Digital Transformation and Enterprise Sustainability: The Moderating Role of Regional Virtual Agglomeration

Abstract

As digital transformation reshapes the world, businesses are becoming increasingly concerned about its impact. This study examines the regional data of 29 provinces and cities in China from 2013 to 2020 and uses the entropy method to calculate the regional virtual agglomeration level. This paper analyzes the panel data of listed companies in China to evaluate the economic impact of digital transformation on the sustainable development of enterprises and the moderating effect of the regional virtual agglomeration level. This study reveals a significant positive U-shaped relationship between digital transformation and enterprise sustainability. The level of regional virtual agglomeration can alleviate the initial negative impact of digital transformation on sustainability and enhance the economic growth capacity of enterprises. This paper found that the impact of digital transformation on sustainable development is particularly pronounced in the eastern region. All in all, based on the transaction cost theory and new geoeconomics, this paper analyzes the impact mechanism of digital transformation on the sustainable development of enterprises, verifies the relationship between the regional virtual agglomeration on digital transformation and sustainable development of enterprises, and identifies the adjustment mechanism. Our research provides theoretical support and practical guidance to promote digital transformation and enterprise sustainable development.

1. Introduction

The world is undergoing a digital transformation [1]. A new round of social change and enterprise innovation emerges as the times require. It is leading to a new round of growth and opportunity. China’s economic development must seize opportunities and unleash the full potential of digital transformation to promote enterprise sustainability and promote the conversion of kinetic energy between new and old. However, changes often contain multiple risks and challenges, and the obstacles faced by digital transformation should not be underestimated [2]. Reviewing the existing research shows that it is difficult to carry out change in enterprises [3], and up to 70% of large enterprises fail to change [4].

The role of digital transformation in enterprise transformation has gradually attracted the attention of the academic circle [5]. Digital transformation is the embodiment of innovation and technological progress that reshapes the business landscape. The emergence of network synergies is breaking the diseconomies of scale in traditional management [6], and platform economies are rapidly emerging [7], redefining enterprise boundaries [8] and dismantling sectional organizations [9]. The sharing and circulation of knowledge are also accelerating [10], and technologies are decomposing production factors and processes into smaller units for more efficient combination and distribution [11]. The use of platforms and network connections allows external resources to be realized [12], and new global social division of labor and collaboration organization models are taking shape. All of these changes profoundly impact our production and lifestyles. However, the above studies on digital transformation are too optimistic [13]. In reality, 87.5% of enterprises fail in the process of digital transformation [14]. Is there a new way to enable enterprises to effectively carry out digital transformation and improve their sustainability?

Virtual agglomeration has emerged as a new trend in the development of enterprises, providing a channel of “temporary proximity” between enterprises, expanding the knowledge spillover effect beyond geographical limitations, and leading to a closed network connection [15]. Nowadays, the production activities of enterprises cover global cyberspace [16], using data aggregation and data processing technology to achieve the optimal allocation of resources. In this new mode of agglomeration, virtual agglomeration, the allocation of resources no longer relies solely on geographical and physical space but on virtual cyberspace. In the traditional real economy, production resources are gathered and allocated through physical space; therefore, most transactions and production processes must have some process that can be handled, described and applied to the administrative or production system [17], forming a top-down or distributed production management mechanism. However, virtual agglomeration changes this. By contrast, with the development of information technology, any information, including non-standardized knowledge that used to be considered difficult to communicate or systematically represent, can now be encoded and decoded through multimedia technology, VR/AR and other new generation information technologies, thus enabling the transfer of information within virtual space and across the limits of physical space. This digitization of information greatly expands the scope of production and transactions, reduces search and transaction costs, improves the efficiency of matching goods and services between buyers and sellers [18], makes transactions a definitive process, and helps to build trust in transactions. The emergence of virtual agglomeration as a new trend of industrial organization in a wave of the digital economy provides a new paradigm for enterprises to better utilize resources, share knowledge, and optimize innovation strategies. It also provides a new path for enterprises to adapt to the digital economy era with the gradual popularization of information and communication technology and the rapid development of efficient and convenient information dissemination methods [19].

Enterprise sustainability refers to the process of integrating economic, social, and environmental concerns into the enterprise’s operations and decision-making [20]. This is conducted to ensure long-term viability, resilience, and competitiveness while contributing to the broader societal goals of sustainable development. In the context of digital transformation, sustainable development takes on new dimensions as businesses adapt to emerging technologies, changing consumer preferences, and evolving regulatory landscapes. Herden [21] proposed that enterprises must ensure that their digital transformation initiatives are developed in a responsible manner, taking into account the potential social and environmental impacts. Mondejar et al. [22] believed that digital security is a key link in the process of digital transformation to ensure enterprise sustainability. With the aim of promoting sustainable economic development and comprehensive societal progress through the enhancement of digital transformation, it is imperative to undertake collaborative endeavors among the government, enterprise entities, and academia.

It is crucial to consider the concepts of enterprise sustainability and virtual agglomeration together with technologies, as they are interrelated and create synergies with each other. Based on the environmental responsibility of enterprises, the application of digital technology could promote the development of more sustainable products and services, improve the visibility and traceability of the supply chain, monitor and manage environmental resources, improve the efficient utilization of energy and reduce the output of waste. This can help enterprises better reach a balance of production and ecological benefits [23]. As the scale of data becomes larger and more diverse, it is important to consider how to solve problems with data collection and processing, how to ensure that there are no quality and security loopholes in the data, and how to deliver the correct data at the correct time and in the correct way. We must prescribe the right remedy to the above problems that must be faced in the process of digital transformation. Only by solving these problems can we better promote the development of a digital economy. On this basis, sustainable economic development and overall social progress can be realized.

Scholars have previously conducted related research on digital transformation and corporate sustainable development. From the perspective of strategic management, Gerard [24] studied how multinational companies can achieve sustainable development through a digital transformation during the epidemic; El [25] used the method of SEM-PLS; to sum up, the elements that contribute to the sustainable development of enterprises in the process of digital transformation; Abdul [26] focused on digital transformation and corporate environmental sustainability and helped decision makers identify and prioritize investment areas. However, there are few empirical studies on the relationship between digital transformation and corporate sustainable development. Therefore, this article analyzes the impact of digital transformation on the sustainable growth of enterprises and the regulating effect of the virtual agglomeration level on the sustainable growth of enterprises to deepen our understanding of the new connotation of virtual agglomeration and broaden the new field of sustainable development of enterprises. Therefore, this research focuses on the regional virtual agglomeration level of 29 provinces and cities in China from 2013 to 2020, as well as the panel data of listed companies during the same period. This article attempts to address the following questions: Can digital transformation positively promote the sustainable development of enterprises? Is it possible that regional virtual agglomeration has a positive impact on the sustainable development of enterprises through digital transformation? What is its underlying mechanism? How can the agglomeration effect of virtual agglomeration be compared with traditional agglomeration?

2. Review of Literature and Hypotheses Development

2.1. Digital Transformation and Enterprise Sustainability

Sustainable development and digital transformation have fundamentally changed industrial society, yet research on them today is often conducted independently. As human beings pay more and more attention to sustainable development, quantitative research examining the relationship between various digital transformations and sustainable development strategies is emerging [27].

Digital transformation is an iterative process that exploits technology to augment economic growth and enhance efficiency by promoting information exchange and knowledge sharing. By utilizing digital-related technologies and relying on regional information technology infrastructure, businesses can fully harness information technology and develop and exploit information resources. By employing advanced technology, enterprises can streamline production processes, increase production efficiency, and reduce production costs. For instance, companies can employ artificial intelligence technologies to formulate production planning scientifically and enhance production efficiency [28]; the Internet of Things technology can be utilized to automate equipment control, decrease equipment failure rates, and improve the stability and efficiency of the production lines [29]; and big data analysis techniques can be applied to examine production data, identify bottlenecks and issues in the production process, and optimize the production process to boost efficiency [30].

Digital transformation can bring numerous benefits to enterprises, including the improvement of internal efficiency, enhancing competitiveness, promoting of inter-enterprise communication and cooperation, and strengthening the connection and interaction between various links in the industry chain [31]. By using information technology, enterprises can establish an information-sharing platform that facilitates data sharing and interoperability, enabling them to better understand the market demand and industry dynamics. This, in turn, strengthens information exchange and cooperation between enterprises.

Digital transformation also promotes innovation and cooperation among enterprises by allowing them to respond faster to changes in consumer needs, accelerating product development and promotion [32]. Additionally, enterprises can explore cooperation opportunities and find partners [33], thereby strengthening their connections and interactions between various links of the industry chain. Moreover, digital transformation can promote knowledge-sharing and technology transfer among enterprises by establishing knowledge-sharing platforms, promoting technology transfer and cooperation, and enabling joint problem-solving in the development of the industry. For example, enterprises can use blockchain technology to establish a digital cooperation platform to make transactions more transparent and traceable, thereby reducing information asymmetry, improving the security and credibility of transactions [34], and promoting innovation and cooperation among enterprises. Enterprises in the industry chain provide a mutually recognized trust mechanism [35]. Digital transformation has the potential to unlock significant values for enterprises by creating new business models [36], increasing efficiency and agility, and driving innovation and collaboration across the industry.

Sustainable progress is essential, and the concept of enterprise sustainability has garnered increasing attention in today’s business landscape. The triple bottom line approach proposed by Elkington [37], as well as the definition of corporate sustainability by Dyllick and Hockerts [38], emphasized the necessity for businesses to balance economic, environmental, and social impacts while addressing the needs of both present and future stakeholders. To achieve enterprise sustainability, companies must consider the resources they consume and the environmental impact they generate. They should endeavor to minimize their ecological footprint and pollution levels [39]. Business managers are also responsible for ensuring that their operations are conducted responsibly, and they should prioritize the health, safety, and quality of life of both the employees and external communities [40].

Enterprise sustainability necessitates that companies should consider the long-term impact of their operations and balance the needs of stakeholders both in the present and the future. By doing so, enterprises can ensure that their progress is sustainable and that they contribute to a more sustainable future for all. At the corporate strategy level, sustainability should be integrated into strategic management. As Baumgartner suggests, incorporating corporate sustainability into strategic management can offer competitive advantages by using ecological responsibility as a means to achieve cost reduction and market differentiation [41]. Times are changing, technology is progressing, and the business environment is rapidly evolving, driven by ecological and social trends that bring about more complex sustainability challenges [42].

A shift in values is essential if new technologies are to contribute more fully to sustainability as part of the new economy [43]. During the initial phases of digital transformation, due to organizational inertia [44], the lack of adaptability to new changes and challenges can easily lead to limitations and stagnation due to organizational inertia, i.e., a mindset and behavior pattern that has developed over time [45]. This can lead to limitations and stagnation. First of all, organizational inertia can lead to a lack of motivation for innovation and change [46]. Throughout the course of digital transformation, enterprises need to continuously innovate and change, and introduce new technologies and business models, but enterprise inertia often makes enterprises lack sensitivity to emerging technologies and business models and lack enthusiasm and initiative for change, which leads to slow progress in the time period when digital transformation moves from plans to practice. Second, organizational inertia also leads to resistance and hindrance in the process of digital transformation [47]. On the way to digital transformation, enterprises need to change their traditional organizational structure and processes and require extensive coordination and communication within the enterprise; however, corporate inertia frequently leads to a lack of knowledge and comprehension of digital transformation among the departments and individuals within the organization [48]. Similarly, enterprise inertia often makes enterprises lack their knowledge and understanding of digital transformation, which leads to hindrances, thus affecting the smooth implementation of digital transformation. Moreover, enterprises in digital transformation also need to invest in a lot of resources, including financial, human and technical aspects [49]. This may lead to short-term negative impacts such as increased costs and decreased efficiency. However, with the continuous promotion of digital transformation, enterprises can gradually remove the inertia of the past, optimize business processes and management methods, improve product and service quality, and enhance market competitiveness, thus bringing positive effects of sustainable development. Consequently, grounded in the aforementioned analysis, the subsequent hypotheses were put forward:

Hypothesis 1 (H1): 

Digital transformation has a U-shaped relational impact on enterprise sustainability.

2.2. The Moderating Role of Virtual Agglomeration

Marshall first identified agglomeration effects as a regional development strategy. He elucidated the benefits of spatial proximity for firms engaged in a collaborative economic framework that involved the procurement and vending of labor, commodities, and amenities [50]. However, the congestion effect and rent-seeking behavior of the traditional spatial agglomeration output often have a negative impact on firm productivity [51]. Driven by technological advances and globalization, the distance between people has been infinitely reduced [52]. The concept of “death of distance” has been gradually accepted [53]. As a new way of cooperation, virtual agglomeration can be gradually used in the sustainable development of enterprises.

The notion of virtual agglomeration emerged in the late 1990s and early 2000s, driven by the growth of e-commerce and digital technologies. It was first introduced in 1997 by a research group of seven universities supported by the EU’s SACFA program. According to the group, a Virtual Industrial Cluster (VIC) is a collection of firms with specific strengths that are able to align their essential strengths and involvement in the virtual enterprise process, thereby sharing market opportunities [54]. Virtual agglomeration refers to economic activities clustering in the digital rather than physical space, including online marketplaces, virtual communities, and digital ecosystems. These digital spaces often have dense networks of economic players, such as firms, consumers, and intermediaries. Digital trunking is an inter-organizational network composed of Internet business communities [55]. The digitization of cluster operations means that traditional, geographically restricted cluster activities can be transformed into digital processes, enabling businesses to utilize digital tools and platforms for collaboration, communication and resource sharing. This approach can greatly enhance the competitive advantage of clusters so that the concept of traditional agglomeration can be extended to include virtual agglomeration. Virtual agglomeration is increasingly important in the global economy as companies use digital technologies to reduce transaction costs, increase efficiency, and access new markets [56]. The digitization of virtual clusters and cluster operations is a component or manifestation of virtual agglomeration. Virtual agglomeration encompasses the broader digital landscape, where various forms of virtual clusters and digital operations come together to promote innovation, growth and competitiveness in the global economy. Due to its potential for enabling firms to access new sources of knowledge and innovation, reducing supply chain costs, breaking the protectionism of traditional clusters, and creating new business models, virtual agglomeration is often regarded as an important catalyst for economic expansion and competitiveness improvement. The development process of virtual agglomeration is shown more clearly in

Table 1. Related literature combing.

Author(s)	Related Overview
Stephen Graham (1998) [57]	Proposed that the virtual environment has a unified spatial location, which can effectively allow the remote transmission of geographic location and provide users with an immersive experience.
Michael and Edward (2001) [58]	Highlighted how the internet might allow for virtual agglomerations by reducing the costs of communication and enabling knowledge spillovers among geographically dispersed entities.
Kvasny and Keil (2006) [59]	Highlighted the role of digital infrastructure in reshaping economic geography and suggested that virtual agglomerations may emerge as a result of increased access to digital technologies and resources.
Kolko (2007) [60]	Highlighted the role of virtual agglomerations in promoting co-location and knowledge spillovers among service firms, enabled by ICT and digital infrastructure.
Wasko et al. (2011) [61]	Undertook a comparison of the virtual agglomeration of economic activities in the apparel industry with traditional industrial agglomeration. Regarded it as an important strategy to ensure regional and national competitiveness in today’s volatile economic environment.
Tian and Han (2021) [54]	Sorted out the connotations, characteristics, modes, incentives, influences and specific applications of virtual agglomeration in detail and evaluated them, respectively, while summarizing the academic differences and future expansion space of virtual agglomeration.

.

According to transaction cost theory, clustering in digital space can reduce transaction costs between economic participants by making it easier and more efficient to share information [62]. By reducing transaction costs, virtual agglomeration can enable firms to access new markets, reduce supply chain costs, and increase efficiency. For instance, online marketplaces can help companies connect more easily with suppliers and customers, thereby reducing the need for expensive middlemen. In addition, digital transformation can help companies collect, process, and analyze information more efficiently, reducing the costs associated with finding and evaluating potential trading partners. In areas with high levels of virtual agglomeration, these cost reductions are likely to be more pronounced due to the increased availability of information and the ease of sharing resources through digital channels [63].

New economic geography shows that transportation costs are the main force of an industrial layout [64]. From the vantage point of the C–P (central–peripheral) model, enterprises have the propensity to congregate in core zones with diminished transportation expenditures and heightened attainability. In contrast to the negative externalities generated by spatial–geographical interactions between enterprises, such as pollution, traffic congestion, and high housing prices [65], clusters in digital space can create more positive externalities, such as knowledge spillovers and economies of scale [66]. If an innovative enhancement implemented by an enterprise enhances the achievements of other firms with no cost to the beneficiary firm, these externalities or spillovers can be “(fully) remunerated [67]”. The moderating effect of regional virtual agglomeration may enhance these network externalities by increasing the density of digital connections and interactions, leading to stronger and more valuable digital ecosystems.

This paper argues that companies during the case of digital transformation are moderated by their level of regional virtual agglomeration, which can weaken the impact of negative factors generated in the trial stage of digital transformation and release the dividends of digital transformation for sustainable business development in advance. Therefore, this paper proposes the following hypothesis.

Hypothesis 2 (H2): 

The level of regional virtual agglomeration positively moderates the U-shaped relationship between digital transformation and sustainable enterprise development.

The research framework is shown in Figure 1.

3. Model Design

3.1. Measurement of Variables

Enterprise sustainability: It refers to a company’s ability to achieve long-term profitability while maintaining enduring competitiveness and prioritizing responsible practices. Therefore, this paper constructed the enterprise sustainability index to measure enterprise sustainability according to the Van Horn sustainability static model. The calculation formula is shown below:

$$\mathrm{Enterprise} \mathrm{sustainability}=\frac{\mathrm{Net} \mathrm{profit}\ast \mathrm{Revenue} \mathrm{retention}\ast (1+\mathrm{Equity} \mathrm{ratio})}{(\frac{1}{\mathrm{Total} \mathrm{asset} \mathrm{turnover}}-\mathrm{Net} \mathrm{sales} \mathrm{profit}\ast \mathrm{Revenue} \mathrm{retention}\ast (1+\mathrm{Equity} \mathrm{ratio}))}$$

This paper adopted the method of Wu, F. et al. [68] to measure digital transformation. To be precise, we gathered the frequency of digital transformation-related keywords used in the annual reports of the listed companies in our sample, aggregated them, and subsequently applied the natural logarithm to formulate digital transformation indices.

Regional virtual agglomeration level: This paper selected the regional digital economy index as a proxy variable for the regional virtual agglomeration level, drawing on Liu, J. et al. [69]. This evaluation index system of the digital economy constructs a comprehensive index evaluation model that can measure the comprehensive evaluation index of the regional virtual agglomeration level through the following steps: standardize the original data; apply the entropy value method to measure the weights of each index in the evaluation system; and adopt the multi-objective linear weighting function method to weight each index to calculate the comprehensive index for each region and criterion level index using the following formula. The weight distribution of the specific indicators is shown in

Table 2. Evaluation index system of digital economy.

Level 1 Dimension	Secondary Dimension	Specific Indicators	Weights
Informatization Development	Informatization Foundation	The quantity of fiber optic cables per unit area	0.0619
		Density of mobile phone base stations	0.0693
		Proportion of information technology professionals	0.0376
	Influence of Informatization	Aggregate Telecommunications Enterprise	0.1024
		Revenue from software enterprise	0.1596
Internet Development	Fixed End Internet Foundation	Density of internet access points	0.0733
	Mobile Internet Foundation	Rate of mobile internet adoption	0.0195
	Impact of Fixed Internet Connectivity	Proportion of broadband internet subscribers	0.0456
	The Impact of the Mobile Internet	Proportion of mobile internet subscribers	0.0217
Digital Trading Development	Digital Trading Basics	Density of websites per 100 enterprises	0.0073
		Utilization of computers by enterprise	0.0327
		Proportion of enterprises engaged in e-commerce	0.0580
	Digital Trading Impact	E-commerce sales	0.1304
		Online retail sales	0.1806

. (1) Data standardization. For the positive and negative indicators in the comprehensive evaluation system, this paper adopted the extreme value method for standardization. The calculation formula is as follows:

Positive indicators:

$$Y_{ij}=\frac{(X_{ij}-X_{i,min})}{(X_{i,max}-X_{i,min})}$$

(1)

Negative indicators:

$$Y_{ij}=\frac{(X_{i,max}-X_{ij})}{(X_{i,max}-X_{i,min})}$$

(2)

In the formula, $X_{ij}$ represents the value of the j-th evaluation index in the i-th region, $X_{i,min}$ and $X_{i,max}$ are the minimum and maximum values of the i-th region in the original data. The higher the value, the higher the level of the indicator. $Y_{ij}$ is the proportion of the j-th index in the i-th region. This formula can be compared each year.

(2) The entropy value method is an objective weighting method. The principle of weighting is to calculate the weight according to the coefficient of variation in the index. If the coefficient of variation is larger, the calculated weight is larger, and vice versa. Therefore, the entropy value method was used to determine the weight in this paper.

The weight of the j-th indicator for the i-th region, denoted by m, was determined by the number of regions:

$$P_{ij}=\frac{Y_{ij}}{{\sum }_{i=1}^m{Y}_{ij}}$$

(3)

Entropy value of the j-th indicator:

$$E_j=\raisebox{1ex}{$-1$}\!\left/ \!\raisebox{-1ex}{$\ln \left(m\right)$}\right.\sum _{i=1}^m{P}_{ij}\ln \left(P_{ij}\right)$$

(4)

Coefficient of variation in the j-th indicator:

$$H_j=1-E_j$$

(5)

Weight of the j-th indicator:

$$W_j=\raisebox{1ex}{$H_j$}\!\left/ \!\raisebox{-1ex}{${\sum }_{j=1}^m{H}_j$}\right.$$

(6)

(3) The composite index-linear weighting function approach for multiple objectives.

The indicator scores for the s-th criterion stratum in the i-th region were determined based on the q total number of indicators included in that particular indicator stratum:

$$Z_{is}=\sum _{i=1}^q{W}_j{Y}_{ij}$$

(7)

The composite evaluation index of the i-th region was determined with n representing the number of indicator layers:

$$F_i=\sum _{s=1}^n{Z}_{is}$$

(8)

Control variables: In this paper, six indicators, including enterprise growth, operating efficiency, enterprise risk, profitability, operational efficiency and investment profitability, were selected as control variables, and the existence of one control in many enterprises was taken as a dummy variable. Its specific index measurement can be seen in

Table 3. Variable definition.

Variable Symbols	Variable Name	Variable Definition and Description
SUS	Enterprise Sustainability	(Net income/total owner’s equity ending balance)
		×[1-Pre-tax dividend per share/(net income for the period/
		Paid-in capital current period-end value)]/(1-numerator)
DA	Degree of Digital Transformation	Regarding the emergence of AI and blockchain technology in enterprise, reporting about
		Cloud computing technology; Big data technology; and Digital technology applications.
		After frequency normalization, the total index was processed by adding 1 to the logarithm
DE	Virtual Agglomeration Level	The regional digital economy index was selected as a proxy variable
YS	Business Growth	(Total operating income in the current year)/current period amounts to the total operating income in the previous year
		(Amount for the same period)/(Amount of total operating income for the same period of the previous year)
JLL	Operating Efficiency	(Net income for this year-Prior year’s net income)/Prior year’s net income × 100%
GG	Enterprise Risk	Financial leverage × Operating leverage
ROA	Profitability	Net income/total assets balance
YYLL	Operational Efficiency	Operating profit/operating income
TZLL	Investment Profitability	Total annual profit or average annual profit/total project investment × 100%
OC	Whether there is a control of multiple	0: No, 1: Yes. Whether there is the same beneficial owner controlling multiple
		The status of listed companies

.

3.2. Model Construction

In order to examine the influence of digital transformation on enterprise sustainability, the subsequent model was developed.

$$SU{S}_{it}={\alpha }_0+{\beta }_{1}D{A}_{it}+{\beta }_{2}Controls+{\sum }^{ }Year+{\epsilon }_{it}$$

(9)

$$SU{S}_{it}={\alpha }_0+{\beta }_{1}D{A}_{it}+{\beta }_{3}D{A}^2{}_{it}+{\beta }_{2}Controls+{\sum }^{ }Year+{\epsilon }_{it}$$

(10)

$$SU{S}_{it}={\alpha }_0+{\beta }_{1}D{A}_{it}+{\gamma }_{1}D{A}_{it}\times D{E}_{it}+{\beta }_{3}D{A}^2{}_{it}+{\gamma }_{2}D{A}^2{}_{it}\times D{E}_{it}+{\gamma }_{3}D{E}_{it}+{\beta }_{2}Controls+{\sum }^{ }Year+{\epsilon }_{it}$$

(11)

In Models (9)–(11), i represent the listed company, and t signifies the year. Controls represents the control variable, and ε is the residual term. The interaction terms of the digital transformation and regional virtual agglomeration level are represented by $D{A}_{it}$ × $D{E}_{it}$ and $D{A}^2{}_{it}$ × $D{E}_{it}$. Model (9) was used to examine the influence of digital transformation on the sustainable growth of businesses, and Model (10) included the quadratic term for digital transformation. If the regression coefficient ${\beta }_3$ was significantly positive, it signified the existence of a notable U-shaped correlation between the two. In Model (11), we added the interaction term of the regional virtual agglomeration level and digital transformation. If ${\beta }_1$, ${\gamma }_1$, and ${\gamma }_2$ are significant, it suggests the combination of two functions [70]. The U-shaped association between digital transformation and sustainable enterprise development can be delineated by devising two prospective functions, denoted as function A and function B.

Function A represents a negative correlation curve relationship between digital transformation and enterprise sustainable development. It suggests that in the early stages of digital transformation, enterprises may lack reasonable digital thinking and effective strategic foresight, leading to the misconception of “emphasizing technology introduction but neglecting professional digital talent training”. This misconception requires a significant number of resources for the introduction of new technologies and may face resistance from existing interest patterns, resulting in a “top-heavy” pathological development mode. Furthermore, while companies find themselves in the early phase of digital transformation, they may be unable to integrate data as a key element into various businesses and effectively use data to help them operate, causing a decline in their sustainable development capability.

Function B, on the other hand, indicates that, under unchanged conditions, as digital transformation achieves certain results, the positive impetus brought by digitalization to enterprise businesses can fully activate the revenue potential of key business scenarios. This activation can help enterprises break the barriers of institutions and mechanisms, further build digital business processes and data platforms, and eventually achieve three-dimensional and comprehensive digital intelligence. This, in turn, can inject new momentum into the sustainable development of enterprises, allowing them to continuously improve their sustainable development capability.

3.3. Sample Source

In this manuscript, this paper takes Shanghai and Shenzhen a-share listed companies as research samples for regression analysis, encompassing a time frame spanning from 2013 to 2020. In order to guarantee the precision and dependability of the research, we applied the following criteria when screening the sample data. First, companies with ST/ST* symbols, which often face abnormal financial conditions and serious financial risks, were removed from the sample. The inclusion of such companies could lead to a significant deviation between the experimental results and the actual situation, resulting in counterfactual phenomena. Next, we eliminated most of the samples containing incomplete information and outliers to maintain data accuracy. Finally, we removed the sample of companies in the financial sector. Following these treatments, we obtained 5960 valid sample data from 889 companies in the China Stock Market & Accounting Research Database (CSMAR).

In order to measure the relevant indicators of the level of regional virtual agglomeration, this paper selected measurement benchmark data from the 2013–2020 China Statistical Yearbook and the National Bureau of Statistics. In order to reflect the role of digital transformation as accurately as possible, this paper excluded some provinces where digitization was in its infancy.

4. Empirical Results and Analyses

4.1. Descriptive Statistics and Correlation Analysis

Descriptive statistics provide a concise summary of the main characteristics of a dataset, including the measurement of centralized trends and dispersion. These statistics help researchers understand the general characteristics of the data, which is essential for subsequent analysis and interpretation. The dataset presented in

Table 4. Descriptive statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
SUS	5960	0.605	0.468	−0.275	2.306
DA	5960	3.444	0.208	3.255	4.429
DE	5960	0.31	0.148	0.073	0.768
YS	5960	0.135	0.263	−0.408	1.393
JLL	5960	0.632	3.777	−5.31	28.133
GG	5960	2.357	2.565	0.941	18.894
ROA	5960	0.066	0.042	0.004	0.223
YYLL	5960	0.12	0.113	−0.019	0.628
TZLL	5960	0.623	3.025	−1.82	25.478
OC	5960	0.393	0.488	0	1

contains 5960 observations and 10 variables. The summary statistics provided include the mean, standard deviation and minimum and maximum values for each variable. The average value of SUS was 0.605, the standard deviation was 0.468, and the range was from −0.275 to 2.306, indicating that there were large differences in the sustainable development capabilities of various enterprises. The mean value of DA was 3.444, the standard deviation was relatively small at 0.208, and the range of DA was narrow, with values between 3.255 and 4.429. This means that digital transformation has become the trend of enterprise development. The mean value of DE was 0.31, the standard deviation was 0.148, and the minimum and maximum values were 0.073 and 0.768, respectively. This reflects that there are certain differences in the level of virtual agglomeration among regions and highlights the scientific nature of setting the level of regional virtual agglomeration as an adjustment variable.

Table 5. Correlation analysis and VIF value.

	SUS	DA	DE	YS	JLL	GG	ROA	YYLL	TZLL	OC	VIF	1/VIF
DA	0.061 ***	1
DE	0.122 ***	0.027 **	1
YS	0.189 ***	0.061 ***	0.001	1							1.04	0.964207
JLL	0.048 ***	0.001	0.025 *	0.052 ***	1						1.02	0.984325
GG	−0.399 ***	0.053 ***	−0.084 ***	−0.098 ***	−0.040 ***	1					1.2	0.835375
ROA	0.834 ***	0	0.007	0.136 ***	−0.027 **	−0.349 ***	1				1.53	0.654232
YYLL	0.453 ***	0.067 ***	0.057 ***	0.025 *	0.042 ***	−0.339 ***	0.545 ***	1			1.5	0.66559
TZLL	0.009	0.028 **	0	−0.030 **	−0.003	−0.025 *	0.031 **	0.025 *	1		1.01	0.993765
OC	−0.027 **	0.01	0.016	−0.059 ***	0.072 ***	0.075 ***	−0.088 ***	−0.028 **	−0.058 ***	1	1.02	0.978371

Note: N = 5960; * p < 0.1, ** p < 0.05, *** p < 0.01.

exhibits the outcomes of the Pearson correlation analysis among the primary variables. This analysis uncovered a significant and positive correlation between digital transformation and enterprise sustainability at the 1% level, suggesting that a higher degree of digital transformation was advantageous for the sustainable growth of companies. Nonetheless, further investigation was necessary to comprehend the precise nature of this relationship. It is worth noting that the correlation coefficients among the variables were all below 0.51, indicating that there was no significant issue of multicollinearity among the variables.

The variance inflation factor (VIF) inspection is a diagnostic tool that is used to assess the severity of multicollinearity in multiple regression analysis. Multicollinearity occurs when independent variables in a regression model are highly correlated, leading to unreliable and unstable estimates of regression coefficients. Therefore, we conducted a variance inflation factor (VIF) test among variables, which confirmed that there was no noteworthy problem of multicollinearity. In this paper, the maximum value of each variable VIF was 1.53, and the minimum value was 1.01, both of which are less than five, indicating that multicollinearity was not obvious.

4.2. Hypothesis Testing

Direct effect test:

Table 6. Regression model.

Variables	(1) SUS	(2) SUS	(3) SUS	(4) SUS
DA	0.0115 ***	0.0035	−0.1805 ***	0.3424 ***
	(0.0015)	(0.0036)	(0.0304)	(0.0767)
DE		−0.0321		2.3951 ***
		(0.0327)		(0.3694)
Sq.DA			0.0257 ***	−1.3287 ***
			(0.0041)	(0.2030)
DA×DE		0.0188 **		−0.0465 ***
		(0.0095)		(0.0104)
Sq.DA × DE				0.1847 ***
				(0.0277)
YS	0.0116 ***	0.0064 ***	0.0053 ***	0.0058 ***
	(0.0012)	(0.0010)	(0.0010)	(0.0010)
JLL	0.0007 ***	0.0004 ***	0.0004 ***	0.0004 ***
	(0.0001)	(0.0001)	(0.0001)	(0.0001)
GG	−0.0022 ***	−0.0017 ***	−0.0017 ***	−0.0016 ***
	(0.0001)	(0.0001)	(0.0001)	(0.0001)
ROA	0.9021 ***	0.9654 ***	0.9688 ***	0.9691 ***
	(0.0094)	(0.0109)	(0.0109)	(0.0108)
YYLL	−0.0101 **	0.0153 ***	0.0185 ***	0.0145 **
	(0.0034)	(0.0046)	(0.0047)	(0.0046)
TZLL	−0.0002 *	−0.0001	−0.0001	−0.0001
	(0.0001)	(0.0001)	(0.0001)	(0.0001)
1.OC	0.0049 ***	0.0026 **	0.0026 **	0.3424 ***
	(0.0007)	(0.0009)	(0.0009)	(0.0767)
Stked	YES	YES	YES	YES
Year	YES	YES	YES	YES
_cons	−0.0365 ***	−0.0223 *	0.3120 ***	−0.0347 ***
	(0.0054)	(0.0125)	(0.0558)	−0.0082
N	5960	5960	5960	5960
R2	0.742	0.738	0.728	0.74

Note: * p < 0.1, ** p < 0.05, *** p < 0.01.

displays the outcomes of the influence of digital transformation on enterprise sustainability. Column (1) shows that after including the control variables, digital transformation was significantly and positively related to enterprise sustainability at the 1% level. Column (2) tests the effect of the regional dummy level on the linear relationship between the two. Column (3) introduces the square term of digital transformation into the experimental model to examine its impact on firm innovation. The results show a positive U-shaped relationship at the 1% level, indicating that at the beginning of digital transformation, enterprises faced internal and external resistance and needed to invest by introducing technology and training talents, leading to a weakening of their sustainable development ability, and a downward trend in the curve. However, as digital transformation gained results and was accepted by enterprises, it injected new momentum into their sustainable development, leading to an inflection point in the curve and an upward trend. Hypothesis H1 was thus proved. Column (4) uses the regional virtual agglomeration level as a moderating variable to interact with the square term of digital transformation. The outcomes still evinced a constructive U-shaped association at the 1% level, with the curve’s point of inflection shifting toward the left. This denotes that the virtual agglomeration level of regions could effectively and positively moderate the effect of digital transformation on sustainable development and alleviate the adverse elements that firms may encounter in their preliminary phase of digitalization. H2 holds.

4.3. Robustness Tests

The explanatory variables were taken as the next period and expressed as L.DA and L.Sq.DA, representing the linear and curvilinear relationship of digital transformation. The regression shows that the significance of the explanatory variable “enterprise sustainability” was generally consistent with the above, indicating that the above findings are robust (see column (1) of

Table 7. Robustness test.

Variables	(1) SUS	(2) TPF
L.DA	−2.1345 ***	2.6274 **
	(−0.4597)	(−1.1987)
L.Sq.DA	0.2980 ***	−0.3343 **
	(−0.0629)	(−0.1642)
Controls	Yes	Yes
Years	Yes	Yes
N	4716	4716
R2	0.738	0.265

Note: ** p < 0.05, *** p < 0.01.

).

Because the production method has not been scientifically regulated [71], enterprises face challenges in ensuring that economic growth is accompanied by a positive impact on environmental protection. Additionally, various abrupt natural calamities and public health emergencies have arisen frequently in recent years, which has impeded the seamless execution of sustainable enterprise development [72]. Therefore, it is very important for enterprises to strengthen the sustainable development of environmental protection. The existing literature has incorporated carbon emissions into sustainable development studies and provided practical examples [73]. As a key indicator, enterprise carbon emissions can effectively measure the sustainable development capabilities of companies in response to these global challenges. In the robustness examination of this study, the independent variable SUS was substituted with the natural logarithm of enterprise carbon emissions augmented by one (TPF), and the outcomes are presented in column (2) of Table 7. The inverted U-shaped configuration between digital transformation and enterprise carbon emissions intimates that as digitalization has advanced, enterprise carbon emissions initially intensified before subsequently diminishing, which is advantageous in furthering enterprise sustainable development. These findings validate the resilience of the regression discoveries in this manuscript.

The COVID-19 pandemic in 2020 has had a major impact on the sustainable development of enterprises; therefore, this paper conducted a robustness test again by excluding the sample data in 2020. The results (

Table 8. Deleted 2020 sample data.

Variables	SUS	SUS	SUS	SUS
DA	0.0987 ***	−0.0008	−1.1894 **	2.0768 **
	(0.0158)	(−0.0392)	(−0.3621)	(0.9126)
DE		−0.0321		17.9615 ***
		(−0.0327)		(4.5117)
Sq.DA			0.1744 ***	−0.2838 **
			(0.049)	(−0.1234)
DA×DE		0.3121 **		−9.9819 ***
		(0.109)		(−2.4672)
Sq.DA×DE				1.4037 ***
				(0.3348)
Controls	Yes	Yes	Yes	Yes
Years	Yes	Yes	Yes	Yes
N	5318	5318	5318	5318
R2	0.722	0.731	0.724	0.734

Note: ** p < 0.05, *** p < 0.01.

) showed that there was still a significant U-shaped relationship between digital transformation and enterprise sustainable development, and the level of regional virtual agglomeration still played a significant positive regulatory role, indicating that the hypothesis of this research conclusion is still valid.

4.4. Heterogeneity Analysis

From the data presented in different series in

Table 9. Heterogeneity analysis of area.

	East	Middle	West
Variables	SUS	SUS	SUS
DA	0.1767	0.0256	0.6100 *
	(0.1168)	(0.3267)	(0.3534)
DE	1.8510 ***	−0.0354	3.6824
	(0.4942)	(2.6764)	(2.6482)
DA×DE	−1.0414 ***	0.0244	−1.9768
	(0.2712)	(1.4633)	(1.4497)
Sq.DA	−0.0253	−0.0020	−0.0798 *
	(0.0158)	(0.0443)	(0.0479)
Sq.DA×DE	0.1482 ***	−0.0045	0.2626
	(0.0369)	(0.1986)	(0.1971)
_cons	−0.3187	−0.0645	−1.1505 *
	(0.2146)	(0.5995)	(0.6481)
Controls	Yes	Yes	Yes
Years	Yes	Yes	Yes
Firms	Yes	Yes	Yes
N	4216	948	796
R2	0.747	0.714	0.750

Note: * p < 0.1, *** p < 0.01.

, it is discernible that there exist disparities in the impacts of digital transformation on sustainable enterprise development at the regional level, with marked regional heterogeneity. In particular, the effect of digital transformation on sustainable business growth in the eastern region was significantly positive, signifying that digitalization in the east exerted a stimulating influence on augmenting enterprise sustainability, as the country was vigorously promoting digital China, smart societies and digital economy, using policy as a guide to achieving long-term goals. As a form of planning, it provided ideological guidance and policy support for enterprises in the region to carry out the “protracted war” of digital transformation, which notably mitigated the obstruction caused by adverse elements encountered by firms in the preliminary phase of digitalization. The eastern region’s arithmetic coordination gives full play to the leading role of the original core enterprises and built knowledge network, thus realizing the knowledge spillover effect, forming a homogeneous group effect, continuously attracting enterprises to digital transformation, and absorbing new members to join the network. Enterprises changed from originally geographical-oriented to collaborative innovation, which was demand-oriented for cooperation, leading to a reduction in enterprises in the preliminary phase of digital transformation due to information asymmetry and production. This reduced the “misjudgment” of enterprises during the initial stage of digital transformation due to information asymmetry, strongly strengthening enterprises’ ability to grasp opportunities and providing full play to the positive regulation effect generated by regional virtual agglomeration. With the advantages of good infrastructure resources, a complete industrial chain, high-quality education resources and a high marketization level, the eastern region provides more room for innovation and vitality for enterprises’ digital development and has a large reserve of related talents, thus advancing the sustainable development of enterprises.

The impact of digital transformation on the sustainability of enterprises in the central region was not significant, and the promotion effect of regional virtual agglomeration was also very little. Specifically, because the central region is mainly dominated by heavy industry, traditional industry accounts for a relatively high proportion, resulting in relatively high costs for digital transformation. Coupled with multiple historical development bottlenecks, digital transformation faces greater resistance, so the transformation process is relatively slow, and it is difficult to fully demonstrate its positive effects. In addition, the central regions did not pay enough attention to the top-level design and overall layout of the areas under their jurisdiction when formulating their respective digital economy development plans, and the gathering of enterprises lacked an internal and effective attraction mechanism. There was no effective mechanism to attract enterprises. At the same time, according to the theory of the “gradient transfer” of industries, with the rising prices of production factors, energy and resource constraints, and serious environmental problems in the eastern coastal region, labor-intensive industries in the east are increasingly shifting to capital-intensive and technology-intensive high-tech industries, and the central region is taking over the labor-intensive industries transferred from the east. The labor-intensive sectors in the central region assume the labor-intensive industries that have relocated from the eastern region, further strengthening the social environment, investment system environment and business environment constraints in the central region. Transferred industries also face the problem of unconformity, as it is difficult to form an efficient agglomeration. Although some development zones have realized the spatial aggregation of enterprises, the depth of cooperation among enterprises is lacking, which leads to the fragility of industrial aggregation relying on soft policy conditions, and the technology and industrial resources quickly gathered are transferred out due to the policy preferences in other regions, which makes the economic development of aggregation less stable and the regulating effect not as obvious.

Although the level of regional virtual agglomeration has not played a significant moderating role in the western region due to infrastructure, mindset and other factors, a significant relationship was shown between digital transformation and enterprise sustainability. This indicates that digital transformation in the west has initially formed a scale effect and started to advance the sustainable development of enterprises. In recent years, the governments of western regions have incorporated digital transformation into their assessment projects to leverage the industrial advantages of western original agricultural products. Since the overall economy of the western region is relatively weak, this digital foundation is relatively backward, and the influence of the digital transformation trend is short; therefore, the digital divide is relatively small [19]. This implies that during the course of digitalization, western firms could expedite the integration of technology into their production and management systems, thereby attaining digital transformation at a more rapid pace. To provide full freedom of play for business development means that digitalization can be developed according to their respective needs, including digital agriculture, the digital countryside of production, supply and the marketing integration of investment, bigger and stronger local agriculture, planting, breeding, agricultural processing industry, logistics, cold chain, transportation and e-commerce sales platform, live with goods platform. On the other hand, because of the comparatively slower economic progress in the western region, there is relatively low market competition and limited interaction and collaboration between enterprises. In addition, in the more remote geographical location of the western region, the logistics costs are relatively high, resulting in cooperation and interaction between enterprises being subject to certain constraints. As a result, they have not established a good relationship of trust between each other, meaning the degree of regional virtual agglomeration would play a relatively weak regulatory role.

5. Discussion

Enterprise sustainability is a concept that aims to balance economic, social and environmental factors. From an economic perspective, enterprise sustainability involves the effective utilization of resources and the promotion of economic growth without sacrificing environmental and social well-being. In this context, digital transformation can perform a crucial function in promoting sustainable economic growth for enterprises.

Improving operational efficiency and reducing costs is a way for digital transformation to promote the sustainable economic growth of enterprises. Through the utilization of digital technologies such as automation, AI and data analytics, companies can optimize processes and reduce waste, resulting in cost savings and increased productivity. In turn, this can boost economic growth while reducing the environmental impact of business operations. Digital transformation can also facilitate the development of new business models that can promote sustainable economic growth. For example, the sharing economy supported by digital platforms allows individuals and businesses to share assets such as vehicles and accommodation, thereby using resources more efficiently. This can reduce waste and promote sustainability by reducing the need for new resource-intensive production. In addition, digital transformation can promote sustainable economic growth by facilitating the development of renewable energy sources. Digital technologies can manage renewable energy systems more efficiently and make it easier to integrate renewable energy into the economy. This can promote economic growth while reducing dependence on nonrenewable resources, resulting in a more sustainable economic paradigm. The empirical analysis in this paper confirms a linear relationship between digital transformation and enterprise sustainability, which is consistent with the findings of George [42], Ren [74], and others. However, an analysis of the Chinese context reveals a U-shaped curve relationship between digital transformation and enterprise sustainability, which validates Li’s [75] research and provides indirect empirical support for Li’s [76] argument that digital transformation is a long and uncertain process that requires patience to achieve positive results. Based on reality, enterprises use machine learning algorithms and big data analysis to help improve resource allocation, optimize processes and predict potential problems. This results in smarter decisions, less waste and sustainable growth. The research results of this paper can provide better theoretical support for this phenomenon. Since this paper has used enterprise carbon emissions as a measure of enterprise sustainability [77], a proxy variable for robustness testing, the concept of circular economy can be elicited for further discussion. One important way in which digital transformation supports sustainable development is by promoting the adoption of circular economy principles. A circular economy is an economic system in which resources are used and reused for as long as possible, waste is minimized, and resources are recovered at the end of their useful life. Technologies such as blockchain can more effectively monitor and manage resources and simplify the process for companies to implement circular economy principles, which in turn, can encourage the adoption of renewable energy and promote environmental sustainability within companies.

Virtual agglomeration may exert a significant impact on the correlation between digitalization and sustainable business growth. However, current research has mainly focused on qualitative studies of virtual agglomeration, with relatively few quantitative studies conducted [78,79]. According to the empirical findings of this paper, the correlation coefficient between digital transformation and sustainable business growth is notably strengthened upon the inclusion of regional virtual agglomeration as a moderating variable. This finding indicates that the level of regional virtual agglomeration could serve as a significant “enhancer”. It can provide theoretical support for the research of digital twinning [80], edge computing [81], decentralized finance [82] and other related aspects.

Specifically, the level of regional virtual agglomeration, as an environmental factor, enables enterprises to achieve more cooperation and knowledge sharing around sustainable practices and promote healthy competition among enterprises. This can drive companies to adopt best practices in digital transformation and sustainable development, resulting in better overall performance, echoing the findings of Kang [83] for the strengthening of empirical research on virtual agglomeration studies. However, the current academic community also maintains a high level of concern about the potential threats that virtual agglomeration may pose. From the perspective of enterprises themselves, nowadays, how to ensure the security of enterprise-sensitive data in the process of cooperation has gradually become the focus of attention within enterprises [84]. Their increased reliance on digital technology and data sharing within virtual agglomerations may raise privacy and security concerns for business managers [85]. On the other hand, this paper proposes a future angle for research from a cultural perspective: based on the social identity theory [86], when individuals improve their capabilities by imitating better individuals, they may develop stronger identification with the better group of individuals, which may lead to a tendency to support their group (better individuals) and discriminate against those outside the group (original individuals). Virtual agglomeration gives companies the opportunity to reach out across geographic space and engage with companies worldwide (better individuals) through supply and demand, which may place global or regional trends over local needs and values (primitive individuals), leading to the erosion of local identities and cultures to the detriment of meeting the company’s obligations for sustainable development at a societal level. Therefore, it is important to ensure that virtual agglomeration is regulated to prevent possible negative impacts, which can be achieved through policies that promote sustainability and hold companies accountable for their environmental and social impacts.

6. Conclusions

This manuscript primarily concentrated on the U-shaped correlation between digital transformation and enterprise sustainability, emphasizing that firms may confront obstacles that impede their sustainability performance during the preliminary phases of digital transformation. As digital transformation progresses, companies can overcome these obstacles, achieving higher levels of sustainability. Empirical findings suggest that, in the initial stages, companies may face increased costs, disruptions in traditional business models, and employee resistance that negatively affect their sustainability performance. The implementation of digital transformation requires significant investments in infrastructure, employee training, and process re-engineering, contributing to a decline in sustainability performance. However, progress in digital transformation leads to increased efficiency, better decision-making capabilities, and the ability to innovate in an environmentally and socially responsible manner. The development of information technology enhances data access and communication, enabling companies to understand and respond to stakeholder needs and driving sustainable growth. The U-shaped relationship emphasizes the importance of persistence and long-term planning in digital transformation journeys.

Apart from the U-shaped association, this manuscript also probed the favorable moderating influence of regional virtual agglomeration on the correlation between digitalization and sustainability. We found that regional virtual agglomeration could help alleviate some of the difficulties faced by firms in the early stages of digital transformation, thereby accelerating this transition to higher levels of sustainability. Transaction cost theory and new economic geography provide valuable perspectives for us to test the moderating role of regional virtual agglomeration. According to the perspective of transaction cost theory, regional virtual agglomeration can help reduce the costs associated with information acquisition, negotiation, and contract execution during the digital transformation process. This reduction in transaction costs allows firms to allocate resources more efficiently to achieve their sustainability goals. Similarly, the new economic geography perspective emphasizes the importance of knowledge spillovers, network effects, and economies of scale, which can be achieved through regional virtual agglomerations. By promoting closer collaboration and knowledge sharing among firms, regional virtual agglomerations can facilitate the diffusion of digital technologies and best practices for sustainable development. In doing so, it can act as a gas pedal, helping companies overcome initial sustainability performance slumps and eventually reach higher levels of sustainable growth.

This manuscript also scrutinizes regional disparities by categorizing China into the eastern, central, and western regions. The outcomes revealed that the influence of digitalization on sustainable enterprise development differed by region, with pronounced regional disparities. In the eastern region, digitalization had a considerably constructive impact on sustainable enterprise development, predominantly owing to the benefits in policy endorsement, infrastructure, and human resources. On the contrary, digitalization has no significant impact on the sustainable development of enterprises in the central region. This was possibly due to the dominance of the heavy industry, greater resistance to digital transformation, and the lack of intrinsic attractiveness of regional agglomeration. In the western region, although the level of regional virtual agglomeration did not have a significant moderating effect, digital transformation showed a positive and significant relationship with enterprise sustainability, indicating that digital transformation in the western region has initially formed a scale effect and started to advance the enhancement of enterprise sustainability. The influence of digital transformation on enterprise sustainability in different regions is determined by various factors such as policies, infrastructure, and talent. Therefore, regional governments and enterprises should tailor their investment and support for digital transformation based on their own characteristics and development needs to foster sustainable enterprise development.

To conclude, this manuscript helped us understand more deeply the complex relationship between digital transformation, sustainable development and regional virtual agglomeration; it enriched the research connotation of the transaction cost theory and new geographic economics at a theoretical level. Studying the U-shaped relationship of digital transformation has important practical significance for the sustainable development of enterprises and can guide enterprises to effectively utilize their own advantages while considering potential risks and challenges. Practical implications include: (1) Helping enterprises make informed decisions when investing in digital technologies, as they can better predict the potential short-term costs and long-term benefits associated with these investments. (2) Enterprises can develop comprehensive strategies that take into account both the initial challenges and ultimate benefits of digital transformation, ensuring sustainable development in a more balanced manner. (3) Enterprises are able to allocate resources more effectively because they can recognize the need for dedicated support and training in the initial stages of digital transformation before realizing the benefits of improved efficiency and productivity in the long run. (4) By understanding U-shaped relationships, companies can better manage the change process associated with digital transformation, prepare employees for initial difficulties, and provide necessary support and resources to help them adapt to a new digital environment. (5) Communicating U-shaped relationships with stakeholders such as investors, employees and customers can help manage their expectations and build trust by demonstrating transparency and a realistic understanding of the challenges and opportunities associated with digital transformation.

In terms of management practices, by studying the regulating effect of the virtual agglomeration level, multiple competitive advantages can be created for enterprises: (1) Help enterprises improve knowledge overflow, optimize operations, reduce waste and achieve efficient and environmentally friendly innovation. (2) Help companies identify opportunities and leverage competition to achieve sustainable growth. (3) Help companies customize solutions for different regions, cultures and industries. In terms of policy formulation, studying the moderating effect of virtual agglomeration can help provide policymakers and regulators with accurate information and help them formulate effective policies, including encouraging enterprises to participate in the virtual agglomeration industry, developing guidelines for responsible data sharing, solving the digital divide, promoting the inclusiveness of enterprises in different cultures and regions.

Additionally, future research could explore the specific mechanisms and factors that contribute to the U-shaped correlation and the moderating role of regional virtual agglomeration, as well as the impacts of other factors, such as industry type and regulatory environment, on the relationship between digital transformation and sustainable enterprise development.