The impact of digital transformation on innovation performance - The mediating role of innovation factors

Abstract

Existing digital transformation research has focused on economic and environmental performance, which few studies directly explored the relationship between digital transformation and innovation. Based on the innovation factor perspective, we explored the relationship between digital transformation and innovation by using firm data between 2009 and 2019. The findings are as follows: (1) The corporate digital transformation was measured through based on textual analysis methods and it was found that digital transformation can promote corporate innovation. (2) Knowledge flow, technical personnel, R&D investment, and innovation awareness are important mediating paths. (3) In the innovation quantity dimension, the mediating role of innovation awareness is greater. And in the innovation quality dimension, the mediating role of technicians is greater. (4) Digital transformation has a greater impact on innovation of non-SOEs, non-high-tech enterprises and non-heavily polluting enterprises, alleviating the gap between different types of firms. The results of this paper alleviate the concerns of digital transformation in developing countries such as China and provide experiences and evidence for them to promote Industry 4.0 and sustainable innovation.

Full text

Heliyon 9 (2023) e13916
Available online 22 February 2023
2405-8440/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Research article
The impact of digital transformation on innovation performance -
The mediating role of innovation factors
Pengyu Chen, SangKyum Kim
*
Department of Economics, College of Business and Economics, Dankook University, South Korea
ARTICLE INFO
Keywords:
Digital transformation
Innovation performance
Innovation factors
Mechanism analysis
ABSTRACT
Existing digital transformation research has focused on economic and environmental perfor-
mance, which few studies directly explored the relationship between digital transformation and
innovation. Based on the innovation factor perspective, we explored the relationship between
digital transformation and innovation by using rm data between 2009 and 2019. The ndings
are as follows: (1) The corporate digital transformation was measured through based on textual
analysis methods and it was found that digital transformation can promote corporate innovation.
(2) Knowledge ow, technical personnel, R&D investment, and innovation awareness are
important mediating paths. (3) In the innovation quantity dimension, the mediating role of
innovation awareness is greater. And in the innovation quality dimension, the mediating role of
technicians is greater. (4) Digital transformation has a greater impact on innovation of non-SOEs,
non-high-tech enterprises and non-heavily polluting enterprises, alleviating the gap between
different types of rms. The results of this paper alleviate the concerns of digital transformation in
developing countries such as China and provide experiences and evidence for them to promote
Industry 4.0 and sustainable innovation.
1. Introduction
Since the beginning of 21st century, the application of digital, networked and intelligent technologies has given rise to the fourth
industrial revolution-Industry 4.0 [1]. Industry 4.0 emphasizes the use of the convergence of information technology and
manufacturing technology to change the current industrial production and service models [2]. However, Industry 4.0 can not be
realized easily, the process of companies transforming from industrialization to Industry 4.0 smart factories is dened as digital
transformation [3]. The proliferation of digital technologies such as big data, cloud computing, Internet of Things, blockchain, and
robotics are disrupting traditional business models and innovation approaches [4]. As an important tool for achieving sustainable
development, innovation is receiving increasing attention in business research [5]. Therefore, it is of great value to explore the impact
of digital transformation on innovation as well as the mechanisms by which digital transformation works.
Existing research on digital transformation has focused on economic efciency, environmental performance and business models
[6]. [7] adopted listed companies data from 2007 to 2019 and found that digital transformation increased the total factor productivity.
An existing literature used digital transformation as an intermediate variable to explore the impact of the digital economy on inno-
vation [8]. The cited authors used the number of patent applications to measure innovation performance and found that the digital
economy could stimulate rm innovation by driving the digital transformation. As can be seen, few studies have directly explored the
* Corresponding author.
E-mail addresses: cpy702018@163.com (P. Chen), iamskkim@dankook.ac.kr (S. Kim).
Contents lists available at ScienceDirect
Heliyon
journal homepage: www.cell.com/heliyon
https://doi.org/10.1016/j.heliyon.2023.e13916
Received 11 August 2022; Received in revised form 9 February 2023; Accepted 16 February 2023

Heliyon 9 (2023) e13916
2
relationship between digital transformation and innovation [9,10]. [11] pointed out that simply using the number of patent appli-
cations to measure rm innovation is imperfect for the reason of invention patents and appearance utility patents have different
characteristics on economic and environmental dimensions. In addition, many scholars mentioned that digital transformation is
disrupting traditional innovation models [4], but few scholars explored how digital transformation affects innovation. An in-depth
study of this relationship provides a new empirical basis for business managers to change their traditional thinking about innova-
tion [12]. Therefore, invention patents and utility patents are used in this paper to capture the quantity and quality of innovation, the
relationship between digital transformation and innovation are explored from the perspective of innovation factor ow, which ll the
gap of digital transformation research.
First, the relationship between digital transformation and innovation are explored by taking using Chinese manufacturing listed
companies as the research objects. Second, based on the factor mobility perspective, the mechanisms of the role of digital trans-
formation (knowledge mobility, technician mobility, R&D investment and innovation awareness) are explored. Finally, based on
enterprise heterogeneity, the enterprises are classied into state-owned and non-state-owned enterprises, high-tech and non-high-tech
enterprises, and heavily polluting and non-heavily polluting enterprises. Through this study, to answer the following questions can be
sought: 1) Can digital transformation increase corporate innovation? 2) What innovation factor ows are being changed by digital
transformation? 3) Which kind of corporate innovations are more sensitive to digital transformation?
This paper has some potential contributions. First, unlike previous studies, this paper focuses on the impact of digital trans-
formation on innovation quantity and innovation quality, which enriches the literature on digital transformation. Second, we intro-
duce the innovation factors from a factor mobility perspective. The functional mechanisms of digital transformation are explored,
providing us with evidence about how digital transformation subverts traditional innovation models. Third, the attitudes of different
types of corporate toward digital transformation are explored through heterogeneity analysis, which provides guidance directions for
managers of specic types of companies. Finally, possible endogeneity issues are tested in digital transformation to ensure the stability
of our ndings. Overall, this study lls an existing gap and provides a clear and valuable empirical basis for policy makers and business
managers.
2. Literature review and research hypotheses
2.1. The digital transformation and innovation performance
The denition of organizational boundaries is that all organizations have boundaries; there are boundaries within and outside the
organization, as well as between organizations [13]. With the development of technology [14], organizational boundaries are
becoming blurred. In other words, the widespread use of digital technology enhanced the ow of information between organizations,
and within organizations, and eliminating inter-organizational boundaries [15]. Digital technologies can help rms break through own
resource and capability deciencies and leverage external resources and capabilities to achieve strategies of expanding into new
markets and developing new products. Based on a study of Chinese manufacturing A-share listed companies during 2016–2019 [16],
found that the application of blockchain technology improved the collaborative innovation performance.
Furthermore, according to the information processing view (IPV), business performance can be improved when information
processing capabilities can meet the needs of the business [17]. Digital technology can enhance the information processing capabilities
of the rm and reduce the interference of non-demanded information [18]. Specically, information is a carrier of knowledge, and
digital technology can help rms keep track of the information they need in a timely manner [19]. [20] found that rms with strong
information processing capabilities were more likely to adjust their innovation strategies timely to respond to changing circumstances
[21]. Digital technologies can also help rms grasp product sales situation and customer demand to adjust their innovation direction
[22]. [23] found that information processing capabilities can positively improve the relationship between information-intensive
customer engagement (ICI) practices and the number of innovations. Based on the above theory, the following hypothesis are
proposed.
H1a. Digital transformation can improve corporate innovation quantity
H1b. Digital transformation can improve corporate innovation quality
2.2. The mediating role of innovation factors
The innovation factor refers to the resources and capabilities associated with innovation, it is argued that innovation factors may be
an important way in which digital transformation affects innovation performance [24]. Considering the availability of data, factors
including knowledge ow, technical staff, R&D investment and innovation awareness are only considered as innovation factors [25,
26]. The information processing view states that information processing capabilities reect a rm’s sensitivity to information and that
rms can more likely maintain a market advantage when they have timely access to the information they need [27]. The application of
digital technologies can improve the rm information processing ability so that they can capture timely information that can help them
innovate. In addition [15], pointed out that the prevalence of digital technology broke through traditional organizational boundaries
and had a signicant impact on organizational boundary theory. Digital technologies weaken the communication barriers and
accelerate the ow of factors between rms [28]. The role of innovation factors can also be explained by resource-based theory, which
states that rms can maintain competitive advantage when they have “heterogeneous” resources. Because then, they gain the ability to
innovate sustainably [29].
P. Chen and S. Kim

Heliyon 9 (2023) e13916
3
Knowledge ow and talent mobility are considered as one of the important paths of action for digital transformation [30]. used rm
data from 1992 to 1998 and found that the use of Internet technologies can overcome distance barriers between rms and increase the
likelihood of referencing knowledge between the two (knowledge ow increased) [31]. used rm data from 117 countries from 2006
to 2011 and found that the application of advanced technologies led to knowledge spillovers, which in turn promote rm innovation.
By using a questionnaire [32], found that digital transformation drives rm knowledge sharing, which in turn stimulates innovation.
Digital technologies may be changing employment patterns and work environments [33]. Digital technologies are making it easier for
companies such as monster, saongroup, 51job, and other job boards to recruit to meet their demand for highly skilled talent. Digital
transformation implies the use of new technologies, leading to a high demand for technical talent [34]. Technologists have a higher
ability to accept, decode and apply new knowledge, increasing the innovation capacity [35]. Based on the above theoretical analysis,
the following hypothesis are proposed.
H2a. Digital transformation can facilitate knowledge ows
H2b. Knowledge ows can stimulate innovation
H3a. Digital transformation can facilitate technology talent mobility
H3b. Technology talent mobility can inspire innovation
In addition, it is believed that R&D investment and innovation awareness are also important role paths of digital transformation.
Digital technologies have signicant advantages in evaluating innovation projects [36], such as project riskiness, protability, etc.
[37] pointed out that efcient digital technology can attract external investors, which in turn boosts R&D investment. In addition, the
use of digital technology stimulates the innovation of business managers. The ability to capture information, the ability to evaluate
projects and various innovation elements encourage business managers to adopt aggressive innovation strategies [38]. [39] mentioned
that digital technology encourages rms to exibly innovate. Both innovation awareness and R&D investment can touch the inno-
vation output of rms [40,41]. Based on the above theoretical analysis, the following hypotheses are proposed.
H4a. Digital transformation can increase R&D investment
H4b. R&D investment stimulates innovation
H5a. Digital transformation stimulates innovation awareness
H5b. Innovation awareness promotes innovation
3. Data and model design
3.1. Sample
We used the original sample of listed Chinese manufacturing companies. Considering the impact of extreme values such as the
nancial crisis in 2008 and covid-19 in 2020 on the regression results, our study years are 2009–2019. Financial data were obtained
from the China Stock Market & Accounting Research Database (CSMAR) and the Wind database. We excluded rms with signicant
information deciencies and nally obtained 16,159 rm-year observations.
3.2. Variables
3.2.1. Dependent variable: innovation performance
We measure the innovation performance from two perspectives: innovation quantity and innovation quality, namely, using patent
applications (Pat) to measure innovation quantity and invention patent applications (Inpat) to measure innovation quality. On the one
hand, patent data depicts a more intuitive picture of a rm innovation level compared to R&D investment [42]. On the other hand,
according to China’s patent law, patents are classied into invention patents, utility patents and design patents. Inventive patents have
high-value characteristics compared to other patent types [43]. We use the natural logarithm of patent +1 to measure rm innovation
performance.
3.2.2. Independent variable: digital transformation (DT)
It is difcult to describe a rm’s digital transformation with a single variable [44], we use a new approach to quantify the digital
transformation of rms - textual analysis. We use the logarithm of the frequency of relevant terms in nancial annual reports to
quantify digital transformation [45]. We retrieved keywords related to digital technologies based on the methods of [8]. First, we used
deep learning methods to obtain 153 related keywords, such as, big data, information technology, intelligence, robotics, Internet of
Things, blockchain, automation, digitization, and cloud computing. Then, the frequency of these keywords in corporate annual
nancial reports was retrieved using text analysis methods. In addition, two independent researchers re-examined the keywords to
ensure the accuracy of the vocabulary. These data acquisitions and methods were based on the Wingo text analytics platform, which is
the rst intelligent platform for text retrieval in China.
3.2.3. Mediation variables: innovation factors
Innovation factors refer to the combination of resources and capabilities associated with innovation, including knowledge capital,
P. Chen and S. Kim

Heliyon 9 (2023) e13916
4
human capital, physical capital, and innovation awareness [46]. We use patent citation frequency to measure knowledge ows (the
percentage of patent citations to the industry average, marked as KF, the number of technicians to measure human capital (TE), and the
natural logarithm of R&D investments (RD) to describe physical capital [47–49]. We explore the frequency of the occurrence of
innovation-related keywords in nancial annual reports to measure innovation awareness marked as IA [50]. Keywords include jìshù
(Technology), chu`
angx¯
ın (innovation), y´
anji¯
u (Research), k¯
aif¯
a (Development), zhu¯
anlì (Patant), f¯
amíng (Invention), k¯
ejì (Science &
Technology), zh¯
ıshi (Knowledge), and other pinyin Chinese words.
3.2.4. Control variables
Referring to Refs. [51,52] study, we control for the following variables: age, rm growth, rm size, cash ow, debt ratio, xed
capital investment ratio, board size, and shareholding concentration. 1) Age is the logarithm of the age of the enterprise. 2) Firm
growth (Growth) is the growth rate of operating income.3) Firm size (Size) is the logarithm of the number of employees. 4) Cash ow
(Cash) is the ratio of cash generated from operations to total assets. 5) The debt ratio (Lev) is the ratio of total liabilities to total assets.
6) Fixed capital investment (Fix) ratio is the ratio of net xed assets to total assets. 7) Board size (BS) is the logarithm of the total
number of board members. 8) Shareholding concentration (SC) is the sum of the proportion of shares held by the top 5 largest
shareholders. Variables are dened in Table 1.
3.3. Model design
We construct the following model to test the relationship between digital transformation and innovation as equation (1).
Pati,t(Inpati,t)=β0+β1DTi,t+∑
8
i=1
γicontrol +
μ
i+δt+
ε
i,t(1)
where the dependent variables are quantity of innovation (Pat) and quality of innovation (Inpat); DT denotes digital transformation;
controls denote the nine control variables; and
ε
i,t is a random disturbance term. Fixed time effects (δt) and individual effects (
μ
i) are
used to mitigate the disturbance of the unobserved terms.
To test the mechanisms of digital transformation, the following model was built as equations (2) and (3).
Medi,j,t=
α
0+
α
1DTi,t+∑
8
i=1
θicontrol +
μ
i+δt+
ε
i,t(2)
Pati,t(Inpati,t)=
α
0+
α
1DTi,t+
α
jMedi,j,t+∑
8
i=1
θicontrol +
μ
i+δt+
ε
i,t(3)
where the Medi,j,t is the innovation factor, including knowledge ows (KF), technicians (TE), R&D investment (RD) and innovation
awareness (IA).
4. Results and discussion
4.1. Descriptive statistics and correlation analysis
Table 2 shows that the mean, maximum and minimum values for quantity of innovation are 2.528, 0 and 8.750, respectively. The
mean value for quality of innovation is only 1.266. This means that most Chinese companies are in a low-quality innovation stage and
Table 1
Variable denition.
Name Variable Denition Previous studies Source
Dependent variable Pat The natural logarithm of using patent applications [42] Wingo
Inpat The natural logarithm of invention patent applications [43] Wingo
Independent variable DT Logarithm of the frequency of digital-related words in nancial annual reports [8] Wingo
Mediation variables KF The percentage of patent citations to the industry average [47] Wingo
RD The natural logarithm of R&D investment [49] CSMAR
TE The number of technicians [48] CSMAR
IA Frequency of innovation-related keywords in nancial annual reports [50] Wingo
Control variables Age Logarithm of the age of the enterprise [51,52] CSMAR
Growth The growth rate of operating income
Size The logarithm of the number of employees
Cash The ratio of cash generated from operations to total assets
Lev The ratio of total liabilities to total assets
Fix The ratio of net xed assets to total assets
BS The logarithm of the total number of board members
SC The sum of the proportion of shares held by the top 5 largest shareholders
P. Chen and S. Kim

Heliyon 9 (2023) e13916
5
prefer quantitative level of innovation. The maximum (6.864), minimum (0) and mean (1.959) values of digital transformation tell us
that most Chinese rms are in the industrialization stage and are in low-level digital transformation. The situation is the same for the
other variables and the development is very uneven among the companies.
In addition, Table 3 provides the results of the correlation regression of the quantity of innovation and the quality of innovation. We
found that the correlation coefcients are both less than 0.7 [53]. And the multicollinearity test nds a VIF value of 1.15 for both,
which is less than 10 [54]. This indicates that there is no multicollinearity in our model.
4.2. Baseline analysis
Table 4 shows the regression results of digital transformation on the innovation quantity and quality. The results show that the
coefcient of digital transformation is signicantly positive, regardless of the xed time effect or the added control variables. Taking
columns (3) and (6) as an example, the impact of digital transformation on the innovation quantity is 0.095 at the 1% signicant level
and the impact on the innovation quality is 0.070 at the 1% signicant level. It is consistent with hypotheses H1a and H1b. This result
suggests that digital transformation has a greater impact on the innovation quantity than on the innovation quality.
4.3. Robustness test
Table 5 shows the regression results for the three robustness tests. Columns (1) and (2) are the results of regressions using granted
patents as a measure of innovation performance. Considering the possible endogeneity problem within the model. We use the number
of Internet users in each province as an instrumental variable (IV) to address the endogeneity problem [55]. Internet users per province
implies a regional level of digitization that can affect the corporate digital transformation without directly affecting the corporate
strategy. The rst stage (column 3) shows that the number of Internet users facilitates the digital transformation. The F-value is 509.22,
which is much larger than 10. This indicates that the choice of our instrumental variable is effective [56]. The second stage regression
results (columns 4 and 5) tell us that digital transformation has a signicant positive impact both in terms of quality and quantity of
innovation. Column (6) and (7) show the regression results after we exclude rms located in municipalities and provincial capitals.
This is because provincial capitals tend to receive more resource bias and attention [57]. All data are from provincial statistical
yearbooks. The results in Table 5 show that regression results are found to be robust.
4.4. The mediating role of innovation factors
Tables 6 and 7 shows the mechanism of action of digital transformation. The path from digital transformation to knowledge ow is
signicant and hypothesis H2a is supported (β =0.0007, p <0.1). The path from digital transformation to technicians is signicant,
with hypothesis H3a supported (β =21.584, p <0.01). Digital transformation to R&D investment (H4a: β =0.356, p <0.01) and
innovation awareness (H5a: β =0.001, p <0.01) on the path was also considered signicant. The coefcient of the mediating variable
remains signicantly positive after the inclusion of the mediating variable. This veries the hypotheses H2b, H3b, H4b, H5b. This
suggests that knowledge mobility, technicians, R&D investment, and innovation awareness are important paths through which digital
transformation promotes innovation.
Table 8 describes the intensity of each effect. It can be seen that at the level of the number of innovations, the direct effects of digital
transformation are much greater than the various indirect effects. Taking the path of knowledge ow as an example, the indirect effect
of digital transformation is 0.002 at the 1% signicant level, while the direct effect is 0.093 at the 1% signicant level. Similarly, at the
level of innovation quality, the direct effects of digital transformation are much larger than the indirect effects. To test the stability of
the various effects, we veried the stability of the results by 1000 bootstrap method [58].
Table 2
Descriptive statistics.
Variable Obs Mean Std. Dev. Min Max
Pat 16,159 2.528 1.570 0 8.750
Inpat 16,159 1.266 1.246 0 8.038
DT 16,159 1.959 1.379 0 6.864
RD 14,966 17.74 1.490 8.006 24.615
IA 16,159 0.900 0.391 0.045 3.394
KF 16,159 0.019 0.062 0.00001 1
TE 16,111 634.656 981.737 29 5287
Age 16,159 2.695 0.423 0 3.951
Growth 16,158 0.253 5.385 −0.991 666
Size 16,153 7.709 1.170 0.693 12.342
Cash 16,158 0.169 0.139 −0.165 0.883
Lev 16,159 0.409 0.556 0.007 42.857
Fix 16,158 0.235 0.142 0 0.902
BS 16,159 8.567 1.619 0 18
SC 16,159 53.621 14.715 23.476 82.996
P. Chen and S. Kim

Heliyon 9 (2023) e13916
6
4.5. Heterogeneity analysis
Table 9 shows the results of the regressions based on rm heterogeneity. We classify the enterprises into state-owned and non-state-
owned enterprises (SOE and Non-SOE), high-tech and non-high-tech enterprises (HTE and Non-HTE), and heavily polluting and non-
heavily polluting enterprises (HPE and Non-HPE). The classication is based on the information provided by the CSMAR database.
The ndings in columns (1) and (2) tell us that the coefcient of digital transformation for SOEs is 0.082 at the 1% signicant level
and the coefcient of digital transformation for non-SOEs is 0.103 at the 1% signicant level. This implies that digital transformation
has a greater impact on innovation in non-SOEs. Columns (3) and (4) show the regression results for HTE and Non-HTE. We nd that
digital transformation has a greater impact on innovation of Non-HTE (β =0.126, p <0.001). The results in columns (5) and (6) show
Table 3
Correlation analysis.
1 2 3 4 5 6 7 8 9 10
1.Pat 1.000
2.DT 0.351* 1.000
3.Age 0.035* 0.076* 1.000
4.Growth 0.100* 0.082* −0.112 1.000
5.Size 0.427* 0.059* 0.139* 0.021* 1.000
6.Cash −0.005 0.045* −0.147* 0.073* −0.196* 1.000
7.lev 0.177* −0.066* 0.136* −0.021* 0.469* −0.441* 1.000
8.Fix −0.111* −0.256* 0.055* −0.090* 0.219* −0.389* 0.215* 1.000
9.BS 0.080* −0.128* 0.009 −0.009 0.249* −0.016* 0.161* 0.097* 1.000
10.SC 0.014 0.061* −0.198* 0.086* 0.001 0.164* −0.183* −0.067* −0.041* 1.000
1.Lnpat 1.000
2.DT 0.327* 1.000
3.Age 0.060* 0.076* 1.000
4.Growth 0.089* 0.082* −0.112* 1.000
5.Size 0.381* 0.059* 0.139* 0.021* 1.000
6.Cash −0.019* 0.045* −0.147* 0.073* −0.196* 1.000
7.lev 0.162* −0.066* 0.137* −0.021* 0.469* −0.441* 1.000
8.Fix −0.088* −0.256* 0.055* −0.090* 0.219* −0.389* 0.215* 1.000
9.BS 0.087* −0.128* 0.009 −0.009 0.249* −0.016* 0.161* 0.097* 1.000
10.SC −0.014 0.061* −0.198* 0.086* 0.001 0.164* −0.183* −0.067* −0.041* 1.000
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
Table 4
Baseline analysis.
(1) (2) (3) (4) (5) (6)
Variables Pat Pat Pat Inpat Inpat Inpat
DT 0.408*** 0.136*** 0.095*** 0.314*** 0.098*** 0.070***
(0.009) (0.011) (0.010) (0.007) (0.009) (0.009)
Age 0.136* 0.146*
(0.070) (0.060)
Growth −0.003* −0.002
(0.001) (0.001)
Size 0.473*** 0.322***
(0.016) (0.014)
Cash −0.082 0.012
(0.076) (0.065)
Lev 0.042** 0.047*
(0.014) (0.012)
Fix −0.363*** −0.210*
(0.098) (0.083)
BS 0.012 0.017*
(0.008) (0.007)
BC 0.002* 0.002*
(0.001) (0.001)
C 1.728*** 1.285*** −2.568*** 0.650*** 0.344*** −2.554***
(0.018) (0.032) (0.224) (0.016) (0.027) (0.191)
Firm Yes Yes Yes Yes Yes Yes
Year No Yes Yes No Yes Yes
Obs 16,159 16,159 16,151 16,159 16,159 16,151
R-sq 0.723 0.756 0.771 0.687 0.725 0.736
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
P. Chen and S. Kim

Heliyon 9 (2023) e13916
7
Table 5
Robustness test.
Replace the dependent variable IV method Replace the sample
(1) (2) (3) (4) (5) (6) (7)
Variables Pat Inpat DT Pat Inpat Pat Inpat
DT 0.112*** 0.107*** 0.535** 0.660** 0.099*** 0.116***
(0.011) (0.010) (0.232) (0.211) (0.012) (0.012)
IV 0.0001***
(0.000)
C −2.207*** −2.355*** 1.222*** −2.948*** −3.184*** −2.329*** −2.207***
(0.234) (0.216) (0.130) (0.460) (0.418) (0.266) (0.254)
Control Yes Yes Yes Yes Yes Yes Yes
Firm Yes Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes Yes
F statistic 509.22
Obs 16,150 16,150 10,021 10,201 10,201 11,368 11,368
R-sq 0.770 0.760 0.141 0.799 0.790 0.760 0.748
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
Table 6
The mediation of innovation factors on Pat.
(1) (2) (3) (4) (5) (6) (7) (8)
Variables RD AI KF TE Pat Pat Pat Pat
DT 0.356*** 0.001*** 0.0007** 21.584* 0.071*** 0.093*** 0.077*** 0.090***
(0.008) (0.003) (0.000) (1.831) (0.011) (0.010) (0.011) (0.011)
RD 0.163***
(0.011)
KF 2.799***
(0.310)
AI 19.788**
(0.035)
TE 0.0002**
(0.011)
Control Yes Yes Yes Yes Yes Yes Yes Yes
Firm Yes Yes Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes Yes Yes
Obs 14,959 16,150 16,150 16,136 14,959 16,150 16,150 14,959
R-sq 0.853 0.782 0.894 0.279 0.773 0.773 0.772 0.773
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
Table 7
The mediation of innovation factors on Lnpat.
(1) (2) (3) (4) (5) (6) (7) (8)
Variables RD AI KF TE Lnpat Lnpat Lnpat Lnpat
DT 0.356*** 0.001*** 0.0007** 21.584* 0.047*** 0.067*** 0.064*** 0.061***
(0.008) (0.003) (0.000) (1.831) (0.009) (0.009) (0.009) (0.009)
RD 0.116***
(0.010)
KF 4.183***
(0.263)
AI 6.525*
(0.029)
TE 0.00003***
(0.00001)
Control Yes Yes Yes Yes Yes Yes Yes Yes
Firm Yes Yes Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes Yes Yes
Obs 14,959 16,150 16,150 16,136 14,959 16,150 16,150 14,959
R-sq 0.853 0.782 0.894 0.279 0.773 0.773 0.772 0.773
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
P. Chen and S. Kim

Heliyon 9 (2023) e13916
8
that digital transformation has a greater impact on innovation in non-HPE (β =0.097, p <0.001) than in HPE (β =0.086, p <0.001).
4.6. Discussion
This study makes several contributions to the literature on digital transformation and innovation. First, based on the information
processing view and organizational boundaries, it is found that digital transformation can signicantly improve the innovation ca-
pabilities (both quantity and quality). This is in accordance with the research of [8], who found that digital transformation promotes
rm innovation. Digital transformation optimizes rm resource allocation, strengthens inter-rm linkages, blurs boundaries between
rms, and offsets R&D costs.
From a factor mobility perspective, it is found that innovation factors play a signicant role in the relationship between digital
transformation and innovation. First, digital transformation can alleviate barriers to knowledge ow. The proliferation of digital
technologies accelerates information processing and ow, thus accelerating knowledge ow [59]. Digital technologies reduce the cost
of knowledge storage and make it easier to store information [60]. Second, digital transformation can attract an inux of skilled
people. Digital transformation may improve the work environment (safety, comfort, and efciency) and attract more technicians [61].
Third, digital transformation can facilitate R&D investments. It can reduce the waste of non-R&D time and achieve efciency in
innovation [62]. Finally, digital transformation can increase the innovation awareness of corporate managers. Digital technologies
help companies to understand how their peers are innovating. Managers are forced to innovate to maintain market position [62].
The mediating role of digital transformation is further explored. At the innovation quantity level, the mediating effect of innovation
awareness is greater than the mediating effect of R&D investment, the mediating effect of technicians, as well as the mediating effect of
knowledge ow. At the level of innovation quality, the mediating effect of technicians is greater than the mediating effect of innovation
awareness, the mediating effect of R&D investment, as well as the mediating effect of knowledge ow. The technicians are considered
as carriers of knowledge and innovation, who can contribute more original ideas and methods to improve the quality of innovation
[63]. In addition, the mediating effects of knowledge ows, technicians and R&D investments on innovation quality are all greater
than those on innovation quantity. Innovation quality implies technological breakthroughs and original innovations [64], which
require a more balanced and well-developed R&D system. This implies that rms need to invest more in technical staff, knowledge
ows and R&D investments [65].
Finally, heterogeneity analysis shows that digital transformation can mitigate imbalances among rms. This is consistent with the
ndings of [45], who found that digital transformation can break monopolies based on rm size. SOEs, HTEs, and HPEs have unique
advantages such as political advantages and policy preferences. Non-SOEs, non-HTEs, and non-HPEs can use digital technology to
overcome social disadvantages and attract social capital inows to compensate for their shortcomings.
Overall, this study expands the denition of organizational boundaries and information processing views in digital transformation
research. With the rapid development of digital technology, more and more companies are using digital technology to improve their
Table 8
Description of each effect.
Paths Indirect effects Direct effects Total effects IR Sobel test Bootstrap test
DT → KF → Pat 0.002*** 0.093*** 0.095*** 0.021 0.002* 0.002**
DT → TE → Pat 0.004*** 0.090*** 0.094*** 0.043 0.004*** 0.004***
DT → RD → Pat 0.006*** 0.071*** 0.077*** 0.078 0.006*** 0.006***
DT → AI → Pat 0.019*** 0.076*** 0.095*** 0.200 0.019*** 0.019***
DT → KF → Inpat 0.003** 0.067*** 0.070*** 0.043 0.003** 0.003**
DT → TE → Inpat 0.007*** 0.061*** 0.068*** 0.103 0.007*** 0.007***
DT → RD → Inpat 0.004*** 0.047*** 0.051*** 0.078 0.04*** 0.04***
DT → AI → Inpat 0.006* 0.064*** 0.070*** 0.086 0.006* 0.006*
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1. Bootstrap test shows
only indirect effect, and TR is the ratio of indirect effects to total effects.
Table 9
Heterogeneity analysis.
(1) (2) (3) (4) (5) (6)
Variables SOE Non-SOE HTE Non-HTE HPE Non-HPE
DT 0.082*** 0.103*** 0.072*** 0.126*** 0.086*** 0.097***
(0.017) (0.013) (0.014) (0.018) (0.017) (0.013)
C −3.145*** −2.439*** −2.910*** −2.441*** −2.979*** −2.315***
(0.422) (0.271) (0.315) (0.421) (0.398) (0.271)
Control Yes Yes Yes Yes Yes Yes
Firm Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
Obs 5980 10,171 9606 6545 6384 9767
R-sq 0.791 0.756 0.779 0.812 0.690 0.788
Note: *** indicates signicance at the p <0.01, ** indicates signicance at the p <0.05, * indicates signicance at the p <0.1.
P. Chen and S. Kim

Heliyon 9 (2023) e13916
9
innovation capabilities. Innovation factors such as knowledge mobility, talent mobility, R&D investment and innovation awareness
can all contribute to rm innovation under the context of digital transformation. Finally, it is found that digital transformation has rm
heterogeneous for rm innovation, which provides new empirical and theoretical perspectives for emerging developing countries such
as China to promote Industry 4.0 and innovation.
5. Conclusions, suggestions and limitations
For the rst time, the relationship between digital transformation and innovation is studied from the perspective of innovation
factors. The impact of digital transformation on rms’ innovation performance and its mechanism of action were investigated by using
Chinese listed rms data between 2009 and 2019. The conclusions are as follows. (1) The corporate digital transformation was
measured through based on textual analysis methods and it was found that digital transformation can promote corporate innovation.
(2) Knowledge ow, technical personnel, R&D investment, and innovation awareness are important mediating paths. (3) In the
innovation quantity dimension, the mediating role of innovation awareness is greater. And in the innovation quality dimension, the
mediating role of technicians is greater. (4) Digital transformation has a greater impact on innovation of non-SOEs, non-high-tech
enterprises and non-heavily polluting enterprises, alleviating the gap between different types of rms. The results of this paper
alleviate the concerns of digital transformation in developing countries such as China and provide experiences and evidence for them
to promote Industry 4.0 and sustainable innovation.
Based on the analysis results, the following policy recommendations are proposed. For the government. First, government should
strengthen digital technology infrastructure to accelerate the digital transformation. Second, the government should establish and
improve policies to offset the cost burden of digitalization for enterprises. Third, the government should improve knowledge protection
regulations and develop active talent admission policies and R&D subsidy mechanisms to facilitate the ow of innovation factors. For
enterprise managers. First, enterprise managers should recognize the advantages of digital transformation and actively adopt digital
transformation strategies and transit positively to Industry 4.0 stage so as to achieve sustainable innovation. Second, managers should
break organizational boundaries, increase communication with other organizations, appropriately adjust the treatment of technical
staff, increase innovation-related training, and break the inherent innovation mindset to achieve sustainable innovation. Finally,
enterprise managers should fully realize that different enterprises have different attitudes toward digital transformation. Namely,
managers of non-state-owned enterprises, non-high-tech enterprises and other enterprises that are at a disadvantage in the economic
market should adopt digital transformation strategies, since digital transformation can reduce the gap between themselves and giant
companies that occupy a dominant position.
This paper still has some limitations. First, this paper uses only four variables to measure innovation factors; in other studies [66],
culture is also considered as an innovation factor. Second, digital transformation may be inuenced by government intervention and
market-driven aspects, which are not considered in this paper. Third, the long-term effects of digital transformation have been
neglected. Finally, innovation factors co-exist, it is difcult to analyze the relationship between them. In future research, we will deeply
explore the impact of digital transformation on innovation.
Author contribution statement
PENGYU CHEN: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data;
Contributed reagents, materials, analysis tools or data; Wrote the paper.
SangKyum Kim: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials,
analysis tools or data; Wrote the paper.
Funding statement
This research did not receive any specic grant from funding agencies in the public, commercial, or not-for-prot sectors.
Data availability statement
Data will be made available on request.
Declaration of interest’s statement
The authors declare no conict of interest.
References
[1] R. Morrar, H. Arman, S. Mousa, The fourth industrial revolution (Industry 4.0): a social innovation perspective, Technology innovation management review 7
(11) (2017) 12–20.
[2] G. Li, Y. Hou, A. Wu, Fourth Industrial Revolution: technological drivers, impacts and coping methods, Chin. Geogr. Sci. 27 (4) (2017) 626–637.
[3] S. Mittal, M.A. Khan, D. Romero, T. Wuest, A critical review of smart manufacturing & Industry 4.0 maturity models: implications for small and medium-sized
enterprises (SMEs), J. Manuf. Syst. 49 (2018) 194–214.
[4] J. Loonam, S. Eaves, V. Kumar, G. Parry, Towards digital transformation: lessons learned from traditional organizations, Strat. Change 27 (2) (2018) 101–109.
P. Chen and S. Kim

Heliyon 9 (2023) e13916
10
[5] T. Tolstykh, L. Gamidullaeva, N. Shmeleva, Elaboration of a mechanism for sustainable enterprise development in innovation ecosystems, J. Open Innov. :
Techn. Mark. Comp. 6 (4) (2020) 95.
[6] H. Gimpel, M. R¨
oglinger, Digital Transformation: Changes and Chances–Insights Based on an Empirical Study, 2015.
[7] G. Zeng, L. Lei, Digital transformation and corporate total factor productivity: empirical evidence based on listed enterprises, in: Discrete Dynamics in Nature
and Society, 2021, 2021.
[8] R. Li, J. Rao, L. Wan, The digital economy, enterprise digital transformation, and enterprise innovation, Manag. Decis. Econ. (2022).
[9] S. Nambisan, M. Wright, M. Feldman, The digital transformation of innovation and entrepreneurship: progress, challenges and key themes, Res. Pol. 48 (8)
(2019), 103773.
[10] F.P. Appio, F. Frattini, A.M. Petruzzelli, P. Neirotti, Digital transformation and innovation management: a synthesis of existing research and an agenda for future
studies, J. Prod. Innovat. Manag. 38 (1) (2021) 4–20.
[11] S. Ren, A.B. Eisingerich, H.T. Tsai, Search scope and innovation performance of emerging-market rms, J. Bus. Res. 68 (1) (2015) 102–108.
[12] A. Hinterhuber, Digital transformation, the Holy Grail, and the disruption of business models: an interview with Michael Nilles, Bus. Horiz. 65 (3) (2022)
261–265.
[13] F.M. Santos, K.M. Eisenhardt, Organizational boundaries and theories of organization, Organ. Sci. 16 (5) (2005) 491–508.
[14] D.M. Gann, A.J. Salter, Innovation in project-based, service-enhanced rms: the construction of complex products and systems, Res. Pol. 29 (7–8) (2000)
955–972.
[15] F. Karlsson, M. Frostenson, F. Prenkert, E. Kolkowska, S. Helin, Inter-organisational information sharing in the public sector: a longitudinal case study on the
reshaping of success factors, Govern. Inf. Q. 34 (4) (2017) 567–577.
[16] Y. Wan, Y. Gao, Y. Hu, Blockchain application and collaborative innovation in the manufacturing industry: based on the perspective of social trust, Technol.
Forecast. Soc. Change 177 (2022), 121540.
[17] J. Winkler, C.P.J.W. Kuklinski, R. Moser, Decision making in emerging markets: the Delphi approach’s contribution to coping with uncertainty and equivocality,
J. Bus. Res. 68 (5) (2015) 1118–1126.
[18] M. Hilbert, Digital technology and social change: the digital transformation of society from a historical perspective, Dialogues Clin. Neurosci. (2022).
[19] F.J.N. Do Adro, J.C.C. Leit˜
ao, Leadership and organizational innovation in the third sector: a systematic literature review, Int. J. Innov. Stud. 4 (2) (2020) 51–67.
[20] P.A. Pavlou, O.A. El Sawy, The “third hand”: IT-enabled competitive advantage in turbulence through improvisational capabilities, Inf. Syst. Res. 21 (3) (2010)
443–471.
[21] A. Carvalho, H. Alves, J. Leit˜
ao, What research tells us about leadership styles, digital transformation and performance in state higher education? Int. J. Educ.
Manag. (2022).
[22] J. Kroh, H. Luetjen, D. Globocnik, C. Schultz, Use and efcacy of information technology in innovation processes: the specic role of servitization, J. Prod.
Innovat. Manag. 35 (5) (2018) 720–741.
[23] T.J. Saldanha, S. Mithas, M.S. Krishnan, Leveraging customer involvement for fueling innovation: the role of relational and analytical information processing
capabilities, MIS Q. 41 (1) (2017).
[24] R.B.B.D. Vasconcelos, J.F.D. Santos, J.A.D. Andrade, Innovation in Micro and Small Enterprises: Resources and Capabilities vol. 25, Revista de Administraç˜
ao
Contemporˆ
anea, 2020.
[25] M. Meyer, Tracing knowledge ows in innovation systems, Scientometrics 54 (2) (2002) 193–212.
[26] M. Smith, M. Busi, P. Ball, R. Van Der Meer, Factors inuencing an organisation’s ability to manage innovation: a structured literature review and conceptual
model, in: Managing Innovation: What Do We Know About Innovation Success Factors?, 2019, pp. 69–90.
[27] R. Moser, C.P.J.W. Kuklinski, M. Srivastava, Information processing t in the context of emerging markets: an analysis of foreign SBUs in China, J. Bus. Res. 70
(2017) 234–247.
[28] X. Fu, Digital transformation of global value chains and sustainable post-pandemic recovery, Transnat. Corp. J. 27 (2) (2020).
[29] J. Mills, K. Platts, M. Bourne, Applying resource-based theory: methods, outcomes and utility for managers, Int. J. Oper. Prod. Manag. (2003).
[30] C. Forman, N. van Zeebroeck, Digital technology adoption and knowledge ows within rms: can the Internet overcome geographic and technological distance?
Res. Pol. 48 (8) (2019), 103697.
[31] C. Paunov, V. Rollo, Has the internet fostered inclusive innovation in the developing world? World Dev. 78 (2016) 587–609.
[32] K.S. Choi, B.H. Hyun, In the era of digital transformation: the effect of government support, network capability and knowledge sharing on innovation
performance through innovative behavior, J. Dig. Converg. 20 (4) (2022) 353–366.
[33] S. Hajkowicz, A. Reeson, L. Rudd, A. Bratanova, L. Hodgers, C. Mason, N. Boughen, Tomorrow’s Digitally Enabled Workforce: Megatrends and Scenarios for
Jobs and Employment in Australia over the Coming Twenty Years, 2016.
[34] T. Schwarzmüller, P. Brosi, D. Duman, I.M. Welpe, How does the digital transformation affect organizations? Key themes of change in work design and
leadership, Manag. Rev. 29 (2) (2018) 114–138.
[35] L. Ma, X. Zhai, W. Zhong, Z.X. Zhang, Deploying human capital for innovation: a study of multi-country manufacturing rms, Int. J. Prod. Econ. 208 (2019)
241–253.
[36] D. Liang, Aesthetic value evaluation for digital cultural and creative products with articial intelligence, in: Wireless Communications and Mobile Computing,
2022, 2022.
[37] F. Bertani, M. Raberto, A. Teglio, The productivity and unemployment effects of the digital transformation: an empirical and modelling assessment, Rev. Evolut.
Polit. Econ. 1 (3) (2020) 329–355.
[38] M. Fitzgerald, N. Kruschwitz, D. Bonnet, M. Welch, Embracing digital technology: a new strategic imperative, MIT Sloan Manag. Rev. 55 (2) (2014) 1.
[39] J. Zhang, J. Long, A.M.E. von Schaewen, How does digital transformation improve organizational resilience?—ndings from PLS-SEM and fsQCA, Sustainability
13 (20) (2021), 11487.
[40] G. Geser, E.M. Hollauf, V. Hornung-Pr¨
ahauser, S. Sch¨
on, F. Vloet, Makerspaces as social innovation and entrepreneurship learning environments: the DOIT
learning program, Discour. Commun. Sustain. Educ. 10 (2) (2019) 60–71.
[41] K. Kim, Moderating effects of legitimacy on the liability-and R&D investment-innovation relationships in manufacturing SMEs, Asian J. Technol. Innovat. 27 (1)
(2019) 23–45.
[42] C. Ma, Z. Liu, Effects of M&As on innovation performance: empirical evidence from Chinese listed manufacturing enterprises, Technol. Anal. Strat. Manag. 29
(8) (2017) 960–972.
[43] Z. Chen, J. Zhang, Y. Zi, A cost-benet analysis of R&D and patents: rm-level evidence from China, Eur. Econ. Rev. 133 (2021), 103633.
[44] M.´
A. Galindo-Martín, M.S. Casta ˜
no-Martínez, M.T. M´
endez-Picazo, Digital transformation, digital dividends and entrepreneurship: a quantitative analysis,
J. Bus. Res. 101 (2019) 522–527.
[45] X. Guo, X. Song, B. Dou, A. Wang, H. Hu, Can digital transformation of the enterprise break the monopoly? Personal Ubiquitous Comput. (2022) 1–14.
[46] R.J. Rajapathirana, Y. Hui, Relationship between innovation capability, innovation type, and rm performance, J. Innov. Knowled. 3 (1) (2018) 44–55.
[47] P. Sharma, R.C. Tripathi, Patent citation: a technique for measuring the knowledge ow of information and innovation, World Patent Inf. 51 (2017) 31–42.
[48] F. Neffke, M. Henning, Skill relatedness and rm diversication, Strat. Manag. J. 34 (3) (2013) 297–316.
[49] M.E. Bontempi, Investment–uncertainty relationship: differences between intangible and physical capital, Econ. Innovat. N. Technol. 25 (3) (2016) 240–268.
[50] S. Chen, M. Bu, S. Wu, X. Liang, How does TMT attention to innovation of Chinese rms inuence rm innovation activities? A study on the moderating role of
corporate governance, J. Bus. Res. 68 (5) (2015) 1127–1135.
[51] S. Liu, J. Du, W. Zhang, X. Tian, G. Kou, Innovation quantity or quality? The role of political connections, Emerg. Mark. Rev. 48 (2021), 100819.
[52] D. Xia, X.Q. Wang, The synergetic impact of environmental and innovation information disclosure on corporate nancial performance: an empirical study based
on China coal listed companies, Technovation 100 (2021), 102179.
P. Chen and S. Kim

Heliyon 9 (2023) e13916
11
[53] B.A. Cohen, R.D. Mitra, J.D. Hughes, G.M. Church, A computational analysis of whole-genome expression data reveals chromosomal domains of gene
expression, Nat. Genet. 26 (2) (2000) 183–186.
[54] V. Cristofori, Y. Rouphael, E. Mendoza-de Gyves, C. Bignami, A simple model for estimating leaf area of hazelnut from linear measurements, Sci. Hortic. 113 (2)
(2007) 221–225.
[55] K. Dengler, K. Hiesinger, A. Tisch, Digital Transformation: the role of computer use, in: Employee Health, Economics & Human Biology, 2022, 101137.
[56] I. Andrews, J.H. Stock, L. Sun, Weak instruments in instrumental variables regression: theory and practice, Ann. Rev. Econom. 11 (1) (2019).
[57] X. Zhu, H. Zhao, Recognition of innovation and diffusion of welfare policy: alleviating urban poverty in Chinese cities during scal recentralization, Governance
31 (4) (2018) 721–739.
[58] G.W. Cheung, R.S. Lau, Testing mediation and suppression effects of latent variables: bootstrapping with structural equation models, Organ. Res. Methods 11 (2)
(2008) 296–325.
[59] M. Sharma, S. Luthra, S. Joshi, A. Kumar, Accelerating retail supply chain performance against pandemic disruption: adopting resilient strategies to mitigate the
long-term effects, J. Enterprise Inf. Manag. (2021).
[60] A. Goldfarb, C. Tucker, Digital economics, J. Econ. Lit. 57 (1) (2019) 3–43.
[61] M. Wolf, A. Semm, C. Erfurth, Digital transformation in companies–challenges and success factors, in: International Conference on Innovations for Community
Services, Springer, Cham, 2018, June, pp. 178–193.
[62] A. Urbinati, D. Chiaroni, V. Chiesa, F. Frattini, The role of digital technologies in open innovation processes: an exploratory multiple case study analysis, R D
Manag. 50 (1) (2020) 136–160.
[63] A. Carneiro, How does knowledge management inuence innovation and competitiveness? J. Knowl. Manag. 4 (2) (2000) 87–98.
[64] G.C. O’Connor, R.W. Veryzer, The nature of market visioning for technology-based radical innovation, J. Prod. Innovat. Manag.: Int. Pub. Prod. Develop. Manag.
Assoc. 18 (4) (2001) 231–246.
[65] K.H. Tsai, M.H. Hsieh, E.J. Hultink, External technology acquisition and product innovativeness: the moderating roles of R&D investment and congurational
context, J. Eng. Technol. Manag. 28 (3) (2011) 184–200.
[66] M.Y. Brannen, Culture as the critical factor in implementing innovation, Bus. Horiz. 34 (6) (1991) 59–67.
P. Chen and S. Kim