Linking Dynamic Capabilities and Market Orientation to Sustainability Through Green Marketing in the Batik Industry

The batik industry represents a vital interaction between Indonesia's cultural heritage and creative economy, contributing to employment, regional development, and international trade. Since UNESCO recognized batik as an intangible cultural heritage of humanity in 2009, visibility and expectations regarding authenticity, quality, and sustainability have increased [1]. However, sustainable growth faces challenges from environmental concerns, such as the use of synthetic dyes and water-intensive processes, shifting consumer preferences for ethical products, and competition from mass-produced textiles [2].

In this dynamic context, two strategic orientations gain attention: dynamic capabilities (DC), which involve sensing opportunities, exploiting them, and reconfiguring resources, and market orientation (MO), reflecting an organization's responsiveness to market needs [3]. Although widely studied in manufacturing and services, the joint impact of these factors on sustainable performance in cultural sectors, especially Southeast Asia’s creative industries, remains limited [4]. The textile sector is one of the most harmful industries to the environment, responsible for climate change and global warming. That is to say, businesses must make use of green ways to both the economy and the environment. The sector now points at economic growth, environmental protection, and social welfare as the three pillars of sustainability [5]. An increasing number of global citizens are realizing the importance of sustainability. The change ensures a steady supply of raw material for the industry and helps in making the production more efficient. According to Müller et al. [6], this finding is being paid more attention to by researchers, especially in developing countries. The use of green technologies, conceptual design and sustainable innovations helps to reduce energy use, waste and depletion of resources [7, 8]. This will accelerate the transition to a greener environment.

Sustainable success involves economic, biological, and social aspects [9]. As part of efforts to achieve this goal, cyclical economics and natural dyes, considered more environmentally friendly, are being used. However, the Indonesian batik industry still faces numerous challenges due to the use of artificial dyes, which have a negative impact on the environment [10], as noted by UNESCO. Some alternatives include reusing old materials to protect the environment [11], creating local motifs such as Batik Betawi Seraci [12], and utilizing plant-based dyes sourced from regional areas [13].

Teece [14] and Wilden et al. [15] defined DC as an organization's capacity to leverage resources to generate new value and enhance its competitiveness. In batik, this is demonstrated by the integration of ancient techniques with modern advancements and the use of environmentally friendly technologies, which promote sustainability [16, 17]. The global market reputation of batik producers has been enhanced by the integration of green marketing strategies into their marketing mix, achieved through the use of natural dyes and effective waste management [18, 19]. The adoption of a DC framework that integrates Industry 4.0 and sustainable productivity improves operational efficiency and flexibility [20], thereby impacting the organization, the environment, and society [21]. Simultaneously, MO drives sustainability by prioritizing customer needs, thereby encouraging innovation and marketing strategies rooted in ecologically sustainable practices [22, 23].

Green marketing links environmental awareness with design, production, and promotional strategies [24]. Although this concept is increasingly discussed, related research remains limited. Empirical proof confirming the connection between DC, MO, and sustainable performance is restricted [25]. Examples include the use of natural colors, minimizing waste, and disseminating sustainability messaging to environmentally aware consumers [26]. Moreover, the mediating function of green marketing remains underexplored [27], and targeted research within the context of batik centers is also notably scarce [28].

Despite growing interest in sustainability within cultural and creative industries, existing studies tend to examine DC, MO, and green marketing in isolation, leaving a limited understanding of how these constructs interact to shape sustainability outcomes in heritage-based sectors like batik. Prior research has also been dominated by manufacturing or service contexts, resulting in theoretical frameworks that may not fully capture the cultural, artisanal, and resource-constrained nature of traditional craft industries. Therefore, a conceptual gap remains regarding how DC and MO jointly stimulate environmentally oriented marketing practices and influence multidimensional sustainability performance. Addressing this gap is crucial for advancing theory and providing evidence specific to cultural industries operating in emerging economies.

This research aims to fill these gaps by examining the correlation between dynamic capacities, MO, and sustainable performance, with green marketing serving as a mediating variable. The study was administered to batik businesses, namely those based in batik hubs in Central Java.

Table 1 presents a complete summary of the reliability and validity test results for the model in this study. The results indicate that the measurement model has a high level of reliability and validity. Cronbach-α values for all constructs were recorded above 0.5 (DC = 0.500; MO = 0.80), confirming internal consistency among the indicators. The average variance extracted (AVE) values were above 0.5 (DC = 0.796; MO = 0.680), and the CR values exceeded 0.6 (DC = 0.680; MO = 0.864), indicating convergent validity. These results highlight that batik firms with DC, a strong MO, and effective green marketing practices are better positioned to achieve enhanced economic, environmental, and social competitive sustainability outcomes. This study tested all first-order models to ensure the authenticity of estimates. The variables are DC, including sensing (DC1), learning (DC2), integration (DC3), and reconfiguration (DC4); marketing orientation (MO), including customer orientation (MO1), competitor orientation (MO2), and inter-functional coordination (MO3).

Table 1. Summary of reliability of reflective construct

2.jpg

Figure 2. Cross-loading between variables

A minimum threshold of 0.70 was set for all components with intelligent features evaluated. Following, AVE values must exceed 0.50, while internal consistency (Cronbach-α and CR) should be above 0.70. Figure 2 presents the validity results based on these threshold values. The cross-loadings confirm discriminant validity, as each construct loads higher on its own factor than on any other [30]. The highest cross-loading was observed in the GMM, ranging from 0.676 to 0.708, which meets the required thresholds. This ensures each construct captures distinct phenomena.

According to Chin et al. [31], R² values are classified as high (> 0.67), moderate (0.33–0.67), weak (0.19–0.33), and unacceptable (< 0.19). Similarly, Hair et al. [32] consider R² values of 0.75 to be substantial, 0.50 to be moderate, and 0.25 to be weak for endogenous variables. Table 2 shows that three constructs, including reconfiguration capability (DC4), inter-functional coordination (MO3), and socially appropriate industrial performance (sense3, have high determinant coefficients). Six constructs, including learning capability (DC2), integration capability (DC3), client identification (MO1), competitor orientation (MO2), and two sustainability measures (SIPa, SIPb), fall into the moderate category, while detection capability (DC1) is weak. Overall, capabilities in learning, integration, reconfiguration, client recognition, competitor recognition, and inter-functional coordination emerge as the main drivers of development. The empirical results show a clear distinction among the three dimensions of SIP. Both the descriptive statistics and the coefficient of determination values indicate that environmental performance (SIPb) consistently lags behind the economic (SIPa) and social dimensions (SIPc). This is reflected in the lower R² value for SIPb (0.544), compared with 0.599 for SIPa and 0.718 for SIPc, suggesting that environmental initiatives are not yet as strongly embedded within batik firms as their economic and social efforts. These findings highlight a performance gap in the environmental domain, reinforcing the notion that ecological practices receive comparatively less attention and support.

Table 2. The coefficient of determination results from the first-order reflective

According to Sarstedt et al. [33], PLS-SEM is commonly used to model higher-order constructs that combine concrete lower-order and abstract higher-order dimensions. This study employs a reflective, formative hierarchy model, using the GMM. Using the two-stage PLS method, mediation effects were tested, ensuring more accurate estimates and stronger validity. GMM integrates eco-friendly elements, including people, processes, and products, derived from first-order constructs.

3.jpg

Figure 3. Cross-loading second order

In Figure 3, an indicator is considered significant if it has the highest factor loading on the target construct compared to other constructs. This figure shows that the loading factor for the DC indicators (DC1 to DC4) is higher for the DC construct than for the different constructs. The loading factor of DC2 to DC4 is 0.729-0.786, which is higher than those of the GMM (0.659), SIPa (0.444), SIPb (0.586), and SIPc (0.623). Hence, the latent variable better predicts the indicators in its block compared to others.

4.jpg

Figure 4. Fornell and Larcker’s for the second-order construct

Based on the Fornell-Larcker test (Figure 4), the constructs–energy capability (DC), MO, financially viable industrial performance (SIPa), environmentally sustainable industrial performance, and socially sustainable industrial performance, satisfy discriminant validity, as their diagonal values exceed those below them. Convergent and discriminant validity were further assessed using the HTMT test, where HTMT < 0.9 indicates validity. These tests confirm that each construct makes a significant contribution to explaining the dependent variable, as shown in Table 3.

Table 3. Heterotrait-monotrait ratio (HTMT) bootstrab - reflective second order

The HTMT test results (Table 3) confirm good convergent validity and construct differentiation for DC to SPIb (0.362) and SPIc (0.309). However, the path from DC to SIPa is negative (-0.005). The negative and significant path from DC to the economic dimension of sustainable industry performance (SIPa) suggests that strengthening DC may create short-term economic pressure for batik firms. This pattern is theoretically reasonable, as the development of DC typically requires substantial investment in learning, experimentation, and resource reconfiguration, activities that increase operating costs and reduce short-term financial gains. Prior studies emphasize that capability upgrading often produces temporary inefficiencies, learning burdens, and adjustment costs before the long-term benefits emerge [35]. In the context of batik MSMEs, which generally operate with limited capital buffers, these transitional costs are likely to be more pronounced, explaining the observed adverse effect on economic SIP. The highest MO variable is SPIc at 0.223. Meanwhile, the GMM shows values ranging from 0.584 to 0.840 across all SPI.

This study presents four main hypotheses: three directly related to the dependent and independent variables, and one involving the interaction effects of mediating factors. Hypotheses regarding direct effects are discussed first, followed by those related to mediating factors, in line with Henseler et al. [36], who emphasized the importance of assessing direct effect hypotheses before conducting analyses with interactive variables in PLS path modeling. All main and sub hypotheses regarding the unidirectional relationships between the dependent and independent factors are supported by the findings, as shown in Table 4. These include the relationships between DC, MO, and GMM with sustainability outcomes. These relationships are significant with p-values less than 0.01, except for the hypotheses regarding DC and economic sustainability (H1a), which were rejected. This study tests hypotheses on the links between DC, MO, green marketing mix (GMM), and SIP. Findings show that DC has a strong influence on GMM but only limited effects on SIP, particularly in environmental and social aspects. MO demonstrates weak impacts on both GMM and SIP, with the lowest effect on environmental SIP. Conversely, GMM acts as the primary driver of SIP, with notable impacts on both economic and social dimensions, although its effect on ecological SIP remains modest. Bias-corrected confidence intervals confirm these results, including the most minor yet statistically significant path from DC to SIPc. Overall, the study highlights GMM’s pivotal role in promoting sustainable performance, while underscoring the need for more robust strategies to address environmental outcomes.

Table 4. Hypotheses testing results

Table 5. Confidence interval bias results

Table 5 reports confidence intervals at 2.5%–95%. Relationships are significant when zero is not included, while intervals crossing zero indicate insignificance. Bias-corrected intervals confirm significance for several hypotheses, including DC–SPIb, DC–SPIc, MO–SPIa, MO–SPIb, MO–SPIc, DC–GMM, GMM–SPIa, GMM–SPIb, and GMM–SPIc.

Table 6. Convergent validity of the second order of reflective variables

This study investigates the mediating role of the GMM in the proposed model. Results show that GMM mediates the link between DC and sustainable economic and environmental performance, as well as between MO and SPIb and SPIc performance. Mediation was tested using PLS-SEM with bootstrapping, ensuring robustness through valid measurement indicators (factor loading, CR, convergent validity). Table 6 confirms strong empirical support for GMM as a mediating mechanism.

This study examines the factors influencing SIP in the batik sector of Central Java, Indonesia. The findings highlight the interplay of DC, MO, and GMM in shaping economic, social, and environmental sustainability. DC, particularly sensing, learning, and integration, contribute positively to social and ecological outcomes, although reconfiguration remains underdeveloped. This suggests that batik firms can recognize opportunities and adopt sustainable practices but face challenges in restructuring processes for long-term adaptation. Strengthening reconfiguration is therefore essential to advance green technologies and sustainable production methods.

MO greatly enhances economic and social sustainability by fostering a robust emphasis on customers and competitors. It has a weaker effect on environmental performance, highlighting the importance of taking proactive steps to protect the environment, such as reducing waste, using environmentally friendly materials, and promoting cleaner practices. The GMM variable also helps DC and MO turn internal strengths into measurable results. You can achieve this by developing new, eco-friendly products, establishing competitive prices, and effectively advertising them. There are still a few new products or ways to discuss the environment in the batik industry. The batik industry demonstrates a high level of social responsibility. However, its environmental sustainability is slightly lower. These results indicate that community involvement is already strong; however, further efforts are needed to mitigate environmental impacts and ensure the industry's long-term sustainability.