrates a significant positive correlation between the digitalization degree of the decision-making layer and the management expense ratio. Furthermore, the management expense ratio is significantly positively correlated with business performance. The positive relationship between the digitalization degree of the decision-making layer and business performance remained significant. This implies that for the decision-making layer, a “digitization willingness of the decision-making layer—management expenses (increase)—(improve) business performance”-positive pathway has been established. When an enterprise’s decision-making layer exhibits a greater willingness for transformation, it adopts various measures, requiring follow-up by relevant management personnel and thereby increasing management expenses. To offset the rise in management expenses, the enterprise enhances efficiency and reduces costs in other areas, such as through the effective use of digital methods to improve efficiency. These measures also contribute to the improvement of business performance. However, the digitization of the execution layer does not have a significant impact on the management expense ratio, indicating that the management expense ratio does not serve as an intermediary in the “execution layer digitization degree—management expense ratio—business performance” pathway. This is because the execution layer has a considerably lower awareness of management expenses compared to the decision-making layer, and, therefore, this pathway does not act as an intermediary for the execution layer.

6.3. Environmental Dimension

This study primarily employs the number of government subsidies (lnsubsidy) as an indicator to characterize the external policy environment of businesses.

Table 11. Mechanism test for government subsidies.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	Managerial Level			Executive Level
	TobinQ	lnsubsidy	TobinQ	TobinQ	lnsubsidy	TobinQ
lnDT	0.0910 ***	0.0845 **	0.0894 ***
	(0.0223)	(0.0378)	(0.0220)
lnsubsidy			0.0201 **			0.0207 ***
			(0.0079)			(0.0079)
lndtasset				−0.5042 **	−0.9171 **	−0.4857 *
				(0.2541)	(0.3941)	(0.2521)
Control Variables	YES	YES	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES	YES	YES
N	6907	7075	6902	6907	7075	6902
$R^2$	0.4682	0.1392	0.4690	0.4656	0.1396	0.4664

*, ** and *** indicate significance at the 10%, 5% and 1% levels, respectively.

indicates a significant relationship between the degree of digital transformation at the decision-making level and government subsidies, suggesting that government subsidies can act as an intermediary in this pathway. When the decision-making level receives government subsidies, it is more motivated to formulate a series of strategies that favor digital transformation. However, the effectiveness of these strategies and their impact on business performance may require time to materialize.

Additionally, the degree of digital transformation at the execution level is significantly negatively correlated with government subsidies, while government subsidies are significantly positively correlated with business performance. Moreover, the degree of digital transformation at the execution level remains significantly negatively correlated with business performance. From this, it can be inferred that government subsidies play a partial intermediary role in this pathway.

Comparing Model (6) with Model (4), the significant inhibitory effect of the degree of digital transformation at the execution level on business digital transformation is somewhat mitigated after receiving financial support from the government. This also indirectly reflects the positive impact of government fiscal policies.

7. Conclusions and Implications

Based on data from A-share listed companies on the Shanghai and Shenzhen stock exchanges during 2007–2020, this study investigates the impact of digital transformation on performance and its mechanisms at various organizational levels. The key findings are as follows:

(1)

Digital transformation has contrasting effects on different organizational levels. Decision-making-level digitalization enhances company performance, while execution-level digital transformation diminishes performance drivers.

(2)

Organizational responses to digital transformation vary significantly across levels. The degree of digital transformation at different organizational levels influences company performance through the dimensions of technology, organization, and the external environment.

(3)

This study’s mechanism testing results reveal that digital transformation at various organizational levels influences company performance by increasing R&D expenditures and personnel inputs, promoting revenue growth, and facilitating the effective utilization of government subsidies for transformation strategies. Controlling decision-making-level management expenses effectively enhances company performance.

From a practical standpoint, this study provides insights for companies undergoing or planning digital transformation:

(1)

Prioritize relevant training to allow execution-level personnel to acquire digital skills, which are more effective than accumulating digital assets. Proficient employees can improve work efficiency and convey the importance of digital transformation to decision-makers.

(2)

Decision-makers should consider the broader context of their company, not just revenue growth, as the latter does not necessarily lead to improved company performance.

(3)

Assess both internal conditions and the level of acceptance and understanding of digital transformation within an organization when planning to implement digital transformation.

(4)

Carefully plan the digital transformation, ensuring alignment between technological development, organizational change, and macro-environmental changes to promote synergy and enhance overall operational efficiency and management capabilities.

Through our research, we found that enterprise digitalization may have opposite effects on different levels of an organization: the degree of digital transformation at the decision-making level can effectively promote the improvement of enterprise performance, while digital transformation at the executive level will weaken enterprise performance drive. The finding corresponding to the decision-making-level dimension of this study confirms existing research conclusions stressing that factors such as the personal characteristics of decision-makers and digital awareness can promote digital transformation and subsequently improve performance (Guo Runping et al., 2021; Tang Xuan et al., 2022) [30,31].

At the same time, we found that the existing literature has focused more on the specific digital technologies and a series of behaviors exhibited during the development and application process of the execution layer, for example, the inherent characteristics of systems such as business intelligence (Ana Marija Stjepi ‘et al., 2020) [32], blockchain technology (Saleem Malik et al., 2021) [33], Cloud ERP (Damali, U et al., 2021) [34], and the application of infrastructure such as information systems (Sven Vegard Buer et al., 2020; Tzu Chieh Lin et al., 2020) [10,35]. However, unlike existing research, this study found that for the execution level, digital transformation may not necessarily help organizations maintain sustainable growth (Wei Haiying, 2021) [36] or increase the demand for high-quality human and knowledge capital, promoting the improvement of capital structures (Sun Yuping et al., 2019) [37]. Instead, it may be met with resistance at the execution level, leading to a gradual increase in resource capabilities and related costs. This can cause a certain degree of systemic imbalance in the overall organization.

Furthermore, this study—distinct from existing research, in which single-dimensional analyses were primarily conducted—reveals significant variations in organizational responses to digital transformation across different hierarchical levels. In terms of transmission pathways, the extent of digital transformation at various organizational levels can impact organizational performance with respect to three dimensions: technological, organizational, and environmental. The mechanism test results indicate that digital transformation at different organizational levels influences organizational performance through increased investment in research and development (R&D) expenses and R&D personnel, promoting revenue growth, and facilitating the effective utilization of government subsidies to implement relevant transformation strategies. Controlling management expenses at the decision-making level can effectively enhance organizational performance, supporting and advancing digital transformation within an organization. Although we obtained the current results through a series of studies, we have not considered the influence of different levels of digital transformation on enterprise performance, nor did we consider under the circumstances in which the opposite influence of different levels will reach a balance. Thus, how can one judge which level of digital transformation has a greater impact on enterprise performance? Under what circumstances will the opposite effects of different levels reach a balance? Future research can discuss how to integrate internal and external resources and information in the process of digital transformation by analyzing the data of a specific enterprise, realizing efficient interaction and cooperation between different levels and promoting the optimization of resource allocation.