Frontiers in Business, Economics and Management ISSN: 2766-824X | Vol. 11, No. 1, 2023 117 The Impact of Digital Transformation on Manufacturing Firm Performance Xiaoxiao Yu1, a 1SolBridge International School of Business, Daejeon, Woosong University, Daejeon, 300-718, South Korea axyu012616@gmail.com Abstract: The manufacturing industry has undergone significant changes with digital technology evolving faster and faster, and the integration of digital solutions to improve enterprise performance. Based on the theoretical literature, the theoretical model of the role mechanism of "the degree of digital transformation - the level of manufacturing services - enterprise performance" is constructed by this paper. The study also verifies the impact of digital transformation on the performance of manufacturing enterprises in China and the mediating role played by the level of servitization. The results of this study are of great significance to policy makers and manufacturing companies seeking to improve their performance through digital transformation and servitization strategies. Keywords: Enterprise Performance, Digital Transformation, Service Level, Fixed effects model. 1. Introduction With the development of the intelligent age, digitalization, intelligence, service and greening are the main features of the new round of scientific and technological revolution and the focus of the new generation of information technology, the party's 19th National Congress report clearly puts forward "to strengthen the Internet, big data and artificial intelligence and the entity of the high degree of integration of the economy, and promote China's industry to the high-end of the global value chain". The digital transformation of businesses has become a major trend and an important step towards building a strong digital nation in China. Manufacturing is the foundation of a country's development and a key force in its economic development. However, many Chinese enterprises face challenges in achieving practical results in their digital transformation. Understanding how digital transformation can improve business performance and learning from successful transformation experiences are now important topics in academia and industry. With the above background, this paper will take the relevant data of manufacturing enterprises in China's Shanghai and Shenzhen cities from 2018 to 2022 as the research object and construct a multiple linear regression model, so as to analyze the effect of digital transformation on the performance of China's manufacturing enterprises, as well as to verify whether the servitization level plays a mediating role between the degree of enterprise's digital transformation and enterprise performance. 2. Theoretical Analysis and Research Hypothesis What is understood is that digital transformation is an ability to adapt to environmental changes by leveraging digital technologies, leading to improved operational efficiency and overall performance. Enterprise performance measures the business outcomes achieved through resource investment and reflects competitiveness. The current policy measures in China that encourage traditional enterprises to undergo digital transformation have provided a great incentive for physical enterprises to undergo digital transformation, and the increased investment in digitalization by traditional enterprises has significantly improved their own corporate performance [1,2]. Efficient asset utilization and cost-controlled innovation drive economic benefits. Based on this, according to the first hypothesis of this paper, digital transformation has a significant positive impact on China’s performance of manufacturing enterprises. The rapid development of digital transformation fosters diverse production models, and servitization through information technology (IT) is an essential means for manufacturing enterprises to create and obtain product value [3,4]. On the one hand, by integrating manufacturing with service business, companies optimize resource allocation, reduce service costs, and expand their service scale using digital technology [5,6]. On the other hand, manufacturing enterprises use the combination package of "product" + "service" establishes interactive customer relationships, boosts user loyalty, repeat purchases, and drives profitability [7,8]. Based on this, this paper proposes the second hypothesis of this paper: servitization in manufacturing has a positive mediating effect between digital technology adoption and firm performance. 3. Data Sources and Study Design 3.1. Sample and data This paper chooses and uses all manufacturing enterprises in Shanghai and Shenzhen cities from 2018-2022 as the research object. The relevant data of Chinese manufacturing enterprises are the research objects, and the listed data are taken from CSMAR database, Wind database, and the company annual reports disclosed by listed companies on a regular basis, whose data are scientific and authoritative. To guarantee the dependability of the study results, samples will be screened on the basis of the following criteria: (1) Excluding ST and *ST category listed companies. (2) Excluding enterprises listed after 2018. (3) Excluding samples with serious missing data on 118 explanatory variables, explained variables, and control variables in the model definition. (4) In order to avoid the negative effect of outliers on the results of the paper, the upper and lower 1% Winsorize shrinkage of all continuous variables was performed. Finally, 1470 sample firms were screened in this paper. 3.2. Variable selection and definitions 3.2.1. Explained variables The explained variable in this paper is corporate performance, ROA is selected as a measure of corporate performance, and ROE is selected as a proxy for ROA for robustness test. 3.2.2. Explanatory variables The digital capital investment rate will be selected to measure the level of digital transformation of manufacturing firms. The digital capital investment rate is the ratio of new electronic equipment hardware investment and software system investment to the total business revenue of the enterprise every year. 3.2.3. Mediating variables In this paper, the servitization level of manufacturing industry is selected as the mediating variable. The ratio of other business income to total income of enterprises is mainly used as an norm to measure the standard of servitization. 3.2.4. Control variables In this paper, enterprise size, proportion of shares held by the first largest shareholder, debt-to-assets ratio and growth rate of operating income are chosen as the control variables. Table 1. Variable definition table Variable Type Variable Symbol Definition Explained Variables Enterprise Performance ROA Return on Total Assets ROE Return on Net Assets Explanatory Variables Digital Transformation Dig Ratio of new electronic equipment hardware investment and software system investment to total business revenue per year Mediating Variables Service Level Serve Ratio of other operating income to total revenue of the enterprise Control Variables Enterprise size Size Log of total assets Gearing Ratio Lev Total liabilities / Total assets Shareholding Ratio of the Largest Shareholder Top1 Number of shares held by the largest shareholder / Total number of shares Operating Revenue Growth rate Growth (Current year's operating revenue - Prior year's operating revenue) / Prior year's operating revenue 3.3. Model design In order to test hypothesis 1, Model1 and Model2 are developed in this paper, where both Model1 and Model2 are multiple linear regression models with the explanatory variable enterprise performance, the explained variable digital transformation and several control variables. The difference is that Model1 uses the ROA indicator for firm performance and Model2 uses the ROE indicator for firm performance. ititititititit GrowthTopLevSizeDigROA   33321 1 (Model1) ititititititit GrowthTopLevSizeDigROE   33321 1 (Model2) To test hypothesis 2, this paper will use stepwise regression to verify the existence of its mediating role.Model3 is a regression model of the mediating variable servitization level and the digital transformation of explanatory variables. Model4 is a regression model of the explained variable firm performance, the explanatory variable digital transformation, the mediating variable servitization level and several control variables. ititit DigServe   1 (Model3) itititititititit GrowthTopLevSizeServeDigROA   654321 1 (Model4) 4. Empirical Tests and Analysis of Results 4.1. Descriptive statistics In this paper, descriptive statistical analyses of the explained variables firm performance, explanatory variables digital transformation and control variables indicators were conducted as follows: Table 2. Descriptive statistics Variable Mean Std. Dev. Min Max Observations ROA 0.0448 0.0725 -0.2529 0.2427 7350 ROE 0.0666 0.1276 -0.5325 0.3556 7350 Dig 0.1739 0.1090 0.0118 0.6109 7350 Serve 0.1739 0.1075 0.0199 0.5475 7350 Size 22.1974 1.1577 20.1403 25.8610 7350 Lev 0.3912 0.1733 0.0685 0.7977 7350 Growth -1.0265 6.5384 -45.4974 17.0923 7350 Top1 0.3144 0.1364 0.0826 0.6770 7350 119 4.2. Relevance analysis In this paper, the relevance analysis of the explained variable firm performance, the explanatory variable digital transformation and the control variable indicators were conducted as follows: Table 3. Correlation analysis ROA ROE Dig Serve Size Lev Growth Top1 ROA 1 ROE 0.941*** 1 Dig 0.014*** 0.035*** 1 Serve 0.221*** 0.181*** 0.086*** 1 Size 0.086*** 0.164*** -0.099*** -0.034*** 1 Lev -0.348*** -0.226*** -0.089*** -0.262*** 0.448*** 1 Growth 0.461*** 0.521*** 0.015 0.064*** 0.028** -0.092*** 1 Top1 0.174*** 0.171*** -0.141*** 0.086*** 0.060*** -0.056*** 0.098*** 1 From the above table, it is clear that The relevance coefficient between ROA and Dig is 0.014 The relevance coefficient between ROA and Serve is 0.221 The relevance coefficient between ROA and Size is 0.086 The relevance coefficient between ROA and Lev is -0.348 The relevance coefficient between ROA and Growth is 0.461 The relevance coefficient between ROA and Top1 is 0.174 and all those are significant at the 1% level 4.3. Multicollinearity test One of the assumptions of multiple linear regression is that there has no high relevance among explanatory variables, otherwise when there has a relevance between the data, the analysis results will be biased. Therefore, in this paper, a multiple cointegration test was conducted before multiple regression, and the test results were as follows: Table 4. Multicollinearity test Variable VIF 1/VIF Lev 1.32 0.755574 Size 1.31 0.762472 Top1 1.04 0.962492 Dig 1.04 0.964568 Growth 1.02 0.976187 Mean VIF 1.15 As can be seen from the above table, the expansion factor of the explanatory variable Digitized Transformation Dig and a number of control variables is less than 5, which indicates that there is no high relevance among the variables, i.e., there is no serious collinearity problem, i.e., the data taken in this paper are valid and can be carried out further empirical analysis. 4.4. Effect model In this paper, we will determine the effect model by LM test and Hausman test. Table 5. LM test and Hausman test Test Chi2 Value P Value LM 64.25 0.0000 Hausman 21.63 0.0014 From the above table, the Chi2 value of 64.25 and p-value of 0.0000 in the LM test results indicate that the random effect is better than the mixed regression. Then from the table, the results of Hausman test, Chi2 value is 21.63 and the p-value is 0.0014<0.01, which indicates that the fixed effect is better than the random effect. So this paper should use fixed effect model. 4.5. Basis regression and robustness test In this paper, multiple linear regressions will be conducted for Model1 and Model2 respectively using fixed effects model.ROA indicator is used as the base regression model for the explanatory variable firm performance in Model1.ROE indicator is used as the robustness check model for the explanatory variable firm performance in Model2. The specific regression results are as follows: 120 Table 6. Base regression and robustness test Model1 Model2 ROA ROE Dig 0.0009* 0.0009* (2.06) (2.03) Size 0.0168*** 0.0347*** (8.49) (10.05) Lev -0.1162*** -0.1067*** (-14.02) (-7.38) Growth 0.0035*** 0.0068*** (41.16) (45.75) Top1 0.0393*** 0.0698*** (2.75) (2.80) _cons -0.2809*** -0.6565*** (-6.32) (-8.47) R2 0.299 0.323 adj. R2 0.093 0.124 F 425.8046 474.6537 N 6450 6450 t statistics in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01 As can be seen from the regression results of Model1, the regression coefficient of Dig is 0.0009 and the t-value is 2.06, with a significant positive relevance at the level of 10%. It illustrates that the deeper the extent of digital transformation of manufacturing enterprises in China, the higher the level of their corporate performance. The explanation for this result is as follows, the increased digital investment of traditional enterprises significantly improves their own corporate performance. After digital transformation, enterprises improve the economic benefits of digital transformation by increasing the efficiency of asset use and improving their independent innovation capacity on the basis of cost control. Then from the regression results of Model2, the regression coefficient of Dig is 0.0009 and the t-value is 2.03, with a significant positive relevance at the level of 10% and remains basically consistent with the results of Model1. It indicates that the regression underlying this paper passes the robustness test and the degree of digital transformation of Chinese manufacturing enterprises has a significant positive effect on their enterprise performance. 4.6. Intermediary mechanism test Table 7. Intermediary mechanism test Model3 Model4 Serve ROA Dig 0.0370*** 0.0015* (3.46) (2.05) Serve 0.0571*** (3.99) Size 0.0329*** (9.46) Lev -0.0966*** (-6.59) Growth 0.0068*** (45.86) Top1 0.0678*** (2.72) _cons -0.0836 -0.6291*** (-1.33) (-8.10) R2 0.003 0.325 adj. R2 -0.247 0.126 F 9.3036 399.3874 N 7350 6450 t statistics in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01 From the regression results of Model3, the regression coefficient of Dig is 0.0370 and t-value is 3.46, which is significant at 1% level. From the regression results of Model4, the regression coefficient of Dig is 0.0015, the t-value is 2.05, which is significant at 10% level; the regression coefficient of Serve is 0.0571, the t-value is 3.99, which is significant at 1% 121 level. In conclusion, it shows that the digital transformation of Chinese manufacturing firms can enhance their service level, which in turn improves their corporate performance. 5. Research Findings and Policy Implications 5.1. Research conclusions Based on the 2018-2022 data of digital transformation, service level, and enterprise performance of sample manufacturing enterprises, this paper uses a multiple linear regression model for empirical analysis and combines theoretical analysis with empirical research to explore the relationship among manufacturing digitalization, service level, and enterprise performance. This paper reaches the following important conclusions: (1) Digital transformation of the sample manufacturing industry can promote the improvement of enterprise performance. It is serving as a long-term strategy to enhance competitiveness and adapt to the digital economy. (2) Servitization in manufacturing industry has a positive mediating effect between digital technology application and enterprise performance. Improving servitization levels enhances production and operational efficiency, resulting in improved firm performance. 5.2. Policy Implications 5.2.1. Increase digital investment and enhance enterprise value chain For enterprises, it suggests enhancing digital investment in infrastructure, such as artificial intelligence and internet technology, while leveraging software tools like big data analysis and industrial network design for improved supply chain collaboration and market responsiveness. The integration and alliance between software and manufacturing industries are vital for driving digital transformation, establishing a collaborative digital platform, and adapting to market demands in China's manufacturing sector. 5.2.2. Promote the service of manufacturing industry and extend the industrial chain To enhance the service aspect of manufacturing enterprises, they should integrate digitalization across multiple fields and diversify their offerings. Extending the manufacturing industry's value chain improves the quality of terminal services and generates revenue through manufacturing services. The development trend of manufacturing service focuses on connecting with consumer demands, enhancing consumer satisfaction, and elevating the service level of traditional manufacturing. References [1] Wang W-L., Yang H-Q. (2017) A literature review on the relationship between investment in informatization and firm performance. Journal of Economic Research, 34: 8-11+5 [2] Wang C. (2021) A study on the mechanism of digital transformation on the innovation performance of enterprises. Contemporary Economic Management, 43(03): 34-42 [3] Dai X. (2016) The evolution of the value of services embedded in China's manufacturing exports and the determination of factors. Economic Research, 51(9): 15 [4] Zhou W-H., Wang P-C., Yang M. (2018) Digital empowerment for mass customization technology innovation. Scientology Research, 36(8): 1516-1523. [5] Jiang C., Li N.. (2015) The impact of service innovation and manufacturing servitization on firm performance. Scientific Research Management, 5: 29-37 [6] Hu, C-P., Wang T. (2016) Transformation and upgrading of manufacturing servitization strategy: a theoretical model of evolutionary path: a case study of three local manufacturing enterprises. Research Management, 11: 119-126. [7] Hu C-P., Hu Q-F. (2020) A theoretical framework for the construction of competitive advantage of servitization strategy in manufacturing industry--a multi-case analysis based on manufacturing industry. China Circulation Economy, 34(04) :87-99 [8] Park Q-X., Sun X-B., Qian Y., Jin Z-H. (2020) Research on the interaction mechanism of technological innovation and business model innovation from the perspective of service transformation--a case study of Shenyang Machine Tool Group. Science and Technology Management, 4: 1-21.