The Impact of Digital Marketing Capability on Firm Performance: Empirical Evidence from Chinese Listed Manufacturing Firms

Abstract

The rapid expansion of e-commerce has pushed firms to adopt more sophisticated digital marketing strategies to reach, engage, and retain consumers. Research has shown that digital marketing significantly enhances firm performance by enhancing marketing-related capabilities, yet overlooks its role in driving transformation across other business functions. Grounded in resource orchestration theory, this study examines how digital marketing resources and capabilities support broader business transformation and comprehensively improve firm performance. Drawing on empirical data from Chinese A-share listed manufacturing firms from 2010 to 2023, this study demonstrates that there is a significant positive relationship between digital marketing capability and firm performance. Notably, this relationship is mediated by production capability and R&D capability. Moreover, the effect is more pronounced in firms operating in highly marketized regions, within competitive industries, and among digitally advanced firms. This study contributes to the digital marketing literature by developing a novel framework for measuring digital marketing capability, and uncovering the mechanisms through which it influences firm performance. In addition, this study contributes to the digitalization literature in the manufacturing sector by demonstrating the strategic role of digital marketing in driving value creation. Implications for digital marketing in manufacturing industry are discussed.

1. Introduction

The emergence of the Internet, along with a wave of new digital technologies—such as information and communication technology (ICT), artificial intelligence (AI) [1,2], cloud computing, big data—is profoundly transforming the global economy and society [3]. These technologies drive the transformation of traditional industries and the development of innovative business paradigms [4]. Furthermore, the COVID-19 pandemic accelerated digital transformation [5], as evidenced by both increased digital consumption among customers [6] and the expansion of online sales activities by firms. Against this backdrop, global e-commerce is undergoing rapid expansion. According to the latest eMarketer report, global e-commerce sales reached $6 trillion in 2024, marking a nearly 20% year-on-year increase [7]. This explosive growth is pushing firms to accelerate digitalization efforts in order to remain competitive [8], particularly by transforming their marketing strategies. In this context, digital marketing (DM) has emerged as a critical lever for engaging customers, optimizing campaigns, and driving performance [9].

As an emerging and rapidly evolving practice, digital marketing has garnered increasing attention from both practitioners and academic researchers. As defined by the American Marketing Association (2021), digital marketing refers to “any marketing methods conducted through electronic devices” [10]. This encompasses a wide range of tools and platforms, including websites, search engines, blogs, social media, video, and email, which businesses use to interact with and influence customers. Prior studies have demonstrated that the adoption of digital marketing practices can generate substantial benefits for firms, including enhanced customer satisfaction [11], higher conversion rates [9], greater brand awareness [12], and increased sales performance [13].

Firm performance is a commonly used indicator to evaluate a firm’s operational efficiency and overall success [14], typically assessed through measures of profitability and asset utilization [15,16]. Empirical studies have shown that digital marketing plays a crucial role in business operations, contributing significantly to firm performance [17], as evidenced by improvements in return on assets (ROA) [18], Tobin’s Q [19], revenue [20], return on investment (ROI) [21,22]. Prior studies have highlighted the mediating role of marketing-related capabilities—particularly those related to market and customer [2]—in linking digital marketing to firm performance [19]. Market-related capabilities, including market sensing [23] and market agility [21], enable firms to better anticipate and respond to market changes. Meanwhile, customer-related capabilities, such as customer linkage [23] and customer relationship management capability [24], strengthen customer engagement and retention, further reinforcing the performance benefits of digital marketing. Building on the above analysis, prior research has primarily focused on the “digital marketing → marketing-related capabilities → firm performance” pathway. However, a critical question remains unaddressed: Can digital marketing, beyond its traditional marketing role, act as a catalyst for transformation across other business functions, thereby enhancing overall firm performance? This study aims to systematically investigate this question.

Existing research on digital marketing has predominantly focused on service industries, such as retail [25], financial services [26], and hospitality and tourism [27,28]. In contrast, manufacturing firms have received comparatively less scholarly attention [29]. This gap partly stems from the traditional focus of manufacturing firms on internal operations, particularly production and assembly activities [30]. Consequently, the digitalization literature on manufacturing firms has mainly concentrated on Industry 4.0 technologies in the production process [31,32], including intelligent manufacturing [33], smart manufacturing [34], flexible manufacturing [35], and advanced manufacturing [36]. However, marketing plays a critical role in enabling high-value creation within the manufacturing sector. The well-known “Smiling Curve” from industrial economics illustrates that value tends to concentrate at the upstream (R&D, design) and downstream (marketing, branding, and services) ends of the value chain, where profit margins are significantly higher [37]. Moreover, digital marketing contributes to firm performance through different mechanisms across industries: in services, it primarily drives revenue by increasing customer spending through personalization and cross-selling [38], whereas in manufacturing, it focuses on cost efficiency by streamlining sales channels, optimizing supply chains, and improving operational productivity [39]. This distinction also reflects the different customer bases and market structures of the two sectors. While digital marketing in services primarily targets individual consumers (B2C), manufacturing firms typically operate within complex industrial value chains. Their customers are often other businesses in the ecosystem, such as upstream suppliers or downstream distributors. Consequently, digital marketing in manufacturing not only enhances firm-level performance but also promotes ecosystem-wide digital integration and capability upgrading [40]. Against this backdrop, this study investigates the role of digital marketing in enhancing firm performance within the manufacturing context.

Rooted in the capability-based view [41], resource orchestration theory emphasizes the pivotal role of organizational capabilities in coordinating resources to create value and sustain competitive advantage [42,43]. Building on this perspective, scholars widely recognize digital marketing capability (DMC) as a key enabler for executing marketing initiatives and achieving strategic goals, making it a foundational asset in the digital era [44]. This study defines digital marketing capability (DMC) as a firm’s ability to integrate and leverage digital resources to extract actionable customer insights, optimize marketing activities, and generate business value. Specifically, data and technology resources enable firms to transform raw data into marketing intelligence, thereby enhancing decision-making and reinforcing competitive advantage [45,46]. Resource orchestration theory highlights the critical role of capability deployment and alignment in driving firm performance [47]. In the manufacturing context, R&D and production constitute the two most knowledge- and asset-intensive functions, directly transforming customer insights into product innovation and operational execution. DMC enhances these functions by embedding real-time market intelligence into R&D, accelerating innovation cycles and improving product-market fit [48]. Simultaneously, DMC supports data-driven production planning, enabling firms to improve efficiency and responsiveness revenue by increasing customer spending through personalization and cross-selling. Compared with other organizational functions—such as procurement or logistics—R&D and production are more central to cross-functional value creation in manufacturing.

This study thus examines DMC from a capability-based perspective and seeks to address the following research questions: (1) Can digital marketing capability enhance firm performance in the manufacturing industry? (2) Through what mechanisms does digital marketing capability influence firm performance? (3) What contextual factors moderate the relationship between digital marketing capability and firm performance? Chinese manufacturing sector provides an ideal context for examining DMC. As the world’s largest manufacturing nation for fifteen consecutive years, China is experiencing a dual transformation: shifting from low- to high-value production while rapidly adopting digital technologies. This process is further accelerated by state-led initiatives such as “Made in China 2025”, which explicitly promote the integration of digital capabilities into manufacturing. Against this backdrop, Chinese manufacturing firms offer a timely and relevant setting to investigate the mechanisms and impacts of DMC.

The remainder of this paper is structured as follows. Section 2 reviews the theoretical background and relevant literature. Section 3 develops the research hypotheses and proposes the research model. Section 4 outlines the research design. Section 5 reports the empirical analysis and its results. Section 6 concludes the paper with a summary of theoretical contributions, practical implications, and future research directions.

2. Theoretical Foundation and Literature Review

2.1. Theoretical Foundation: Resource Orchestration Theory

Resource orchestration theory integrates the resource-based view and the capability-based view, proposing that a firm’s sustained competitive advantage stems from the strategic combination of its resources and capabilities. The theory emphasizes that dynamic management of resources is crucial for value creation. Firms build competitive advantages by evolving their resources, developing capabilities, and effectively leveraging them to achieve operational success [47]. It highlights the interdependent relationship between resources, capabilities, and value creation, offering a comprehensive framework for how firms can strategically deploy resources to drive value and sustain a competitive edge.

Resource orchestration theory consists of three core subprocesses: (1) Structuring the resource portfolio, which involves acquiring valuable resources and discarding unnecessary ones, thus creating a resource pool crucial for the firm’s growth. (2) Bundling resources, where firms integrate resources to strengthen their capabilities. (3) Leveraging capabilities, in which firms combine their resource bundles with capabilities to fully drive value creation [42,43]. These subprocesses illustrate how firms create and sustain competitive advantages by dynamically managing resources and capabilities. In this process, resources form the foundation for competitive advantage, while capabilities—developed from these resources—enable firms to deploy and utilize them effectively, achieving operational success and strategic objectives.

This theory also provides valuable insights into the relationship between digital marketing capability and firm performance. By leveraging digital marketing resources, firms lay the foundation for their digital marketing capability. This capability, in turn, serves as key enablers of value creation, helping firms achieve growth and performance goals. Thus, resource orchestration theory is essential for understanding the mechanisms linking digital marketing resources, capabilities, and firm performance.

2.2. Definition and Measurement of Digital Marketing Capability

Recent studies have increasingly explored the concept of digital marketing capability, offering various definitions that reflect the evolving understanding of this construct. For instance, Herhausen et al. (2020) define digital marketing capability as the ability to “perform a coordinated set of digital-related tasks, utilizing digital resources, for achieving a competitive advantage” [49]. Wang (2020) emphasizes digital marketing capability that helps firms benefit from digitalization [8]. Masrianto et al. (2022) describe it as the ability of a firm to use information technologies to facilitate deep interactions with customers [44], while Homburg et al. (2022) focus on digital tools and processes used to engage customers and partners [18]. Apasrawirote et al. (2022) further define it as a technology-enabled capability to access and utilize customer data to generate marketing value [50]. Overall, these definitions reflect a shared understanding of digital marketing capability as the firm’s capacity to deploy digital technologies and data-driven processes to build and manage customer relationships for long-term business success.

Drawing from diverse theoretical perspectives, scholars have proposed various approaches to operationalize digital marketing capability. For instance, Wang (2020) defines digital marketing capability as the relational competencies that strengthen connections with customers, suppliers, and channel partners [8]. Accordingly, five measurement dimensions are proposed: customer-linking digital capability, market-sensing digital capability, channel-bonding digital capability, the capability to create durable relationships with suppliers through digital platforms, and the capability to use digital marketing to retain customers [8]. From the technology utilization perspective, Homburg et al. (2022) conceptualize digital marketing capability as an integrated use of tools such as social media marketing, mobile marketing, content marketing, search engine marketing, web analytics, marketing automation, and email marketing [18]. Similarly, Apasrawirote et al. (2022) identify four core components of digital marketing capability: social media marketing capability, digital marketing strategy, digital relationships, and leadership capability [50].

Although some definitions of digital marketing capability emphasize the importance of data access and utilization [50], these aspects are often underrepresented in existing measurement frameworks. To address this gap, the present study proposes a new measurement approach that explicitly incorporates the data utilization perspective.

2.3. Research on the Impact of Digital Marketing on Firm Performance

Since the concept of digital marketing emerged, scholars have extensively studied its impact on firm performance. Representative studies by Purba et al. (2021) and Gharios et al. (2024) have found a significant association between digital marketing and firm performance, such as profitability [17,26]. Similarly, empirical research by Wu et al. (2024) and Verma et al. (2024) demonstrates that the adoption of digital marketing or digital marketing strategies leads to notable improvements in firm performance, including metrics like profits, ROI, and other financial metrics [22,51]. Moreover, research by Jung et al. (2023) and Erhan et al. (2024) validates the positive impact of digital marketing innovation on firm performance, measured through indicators such as Tobin’s Q [19], ROA, and revenue [20].

Among these studies, a growing body of research has started to focus specifically on the role of digital marketing capability (DMC). For instance, Liu (2022), Homburg et al. (2022), and Chinakidzwa et al. (2022) consistently report that DMC has a positive effect on firm performance, particularly in enhancing profitability [13,18,23]. These findings highlight DMC as a critical enabler of firm performance in the digital era.

2.4. Cross-Functional Mechanisms of Digital Marketing

While prior studies predominantly investigate the direct effects of digital marketing on firm performance, research on the underlying mechanisms has mainly centered on marketing-related variables. Digital marketing primarily operates in customer-facing and market-related domains, where it reshapes marketing processes and alters consumer behavior [2].For example, Liu (2022) suggests that digital marketing positively influences firm performance by enhancing market sensing capability and customer-linking capability [23]. Yusuf et al. (2022) find that digital marketing mix strategies significantly promote performance indicators, such as profits, by improving customer relationship management capabilities [24]. Hossain et al. (2022) highlight that marketing analytics capability enhance financial performance by strengthening market sensing, market grasping, market reconfiguration, and dynamic marketing capabilities [52]. Jung et al. (2023) demonstrate that digital marketing innovation improves firm performance through marketing capability, with the indirect effect being stronger than the direct effect [19].

However, emerging studies in the broader digitalization literature suggest that digital technologies also reshape non-marketing functions, such as R&D and production. In the R&D domain, digital transformation has been found to enhance both explorative and exploitative innovation capabilities by improving knowledge recombination, accelerating idea generation, and shortening development cycles [48,53]. In production, advanced technologies—including digital twins, demand forecasting, and supply chain analytics—enable real-time monitoring, predictive maintenance, and agile production planning. These advancements not only improve operational efficiency and reduce costs, but also enhance production flexibility and responsiveness to market dynamics [54,55].

Building on these insights, this study extends the scope of digital marketing research by exploring whether DMC can similarly drive firm performance through its influence on core operational functions—namely, R&D and production. Focusing on manufacturing firms, which encompass key business functions such as R&D, procurement, production, marketing, and service, this study seeks to clarify how DMC affects firm performance through cross-functional pathways. In doing so, it enriches the existing framework of digital marketing research by emphasizing the value creation mechanisms specific to manufacturing contexts.

3. Hypothesis Development

3.1. Digital Marketing Capability and Firm Performance

Digital marketing capability is generally regarded as a firm’s ability to plan, implement, and manage digital marketing activities and processes [44]. Building on Krishen et al.’s (2021) definition, which highlights the pivotal role of data in digital marketing practices [56], this study defines digital marketing capability as a firm’s core ability to leverage data to drive digital and intelligent transformation in marketing activities, thereby generating business value. In the digital era, data has become a strategic asset for firms—often described as the ‘new oil’ powering value creation [57]. Specifically, marketing data enables firms to extract actionable insights, build predictive models [58], and ultimately support data-driven decisions.

In traditional business management, firms have already leveraged market information to improve product and service offerings. However, limited technology development has constrained the scale and depth of information utilization. In the digital era, advancements in digital technologies have led to explosive data growth and enhanced accessibility to data sources for firms, breaking the traditional boundaries [45]. Digital marketing capability allows firms to systematically process market data and customer data, transforming raw information into actionable intelligence. Regarding market data, firms can collect large-scale market intelligence, filter and analyze relevant information, and derive deeper marketing insights. This enhances market sensing, improves demand forecasting, and enables precise product positioning, thereby increasing decision-making accuracy [59] and reducing uncertainty [52]. Meanwhile, customer data allows firms to track consumer behavior and predict preferences [58]. Prior research indicates that leveraging customer data supports precision marketing, new product development, and strategic realignment, all of which improve customer experience and business performance [60]. By integrating both market and customer data, firms with digital marketing capability can establish stronger digital linkages with external environments, optimize operations, and make data-driven decisions. This leads to greater supply chain efficiency [61], enhanced customer satisfaction [60], revenue growth, and cost reductions [52], ultimately strengthening firm performance and competitive advantage [62]. Therefore, we hypothesize the following:

Hypothesis 1 (H1).

Digital marketing capability can effectively improve firm performance.

3.2. The Mediating Role of R&D Capability

Traditional firms have primarily relied on internal knowledge and expert opinion for research and development (R&D), largely drawing on organizational and individual experience. However, this closed and insular approach often falls short in responding to diverse market demands and uncertainty. Successful new product development (NPD) requires cross-functional information integration. In the digital era, the relationship between new product development and market has attracted attention [63], and firms gain deeper insights into markets and customers to drive innovation [60]. Digital marketing capability plays a crucial role in enabling data-driven R&D, which leverages data analytics to extract meaningful value and facilitate new product development [64]. By analyzing market data, firms can gain a clearer understanding of external dynamics. This includes tracking customer preferences [65], monitoring competitors, and evaluating distribution channels. With these insights, firms can identify emerging opportunities and adjust R&D strategies accordingly. A more adaptive R&D approach enables firms to develop superior value propositions and respond more effectively to market changes [66]. Meanwhile, customer data provides direct insights into consumer preferences and behaviors. Firms can use this data to distinguish between common and unique customer needs. This targeted understanding supports product design by offering concrete reference points for feature development. As a result, firms can create more tailored, customer-centric innovations that better meet market demands. Research on customer co-creation suggests that involving customers in NPD fosters greater innovation and enhances product success [67]. By leveraging digital marketing capability, firms can break down data silos between marketing and R&D departments, fostering a more integrated approach to innovation. R&D capability is critical in competitive markets, as it drives firm growth and long-term competitiveness [68]. Firms that continuously explore new products and refine existing ones can expand into new markets, generate additional revenue, and enhance overall performance [69]. Therefore, we propose the following hypothesis:

Hypothesis 2 (H2).

Digital marketing capability can increase R&D capability, which will have a contributing effect on firm performance.

3.3. The Mediating Role of Production Capability

For manufacturing firms, production is the core business function, and production capability is a crucial determinant of firm survival, growth, and competitive advantage. Traditionally, marketing and manufacturing operations have been perceived as distinct functions, with little alignment in goals or daily coordination [70]. However, with the advancement of digitalization, production processes have undergone significant transformation. Firms now recognize and prioritize the role of the demand side, paying greater attention to market dynamics and customer needs [71]. As a result, the connection between marketing and production activities has become increasingly integrated [72]. By leveraging digital marketing capability, manufacturing firms can continuously monitor market fluctuations and evolving customer preferences [73]. This enables them to collect real-time, data-driven insights and generate visualized, actionable predictions, which can then be communicated to the production team [74]. With these insights, firms can make informed production decisions, leading to better resource allocation and enhanced production agility [75]. This ability to swiftly adapt to market changes strengthens their responsiveness and competitiveness. Enhanced production capability brings two key advantages. First, from an operational perspective, it improves production efficiency and shortens manufacturing cycles [76]. This allows firms to operate with greater flexibility and speed [77], optimize capacity planning, minimize inventory buildup, and reduce operational costs. Second, from a market perspective, improved production capability enables firms to capture greater market share by responding promptly to customer demands [78]. Faster manufacturing cycles shorten delivery times, enhance customer satisfaction, and ultimately drive sales revenue, leading to improved firm performance [79]. Based on these arguments, we formulated the following hypothesis:

Hypothesis 3 (H3).

Digital marketing capability can improve production capability, thus contributing to firm performance.

Figure 1 illustrates the research model, integrating the hypotheses and key variables.

4. Research Design

4.1. Data Collection

The article selects China’s A-share listed firms from 2010 to 2023 as the initial sample. Data collection involves multiple sources: digital marketing capability data is derived from the annual reports of target firms, accessed via Cninfo (http://www.cninfo.com.cn (accessed on 12 July 2024)); data on firm performance and control variables are collected from the China Stock Market & Accounting Research (CSMAR) database (https://data.csmar.com (accessed on 17 July 2024)); regional-level data is obtained from the Statistical Yearbook of the National Bureau of Statistics of China.

To maintain the validity of the data, the following treatments are applied to the sample: (1) Removal of firms that are classified as ST, *ST, PT, or those that have been delisted; (2) Winsorization of continuous variables at the 1% and 99% percentiles to reduce the impact of outliers; (3) Retention of only those samples that have no missing data for at least three consecutive years. Ultimately, 24,779 valid samples are retained, forming an unbalanced panel for analysis. The analysis is then conducted using STATA 16.0 software.

4.2. Model

To examine the impact of digital marketing on firm performance, this study constructs the following two-way fixed effects model.

$${FP}_{it}={\alpha }_0+{\alpha }_1{DMC}_{it}+{\sum {\alpha }_{2}CV}_{s_{it}}+\lambda +\gamma +{\epsilon }_{it}$$

(1)

In model (1), the subscript i represents the firm, and t represents time. ${FP}_{it}$ represents the firm’s performance, ${DMC}_{it}$ represents the firm’s digital marketing capability, and ${CV}_{s_{it}}$ represents the control variables. Additionally, the model also controls for time fixed effects (λ) and industry fixed effects (γ), with ${\epsilon }_{it}$ representing the random error term. The coefficient ${\alpha }_1$ of ${DMC}_{it}$ is the focus of this study, reflecting the impact of digital marketing capability on firm performance. If ${\alpha }_1$ is significantly positive under the full sample, it suggests that digital marketing capability helps improve the firm performance.

4.3. Variables

4.3.1. Dependent Variable: Firm Performance

Following Guo et al. (2023) [79], this study measures the dependent variable, firm performance, using return on assets (ROA). As one of the most widely used indicators in firm performance research, ROA is closely linked to several other profitability metrics. It is calculated as the ratio of net profit to total assets, representing the profit a firm generates per unit of assets. ROA reflects a firm’s ability to utilize its assets effectively and serves as a key indicator of its performance.

4.3.2. Independent Variable: Digital Marketing Capability

Research in the field of organizational capabilities suggests that firms achieve strategic objectives by repeatedly coordinating and deploying resources [80]. In line with this view, digital marketing scholars emphasize that digital marketing capability reflects a firm’s ability to integrate and leverage digital resources to support marketing goals. Building on definitions that underscore the central role of data in digital marketing, the study proposes a measurement framework from the perspective of data resource utilization to more precisely capture this capability.

Notably, scholars working on data-driven research themes emphasize that merely possessing large volumes of data does not inherently generate value for firms [81]. Instead, value creation depends on the firm’s ability to transform raw data into actionable information and knowledge through analytics [46]. This transformation process not only relies on data as essential input, but also requires substantial support from technology resources. Prior research highlights that deriving meaningful insights involves leveraging a combination of technologies such as machine learning, data extraction and cleaning, statistical analysis, and programming expertise—all of which are critical to converting data into business intelligence [82]. As these technologies operate in tandem with data itself, data and technology resources function as complementary assets that jointly enable firms to transform raw data into actionable insights, thereby unlocking value and securing competitive advantage [45,46]. Because this transformation process lies at the heart of effective digital marketing, both types of resources are fundamental in shaping a firm’s digital marketing capability. Their synergy strengthens the multiplier effect of digital marketing by deepening managerial insight [83], generating business value [84], and ultimately enhancing firm performance [85].

Following recent studies on digital transformation measurement [48,86,87], this study uses keyword frequency analysis of corporate annual reports to quantify DMC. This method has been widely recognized in the literature for its objectivity and scalability. Unlike primary data from managerial surveys, annual reports are publicly disclosed documents with standardized formats, required by law for listed firms. As a result, this approach reduces subjective bias, ensures data comparability across firms and years, and supports large-scale longitudinal analyses. Based on the two resources mentioned above, this study first constructs an initial keyword dictionary. This is done by referencing important documents and research reports on digital marketing, reviewing authoritative literature on the subject, and extracting relevant keywords related to digital marketing from corporate annual reports. Various methods are employed to complement each other. Next, synonym expansion is performed to identify and extract semantically similar keywords. For instance, terms related to marketing data are expanded to include order data, transaction data, customer data, membership data, and other relevant terms. Additionally, expert consultation in the digital marketing field is conducted to avoid overlooking key terms. Based on expert feedback, the preliminary dictionary is evaluated and supplemented, leading to the final refined list of keywords related to digital marketing capability, as shown in Figure 2. On this basis, Python’s (v3.0) Counter class is used to perform a quantitative text analysis. This analysis examines the frequency of the given keywords, yielding 27,779 data points digital marketing capability for the period 2010–2023.

4.3.3. Mediating Variable

R&D capability. In existing literature, patents are widely recognized as a key indicator of a firm’s R&D capability [88]. Among them, patents can be classified into filed and granted patents. This study adopts granted patents, which better reflect a firm’s R&D strength and achievements, to measure the actual effectiveness of its R&D activities.

Production capability. TFP, widely used in previous studies, serves as a key indicator of firms’ production efficiency [89]. Its estimation methods include the Olley-Pakes (OP) method [90], Levinsohn-Petrin (LP) method [91], and Ackerberg-Caves-Frazer (ACF) method [92]. Following Lai et al. (2021), this study adopts the ACF approach to measure firms’ production capability [93].

4.3.4. Control Variables

In addition to the independent variables, other factors may also influence the dependent variable. To avoid omitted variable bias and ensure the reliability of the estimated coefficients, this study includes control variables related to firm characteristics, financial features, governance structure, and macroeconomic factors. Based on Yao et al. (2021) [15], the selection of control variables is described as follows.

Firm characteristic. SIZE is measured by the natural logarithm of the firm’s total assets. AGE is the natural logarithm of the firm’s age plus one.

Financial features. Leverage ratio (LEV) compares total liabilities to total assets. Growth rate of operating income (GROW) measures the increase in operating income for the current year compared to the total income of the previous year. Cash flow ratio (CASH) compares the year-end balance of cash and cash equivalents to current liabilities. Intangible assets ratio (INT) reflects the proportion of net intangible assets to total assets.

Governance structure. Board size (BS) refers to the total number of board members. The proportion of independent directors (IND) is the share of independent directors in the total board size. Stakeholder (STA) is determined by the background of the controlling shareholder, with state-owned firms coded as 1 and non-state-owned firms coded as 0.

Macroeconomics. GDP of region (GDP) is measured by the per capita value of regional GDP. Finance of region (FIN) is calculated as the ratio of total loans from regional financial institutions to regional GDP.

Description of key variables is shown in

Table 1. Description of key variables.

Type of Variables	Name of Variables	Symbol	Measurement	Source of Data
Dependent variable	Firm Performance	FP	ROA (=Net profit/total assets)	CSMAR database
Independent variable	Digital marketing capability	DMC	Logarithm of frequency (+1 adjustment) of digital marketing capability-related keywords in corporate annual reports	Corporate annual reports
Mediating variables	R&D capability	PAT	Granted patents	CSMAR database
	Production capability	TFP	Total factor productivity	CSMAR database
Control variables	Size	SIZE	Logarithm of total assets at year-end	CSMAR database
	Age	AGE	Logarithm of firm age (+1 adjustment)	CSMAR database
	Leverage ratio	LEV	Ratio of total liabilities to total assets	CSMAR database
	Growth rate of operating income	GROW	Ratio of current year operating income increase to previous year’s total operating income	CSMAR database
	Cash flow ratio	CASH	Ratio of year-end cash and cash equivalents to current liabilities	CSMAR database
	Intangible assets ratio	INT	Ratio of net intangible assets to total assets	CSMAR database
	Stake holder	STA	State-owned = 1, non-state-owned = 0	CSMAR database
	Board Size	BS	Number of board members	CSMAR database
	Proportion of independent directors	IND	Ratio of independent directors to total board size	CSMAR database
	GDP of region	GDP	Logarithm of regional GDP	Statistical Yearbook
	finance of region	FIN	Ratio of total loans from regional financial institutions to regional GDP	Statistical Yearbook

.

5. Empirical Analysis

5.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variables	Obs	Mean	Std. Dev.	Min	Median	Max
DMC	24,779	0.4286872	0.8078952	0	0	2.639057
FP	24,779	0.0438381	0.8770831	−0.37211	0.0407	0.207
SIZE	24,779	21.93657	1.18391	19.191	21.7832	26.95311
AGE	24,779	3.205675	0.207801	2.484907	3.218876	3.688879
LEV	24,779	0.3952418	0.2065144	0.0529	0.3807	1.1285
GROW	24,779	0.262408	0.8047832	−0.846258	0.113057	8.471219
CASH	24,779	0.9330281	1.503803	0.007261	0.404332	8.78901
INT	24,779	0.0458107	0.0381926	0	0.03741	0.3186466
STA	24,779	0.0853659	0.279433	0	0	1
BS	24,779	8.483861	1.583361	5	9	15
IND	24,779	37.42258	5.374298	0	33.33	57.14
GDP	24,779	10.52822	0.7300792	8.063869	10.58067	11.61513
FIN	24,779	3.483291	1.183525	1.687915	3.202359	7.035455

shows the descriptive statistics for the variables. Regarding the independent variable, digital marketing capability (DMC) has a maximum value of 2.64, a minimum value of 0, a sample mean of 0.43, and a median of 0. This suggests considerable variation in the level of digital marketing capability across firms, with majority demonstrating relatively low digital marketing capability. For the dependent variable, firm performance (FP) has a maximum value of 0.21, a minimum value of −0.37, a mean of 0.04, and a median of 0.04. The data reflects substantial differences in firm performance, with most firms having room for improvement. Additionally, the descriptive statistics for the control variables are consistent with those reported in the existing literature, and therefore, no further discussion is provided here.

5.2. Main Effect Test

Table 3. Main effect test.

	(1) FP	(2) FP
DMC	0.0105 ***	0.00485 ***
	(5.23)	(2.82)
SIZE		0.0153 ***
		(19.00)
AGE		−0.0102 ***
		(−3.06)
LEV		−0.176 ***
		(−30.16)
GROW		0.00176 **
		(2.36)
CASH		−0.00171 ***
		(−3.37)
INT		−0.153 ***
		(−7.82)
STA		−0.00417 **
		(−2.25)
BS		−0.0000718
		(−0.13)
IND		−0.000161
		(−1.13)
GDP		0.00534 ***
		(4.39)
FIN		−0.00104 *
		(−1.70)
Constant	0.0392 ***	−0.233 ***
	(44.38)	(−9.86)
Industry FE	YES	YES
Year FE	YES	YES
N	24,753	24,203
R2	0.027	0.250

Notes: ***, ** and * represent statistical significance at the 1%, 5%, and 10% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

reports the baseline regression results for the impact of digital marketing capability on firm performance. Column (1) presents the estimate with only the core independent variable, excluding the control variables. Column (2) presents the results with control variables included, indicating that digital marketing capability significantly improves firm performance. Notably, the impact of digital marketing capability on firm performance remains statistically significant at the 1% level, irrespective of whether control variables are included. The empirical results support H1, suggesting that digital marketing capability can effectively enhance firm performance.

5.3. Robustness Test

5.3.1. Alternative Period

The COVID-19 pandemic had an unprecedented impact on business operations worldwide from 2021 to 2023. Consequently, the data during this period may contain numerous outliers, which could affect the accuracy and reliability of the analysis. Therefore, to better assess the impact of digital marketing on firm performance, this study excludes data from the pandemic period in the robustness check.

Table 4. Alternative Period.

	(1) 2010–2020	(2) 2010–2020	(3) 2021–2023
DMC	0.013222 ***	0.00659 ***	0.00450 **
	(5.25)	(2.95)	(1.97)
SIZE		0.0139 ***	0.0190 ***
		(15.27)	(17.32)
AGE		−0.0101 ***	−0.0115 ***
		(−2.63)	(−2.78)
LEV		−0.166 ***	−0.207 ***
		(−26.20)	(−21.30)
GROW		0.00242 ***	−0.00123
		(2.97)	(−0.74)
CASH		−0.00164 ***	−0.00209 **
		(−2.97)	(−2.40)
INT		−0.140 ***	−0.213 ***
		(−6.43)	(−6.97)
STA		−0.00515 **	−0.000538
		(−2.53)	(−0.15)
BS		0.000244	−0.00123
		(0.45)	(−1.20)
IND		−0.0000851	−0.000443 *
		(−0.56)	(−1.80)
GDP		0.00621 ***	0.00364 **
		(4.80)	(2.08)
FIN		−0.000528	−0.00239 ***
		(−0.79)	(−2.80)
Constant		−0.222 ***	−0.258 ***
		(−8.46)	(−7.71)
Industry FE	YES	YES	YES
Year FE	YES	YES	YES
N	18,678	18,300	5903
R2	0.0364	0.248	0.289

Notes: ***, ** and * represent statistical significance at the 1%, 5%, and 10% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

shows that after removing samples from 2021 to 2023, the dataset is reduced to 18,678 valid observations. For comparison, the analysis considers two model specifications to test robustness. Column (1) reports estimate results with industry and time fixed effects but without firm-level and macro-regional controls, while Column (2) incorporates these additional controls in a fixed-effects model. The findings reveal that between 2000 and 2020, digital marketing capability is significantly positively associated with firm performance. To further verify the findings, the study extends the analysis to the post-public health crisis period (2021–2023). As shown in Column (3), the estimated coefficient of the core independent variable remains significantly positive, confirming the reliability of the conclusions.

5.3.2. Alternative Key Variable

This study examines the robustness of the model by employing different proxy for the dependent variable, and by using alternative measures for the independent variable, separately. On the one hand, this study employs return on equity (ROE) in place of ROA as an alternative indicator of firm performance. ROE, defined as the ratio of net profit to average shareholder equity, is a key indicator of a firm’s profitability. While ROA reflects a firm’s overall asset profitability, ROE focuses on returns to shareholders’ capital. Given their complementary perspectives, these two metrics are often used interchangeably in robustness tests. To assess whether the findings remain consistent under this alternative measure,

Table 5. Alternative Key Variables.

	(1) FP_ROE	(2) FP_ROE	(3) FP_ROA	(4) FP_ROA	(5) FP_ROA	(6) FP_ ROA
DMC	0.0194 ***	0.00952 ***
	(5.13)	(2.82)
DMC_BW			0.00512 ***	0.00316 ***
			(6.93)	(5.37)
DMC_IO					0.401 ***	0.364 ***
					(12.95)	(13.45)
SIZE		0.0372 ***		0.0140 ***		0.0124 ***
		(19.22)		(15.40)		(13.15)
AGE		−0.0173 **		−0.00992 ***		−0.00277
		(−2.47)		(−2.59)		(−0.67)
LEV		−0.346 ***		−0.166 ***		−0.167 ***
		(−19.48)		(−26.21)		(−25.61)
GROW		0.00784 ***		0.00245 ***		0.00368 ***
		(4.48)		(3.01)		(4.52)
CASH		−0.0103 ***		−0.00159 ***		−0.00123
		(−10.00)		(−2.88)		(−1.56)
INT		−0.298 ***		−0.140 ***		−0.102 ***
		(−7.55)		(−6.43)		(−4.59)
STA		−0.00862 *		−0.00512 **		−0.00732 ***
		(−1.83)		(−2.52)		(−3.50)
BS		−0.000513		0.000251		0.000506
		(−0.40)		(0.46)		(0.89)
IND		−0.000379		−0.0000858		−0.0000953
		(−1.22)		(−0.57)		(−0.63)
GDP		0.0112 ***		0.00623 ***		0.00492 ***
		(4.50)		(4.82)		(3.72)
FIN		−0.00184		−0.000511		0.000313
		(−1.45)		(−0.77)		(0.46)
Constant	0.0533 ***	−0.647 ***	0.0382 ***	−0.226 ***	−0.0557 ***	−0.290 ***
	(30.77)	(−12.28)	(39.22)	(−8.62)	(−7.84)	(−10.47)
Industry FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	24,619	24,083	18,678	18,300	15,783	15,501
R2	0.016	0.151	0.038	0.249	0.072	0.269

Notes: ***, ** and * represent statistical significance at the 1%, 5%, and 10% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

presents results from two model specifications. Column (4) includes industry and time fixed effects but excludes firm-level and macro-regional controls, whereas Column (5) incorporates these additional controls in a fixed-effects model. Across both models, the core independent variable remains positive and statistically significant, reaffirming the stability of the baseline regression results.

On the other hand, the independent variable is also subjected to robustness checks. In the main analysis, digital marketing capability is measured by calculating the frequency of dictionary-based keywords extracted from annual reports. To refine this measure, the dictionary is reconstructed by removing some buzzwords—such as terms ending with “-end” or “-cloud”—that have emerged alongside rapid technological iterations in recent years. These terms may capture trend-following rhetoric rather than substantive capability improvements. After eliminating these terms, the recalculated indicator, denoted as DMC_BW, serves as a revised measure of digital marketing capability. Columns (3) and (4) in Table 5 demonstrate that DMC_BW maintains a positive and statistically significant relationship with ROA at the 1% level, confirming the robustness of the findings.

Furthermore, an external proxy for digital marketing capability is introduced, denoted as DMC_IO. This measure is derived from a stochastic frontier production function that models the efficiency of converting marketing-related inputs into outputs. The input set includes four components: (1) marketing expenses related to promotional activities; (2) accounts receivable, which capture aspects of customer relationship management; (3) the customer base size; and (4) digital-related intangible assets, such as proprietary software, databases, or digital platforms. The output is defined as total sales revenue. As shown in Columns (5) and (6) of Table 5, DMC_IO also exhibits a positive and statistically significant effect on ROA (p < 0.01), further validating the robustness of the results.

5.3.3. Heckman’s Two-Stage Test

The digital marketing levels of sample firms are influenced by various factors, which may lead to sample self-selection bias. To address this issue, this study employs the Heckman’s two-stage method, a widely used technique for correcting such biases. In the first stage of the Heckman model, we conduct a Probit regression with dumDM as the dependent variable. dumDM is coded as 1 if the firm has engaged in digital marketing practices, and 0 otherwise. To further account for endogeneity concerns in estimating the impact of DMC on firm performance, we introduce an exogenous variable. Specifically, this study employs the number of domain names (DN), representing regional digital infrastructure that supports digital marketing development [94]. This satisfies the relevance condition. Importantly, DN is unlikely to directly affect firm performance or influence it through unobserved confounders, thus meeting the exclusion restriction. After incorporating the exogenous variable DN and retaining all other original control variables, the Inverse Mills Ratio (IMR) is obtained. In the second stage, IMR is included as a control variable in the original model for further regression analysis. As shown in

Table 6. Heckman‘s two-stage test.

	(1) dumDM	(2) FP
DN	0.0001 *
	(1.78)
DMC		0.00262 *
		(1.94)
IMR		0.117 **
		(2.31)
SIZE	0.0461 ***	0.0189 ***
	(6.26)	(9.33)
AGE	−0.281 ***	−0.0375 ***
	(−7.81)	(−3.19)
LEV	−0.321 ***	−0.201 ***
	(−6.09)	(−14.18)
GROW	−0.00306	0.00292 ***
	(−0.39)	(6.46)
CASH	0.0173 ***	0.000651
	(2.64)	(0.79)
INT	−1.117 ***	−0.206 ***
	(−6.05)	(−4.27)
STA	−0.178 ***	−0.0172 **
	(−5.60)	(−2.25)
BS	−0.0134 **	−0.00143 **
	(−2.10)	(−2.06)
IND	0.00343 *	0.000147
	(1.85)	(0.82)
GDP	0.185 ***	0.0232 ***
	(12.09)	(2.77)
FIN	0.113 ***	0.0101 **
	(16.10)	(2.05)
Constant	−3.515 ***	−0.642 ***
	(−12.85)	(−3.37)
N	37,087	35,124
R2	-	0.235

Notes: ***, ** and * represent statistical significance at the 1%, 5%, and 10% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

, DN exhibits a positive and statistically significant effect on dumDM in Column (1), demonstrating the strong explanatory power of the exogenous variable. In Column (2), after controlling for IMR, the regression results remain highly consistent with previous findings, confirming the robustness of the main conclusions.

5.4. Mechanism Test

To test Hypotheses 2 and 3, which examine the mediating effects of R&D capability and production capability on the relationship between digital marketing and firm performance, this study constructs mediation models (2) and (3) for stepwise testing:

$${Mediator}_{it}=a_0+a_1{DMC}_{it}+{\sum a_{2}CV}_{s_{it}}+\lambda +\gamma +{\epsilon }_{it}$$

(2)

$${FP}_{it}=b_0+b_1{Mediator}_{it}+b_2{DMC}_{it}+{\sum b_{3}CV}_{s_{it}}+\lambda +\gamma +{\epsilon }_{it}$$

(3)

In model (2) and (3), the subscript i represents the firm, and t represents time. ${FP}_{it}$ represents the firm’s performance, and ${DMC}_{it}$ represents the firm’s digital marketing level. ${Mediator}_{it}$ represents the mediating variable, and ${CV}_{s_{it}}$ represents the control variables. The coefficient $a_1$ of ${DMC}_{it}$ in model (2), along with the coefficients $b_1$ of ${Mediator}_{it}$ and $b_1$ of ${DMC}_{it}$ in model (3), are the focus of this study. If $a_1$, $b_1$ and $b_2$ are all significantly positive, it suggests that digital marketing capability can enhance the mediating variable, thereby improving the firm performance.

Table 7. Mediating effect test.

	(1) PAT	(2) FP	(3) TFP	(4) FP
DMC	0.000138 ***	0.00479 ***	0.0365 ***	0.00391 **
	(4.87)	(2.66)	(21.67)	(2.27)
PAT		0.000133 ***
		(4.73)
TFP				0.0363 ***
				(21.43)
SIZE	0.0158 ***	0.0157 ***	−0.000302	−0.000357
	(18.62)	(18.58)	(−0.27)	(−0.32)
AGE	−0.00858 **	−0.00849 **	−0.0108 ***	−0.0107 ***
	(−2.41)	(−2.39)	(−3.26)	(−3.24)
LEV	−0.177 ***	−0.177 ***	−0.179 ***	−0.178 ***
	(−28.95)	(−28.95)	(−30.26)	(−30.25)
GROW	0.00266 ***	0.00268 ***	0.00304 ***	0.00307 ***
	(3.47)	(3.5)	(3.94)	(3.99)
CASH	−0.00137 **	−0.00137 **	−0.000704	−0.000715
	(−2.17)	(−2.18)	(−1.22)	(−1.24)
INT	−0.140 ***	−0.141 ***	−0.0836 ***	−0.0843 ***
	(−6.98)	(−7.01)	(−4.36)	(−4.40)
STA	−0.00549 ***	−0.00537 ***	−0.00581 ***	−0.00565 ***
	(−2.88)	(−2.81)	(−3.18)	(−3.08)
BS	−0.0000606	−0.0000587	0.000312	0.000301
	(−0.10)	(−0.10)	(0.58)	(0.56)
IND	−0.000185	−0.000187	−0.0000886	−0.0000927
	(−1.24)	(−1.26)	(−0.62)	(−0.65)
GDP	0.00542 ***	0.00533 ***	0.00243 **	0.00234 **
	(4.32)	(−4.26)	(2.1)	(2.02)
FIN	−0.000876	−0.000936	−0.00282 ***	−0.00289 ***
	(−1.37)	(−1.46)	(−4.51)	(−4.63)
Constant	−0.253 ***	−0.251 ***	−0.0989 ***	−0.0963 ***
	(−10.24)	(−10.19)	(−4.10)	(−4.01)
Industry FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
N	21,927	21,927	21,927	21,927
R2	0.25	0.251	0.312	0.312
Sobel Test	Z = 8.912 ***		Z = 6.911 ***

Notes: *** and ** represent statistical significance at the 1% and 5% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

presents the results of testing the mediating role of R&D capability in the relationship between digital marketing capability and firm performance. Column (1) shows a significantly positive coefficient for the impact of digital marketing capability on R&D capability, indicating that stronger digital marketing capability is linked to higher R&D capability. Column (2) also shows a positive relationship between R&D capability and firm performance. Moreover, as the direct effect of digital marketing capability on firm performance remains statistically significant after controlling for R&D capability, the results indicate a partial mediation effect. The Sobel test yields a statistically significant result (Z = 8.912, p < 0.01), confirming the presence of a mediation effect. In addition, a bootstrap analysis with 500 iterations shows that the bias-corrected 95% confidence interval for the indirect effect does not include zero, further validating the mediation pathway. Together, these results suggest that digital marketing capability enhances firm performance by strengthening R&D capability, thus providing support for H2.

Column (3) shows a positive association between digital marketing capability and production capability. Column (4) incorporates both digital marketing capability and production capability as predictors of firm performance. The results confirm that production capability contributes to firm performance while the effect of digital marketing capability remains significant, indicating a partial mediation effect. The Sobel test yields a statistically significant result (Z = 6.911, p < 0.01), confirming the mediation effect. In addition, bootstrap analysis with 500 iterations shows that the bias-corrected 95% confidence interval for the indirect effect does not include zero, further validating the mediation pathway. These findings suggest that digital marketing capability enhances firm performance through its positive effect on production capability, thus providing support for H3.

5.5. Heterogeneity Analysis

This study further explores the heterogeneous effects of digital marketing on firm performance. Specifically, it examines these effects at the regional level (regional marketization degree), industry level (industry competition intensity), and firm level (firm digitalization degree).

Firstly, this study employs the Marketization Index to measure regional marketization degree, drawing data from the China Provincial Marketization Index Report. This index captures five key dimensions: the relationship between government and market, the development of the non-state economy, the maturity of product and factor markets, the maturity of market intermediaries, and the legal environment supporting market operations. A higher index value indicates stronger marketization, while a lower value reflects weaker marketization. To examine how marketization influences the relationship between digital marketing capability and firm performance, this study classifies regions into high- and low-marketization groups using the index’s median as a threshold. Separate regressions are conducted for each group, with results presented in

Table 8. Heterogeneity analysis.

Variable	(1) Regional Marketization—LOW	(2) Regional Marketization—HIGH	(3) Industry Competition—LOW	(4) Industry Competition—HIGH	(5) Firm Digitalization—LOW	(6) Firm Digitalization—HIGH
DMC	0.00147	0.00670 ***	0.00329	0.00687 ***	0.00441	0.00435 **
	(1.91)	(1.56)	(1.40)	(2.89)	(0.94)	(2.42)
SIZE	1.272 ***	1.426 ***	0.0158 ***	0.0146 ***	0.0133 ***	0.0168 ***
	(12.62)	(13.48)	(13.79)	(15.06)	(11.25)	(17.80)
AGE	−1.572 ***	−2.334 ***	−0.00670	−0.0134 ***	−0.00197	−0.0156 ***
	(−3.19)	(−4.42)	(−1.59)	(−2.89)	(−0.38)	(−4.16)
LEV	−7.081 ***	−7.900 ***	−0.180 ***	−0.173 ***	−0.163 ***	−0.190 ***
	(−12.45)	(−11.07)	(−21.58)	(−23.41)	(−20.82)	(−24.96)
GROW	−0.232 ***	−0.400 ***	0.00157	0.00191	0.00327 ***	0.000300
	(−2.58)	(−3.88)	(1.64)	(1.63)	(2.88)	(0.32)
CASH	0.0720	−0.0271	−0.00274 ***	−0.000869	−0.00117 *	−0.00236 ***
	(1.39)	(−0.40)	(−3.96)	(−1.29)	(−1.71)	(−3.59)
INT	−5.181 *	−2.398	−0.152 ***	−0.152 ***	−0.120 ***	−0.193 ***
	(−1.89)	(−0.86)	(−5.46)	(−6.23)	(−4.46)	(−7.79)
STA	−0.283	0.399	−0.00412	−0.00416 *	−0.00597 **	−0.00213
	(−1.24)	(1.19)	(−1.60)	(−1.76)	(−2.38)	(−0.88)
BS	0.140 **	0.276 ***	−0.000577	0.000418	−0.000490	0.000294
	(2.04)	(3.57)	(−0.82)	(0.56)	(−0.64)	(0.45)
IND	0.102 ***	0.0978 ***	−0.000229	−0.000106	−0.000332 *	−0.0000242
	(5.87)	(4.58)	(−1.18)	(−0.58)	(−1.70)	(−0.14)
GDP	0.283 *	0.698 ***	0.00488 ***	0.00577 ***	0.00535 ***	0.00565 ***
	(1.81)	(4.04)	(3.00)	(3.51)	(3.10)	(4.02)
FIN	−0.523 **	0.142	−0.00148 *	−0.000675	−0.000243	−0.00143 **
	(−2.09)	(1.33)	(−1.75)	(−0.89)	(−0.28)	(−1.98)
Constant	46.53 ***	38.23 ***	−0.241 ***	−0.221 ***	−0.212 ***	−0.254 ***
	(14.50)	(11.16)	(−7.61)	(−7.14)	(−6.03)	(−9.24)
Industry FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	10,040	7888	12,170	12,033	10,082	14,121
R2	0.147	0.164	0.251	0.252	0.273	0.244

Notes: ***, ** and * represent statistical significance at the 1%, 5%, and 10% levels, respectively; t-values are reported in parentheses; the robust standard errors are clustered at the firm level.

. The estimates in Columns (1) and (2) show that digital marketing capability exhibits a significantly positive relationship with firm performance in highly marketized regions. In contrast, the effect is not statistically significant in less marketized regions. A possible explanation is that firms in highly marketized regions operate in a more conducive environment for digital marketing. They benefit from a well-developed infrastructure, easier access to capital, technology, and human resources, as well as greater flexibility in resource allocation and business operations. This enables them to swiftly adjust marketing, R&D, and production strategies, effectively unlocking the performance-enhancing potential of digital marketing capability.

Secondly, this study uses the Herfindahl-Hirschman Index (HHI) to measure industry competition intensity. The HHI captures the number, size, and distribution of firms within a market. A lower HHI indicates a more competitive industry with less concentration, while a higher HHI suggests greater market dominance and less competition. To explore how industry competition affects the relationship between digital marketing and firm performance, the sample is split into high- and low-competition groups based on the median HHI. Separate regressions are then performed for each group, with results presented in Columns (3) and (4) of Table 8. The findings reveal that in highly competitive industries, digital marketing significantly influences firm performance. However, the effect is not statistically significant in industries with lower competition. A possible explanation is that in more competitive markets, firms face higher operational pressures, making digital marketing crucial for brand development and market penetration. In such environments, where product imitation and homogeneity are prevalent, digital marketing capability provides firms with the tools to gauge market demands, accelerate new product development, and refine production strategies. These capabilities help firms stand out, enhance their competitive positioning, and ultimately improve performance.

Lastly, this study uses the Digitalization Index of firms to assess the extent of digital transformation within firms. This index captures various dimensions, such as the integration of digital technologies, the establishment of modern information systems, the adoption of internet-driven business models, and the implementation of smart manufacturing practices. A higher index value indicates a more advanced state of digitalization, while a lower value reflects limited digital adoption. To examine how digitalization moderates the relationship between digital marketing and firm performance, the sample is divided into high- and low-digitalization groups based on the median index value. Separate regressions are conducted for each group, with results presented in Columns (5) and (6) of Table 8. The results reveal a digitalization-contingent effect: digital marketing capability significantly influences firm performance in highly digitalized firms, but its impact is not significant for less digitalized firms. A possible explanation is that firms with a higher degree of digitalization already have a solid foundation in strategic leadership, organizational structure, and data utilization. These advantages enable digital marketing capability to integrate more seamlessly with other business functions, fostering operational efficiency and strengthening its role in value creation. As a result, this synergy leads to improved firm performance.

5.6. Results

This study demonstrates that digital marketing capability positively impacts firm performance. The results remain robust after adjusting the sample period, replacing key variables, and conducting Heckman’s two-stage tests. Furthermore, this effect operates through two key mechanisms: R&D capability and production capability. From the perspective of digital marketing capability and R&D capability, firms with strong digital marketing capability can accurately track market trends and customer preferences, providing data-driven insights for product design. This enables the development of products that cater to diverse and personalized customer needs. Regarding digital marketing capability and production capability, digital marketing capability allows firms to assess and anticipate market shifts and emerging customer demands, facilitating proactive production planning. This ensures alignment between production and market needs, enhances manufacturing efficiency, and ultimately contributes to firm performance. For manufacturing firms, digital marketing capability amplifies its value by simultaneously enhancing R&D and production capabilities, two core business functions crucial to firm success. Heterogeneity tests further indicate that the positive impact of digital marketing capability on firm performance is more pronounced among firms in highly marketized regions, competitive industries, and those with stronger digitalization foundations.

6. Discussion

6.1. Theoretical Contributions

This study contributes to the existing literature in three ways.

Firstly, this study deepens the understanding of how digital marketing capability influences firm performance by examining its indirect effects through R&D capability and production capability. Existing research has primarily focused on marketing-related mediators, including market sensing capability [23,52], market grasping capability, market reconfiguration capability, dynamic marketing capability [52], customer-linking capability [23], and customer relationship management capability [24]. In contrast, this study highlights the cross-functional value of digital marketing in enhancing innovation and operational efficiency, thereby offering a more comprehensive perspective on its performance implications.

Secondly, this study contributes to the digital marketing literature by developing a measurement of digital marketing capability from the data utilization perspective. Previous studies have approached the measurement of digital marketing capability from various angles. One stream emphasizes relationship management, using indicators such as customer-linking digital capability and the ability to build long-term relationships with suppliers [8]. Another focuses on technology-enabled marketing practices, capturing firms’ capabilities in areas such as social media [50], mobile, content, and search engine marketing [18]. Although existing definitions of digital marketing capability emphasize the importance of data access and utilization [50], current measurement approaches have yet to incorporate corresponding indicators. Informed by resource orchestration theory and grounded in insights from data analytics research, this study argues that firms build digital marketing capability by integrating and leveraging two core resources: data and technology. This perspective responds to Kannan and Li’s (2017) call for deeper inquiry into value creation through data utilization [65], by highlighting how firms strategically mobilize data assets in digital marketing contexts.

Thirdly, this study enriches the literature on digitalization in the manufacturing industry. Existing research on digitalization in manufacturing firms has predominantly focused on the adoption of Industry 4.0 technologies in the production process [31,32], including intelligent manufacturing [33], smart manufacturing [34], flexible manufacturing [35], and advanced manufacturing [36]. In contrast, the digitalization of front-end activities, such as e-commerce [77,95], has received considerably less scholarly attention. By demonstrating significant effects of digital marketing capability, this study enriches theoretical understanding of how manufacturing firms leverage digital marketing to achieve high-value creation in the digital era.

6.2. Practical Implications

Firstly, in the digital era, manufacturing firms should recognize that digital marketing activities are essential for breaking free from low-value segments of the “smile curve” and driving value creation. Digital marketing capability serves as a key enabler, facilitating stronger consumer connections, optimizing products and services, and accelerating business growth. To fully capitalize on this capability, firms must increase investment, refine their development strategies, and integrate digital marketing capability into their overall strategic planning. A crucial aspect of this process is leveraging data and digital technologies to sense market dynamics and understand customer needs. Firms should build a comprehensive data collection and analysis system to capture real-time market trends, customer feedback, and competitor activities. By systematically analyzing and applying these insights, firms can identify shifts in consumer demand, anticipate market changes, and establish agile response mechanisms. These findings provide actionable pathways for manufacturing firms in developing economies, particularly for original equipment manufacturers (OEMs), to leverage DMC for value chain upgrading and performance enhancement.

Secondly, firms should recognize the critical role of internal capability building in translating digital marketing capability into performance gains. Specifically, firms should leverage digital marketing capabilities to facilitate the advancement of R&D and production functions, while continuously improving coordination mechanisms. To achieve this, managers should actively promote cross-departmental collaboration, institutionalize cooperative workflows, and formalize decision-making protocols that safeguard interdepartmental synergy. At the technological level, firms should invest in integrated data platforms to enable the seamless flow of marketing information into R&D and production, eliminating data silos and enhancing real-time responsiveness. In terms of human capital, firms must provide targeted training to improve employees’ digital literacy and analytical skills. From a cultural perspective, fostering open communication, encouraging knowledge sharing, and cultivating an “internal customer” mindset are essential to building a collaborative environment. Collectively, these initiatives drive the cross-functional integration necessary for firm-wide digital transformation and operational upgrading.

Thirdly, firms should adopt differentiated digital marketing strategies based on their regional, industry, and capability profiles. Firms in highly marketized regions are advised to leverage local policy support—such as digital transformation subsidies—while those in less marketized areas should collaborate with industry associations to improve regional digital infrastructure. In highly competitive industries, firms should allocate specific budgets for competitive intelligence and real-time monitoring of key market indicators to enhance the strategic use of DMC. Additionally, given the digital capability threshold effect, firms with weaker digital foundations need to prioritize fundamental capacity building, such as employee training and basic data system construction. In contrast, firms with strong digital capabilities are encouraged to develop integrated data platforms connecting marketing, R&D, and production, enabling faster conversion of market insights into actionable strategies across business functions.

6.3. Limitations and Future Research

This study has several limitations that offer opportunities for future research.

First, our sample focuses on Chinese A-share listed manufacturing firms, which may limit the generalizability of the findings to non-listed firms, other industries, or different national contexts. In particular, the unique business environment in China—characterized by strong government-led digitalization efforts, policy-driven resource allocation, and uneven regional digital infrastructure—may shape the dynamics of digital marketing capability (DMC) adoption in ways that differ from those in developed economies. Future research could extend this study by conducting cross-country comparisons to examine whether similar mechanisms hold in other institutional settings.

Second, although we employed multiple robustness checks—including variable substitution and Heckman correction procedures—to mitigate endogeneity concerns, potential bias from omitted variables may still exist. Unobserved factors such as managerial style, organizational culture, or firms’ internal digital mindsets could simultaneously influence both DMC deployment and performance outcomes. Addressing these factors in future studies, perhaps through richer firm-level data or qualitative inquiry, would further validate the causal pathways proposed.

Third, the measurement of digital marketing capability relies on a keyword dictionary method applied to annual reports. While this approach allows for large-scale, replicable data collection, it may not fully capture the breadth and depth of firms’ actual digital marketing practices. Although we validated the measure using alternative proxies such as marketing expenses, future research could improve measurement precision by incorporating practice-based indicators, survey data, or multi-source triangulation.