The term green marketing was introduced in 1975 by the American Marketing Association in a workshop on Ecological Marketing (Zhu & Sarkis, 2016). This construct is also referred to as environmental marketing, ecological marketing, social marketing and sustainability marketing (Zhu & Sarkis, 2016). In the context of this study, it is defined as the efforts by a company to design, encourage, price and distribute products in a way that endorses environmental conservation (Polonsky, 2011).

This study measures green marketing practices as a unidimensional construct (refer to Figure 1) comparable to a study by Nuryakin and Maryati (2022) of green marketing performance. At the outset, this study hypothesises that there is a positive relationship between green marketing practices and green logistics practices. One of the functions that resonates with both constructs is distribution. In marketing, distribution implies creating awareness of the product or service (Halik et al., 2023) while in a logistics context, it denotes delivery (Konstantakopoulos et al., 2022). Therefore, fulfilling the purpose of green marketing is contingent on effective logistics that is appropriate – and thus ‘green’. Green marketing practices have further been hypothesised to also influence operational performance and environmental performance. According to Pandey et al. (2018), green marketing practices encourage green purchasing, which directly influences the methods used in the production and packaging process. This state of affairs implies that green marketing practices are likely to dictate operations as well as the performance thereof. Pandey et al. (2018) add that ecological degradation is directly associated with people’s consumption patterns and, therefore with the methods applied in the production and packaging processes. This reality suggests that green marketing practices would improve consumption patterns and therefore influence environmental performance in a positive way. In the same vein, this study contends that there is a positive relationship between green marketing practices and supply chain performance. Given that green marketing practices are reflected in the environmental aspect of the Triple Bottom Line (TBL) concept developed by Elkington (1994), it is argued that the concept is correlated with supply chain performance, which is at the core of the TBL approach (Huang et al., 2021). In addition, this study proposes that both operational performance and environmental performance mediate the relationship between green marketing practices and supply chain performance, respectively. While this study acknowledges the relationship between green marketing practices and supply chain performance, it maintains that operational performance intermediates and should be prioritised on the premise that it enables green marketing practices – which ultimately influence supply chain performance. Equally, green marketing practices need to be realised and identified on the balance sheet of environmental management accounting in order to reinforce the relationship between green marketing practices and supply chain performance. Drawing from this literature, the following hypotheses were therefore developed:

The concept ‘green logistics’ entails planning, controlling and implementing the flow of logistics through incorporating modern logistics methods with the objective of reducing environmental hazards (Tan et al., 2020). The methods adopted are usually orientated towards logistics environmental management, low-carbon warehousing and packaging as well as low-carbon transportation, fleet management, alternative energy and logistics innovation (Zhang et al., 2014). Brahme and Shafighi (2022) also emphasise that green logistics should be operationalised to the satisfaction of customers, and should support the organisation’s goals. For the purpose of this study, green logistics practices are defined as:

Although the concept can be examined as a multidimensional construct (see Agyabeng-Mensah et al., 2020b), it is conceptualised as a unidimensional construct (refer to Figure 1) consistent with the literature (e.g. Agyabeng-Mensah & Tang, 2021) and befitting of the purpose of this study.

The measure of green logistics practices as a predictor is corroborated in the literature (e.g. Agyabeng-Mensah et al., 2020a; Agyabeng-Mensah & Tang, 2021). Accordingly, this study predicts that green logistics practices influence operational performance, environmental performance and supply chain performance. Kim et al. (2024) asserted that green logistics practices are important because they help businesses to create value. This value can stem from green practices that minimise environmental degradation related to greenhouse emissions, transportation, packaging and noise pollution (Lai & Wong, 2012) and more importantly applied without detracting from the performance of the firm. This study therefore hypothesises that green logistics practices influence operational performance in a positive way, but only if there is definitive value created. Equally, this study conceives that there is a positive relationship between green logistics practices and environmental performance. The practices that define green logistics practices are regarded as having a progressive influence on environmental performance, particularly if they are regulated by the organisation (Agyabeng-Mensah et al., 2020b). This study further hypothesises that there is a positive relationship between green logistics practices and supply chain performance. According to Huang et al. (2024), green concepts focus on diminishing ecological footprints from supply chain processes while also improving corporate profits. However, because supply chains have evolved and become more complex, green practices need to be implemented with careful planning and attention, and should be supported by coordination and collaboration. Furthermore, this study acknowledges that environmental performance and operational performance mediate the relationship between green logistics practices and supply chain performance, respectively. There is contention in the literature that green logistics practices need to be reflected in the firm’s environmental performance – indeed, the Global Reporting Initiative (GRI, 2014) recommends this if the impact on supply chain performance is to be credible. Because optimisation and sustainability are mandated in firm operations, green logistics practices cannot improve supply chain performance if such operational performance is not realised. For these reasons, this mediation is considered as a hypothesis. In line with this literature, the following hypotheses were therefore developed:

Considerable attention has been given to several performance measures over the past three decades. These include measures such as the TBL, the balanced scorecard, competitive advantage, stakeholder performance, innovation performance, varying financial outcomes and so forth (e.g., see Berman et al., 1999; Brower & Mahajan, 2013; Zaharia & Zaharia, 2021). Given that these measures determine performance in specific value chain activities, it is understandable that there would be differences in the definition of operational performance. This study adopts a definition from Ferdows and De Meyer (1990), who assert that operational performance relates to advances in cost minimisation, quality, delivery and flexibility. While operational performance may be a multidimensional construct (see Saryatmo & Sukhotu, 2021), this study measures the construct as a single dimension (refer to Figure 1).

Furthermore, this study proposes that operational performance influences environmental performance and supply chain performance in a positive way. Before the 1960s, the objective for many organisations was to prioritise operational performance by maximising revenue without much regard for the cost to the environment (Antão et al., 2016). However, since the 1980s, environmental conservation has been imposed and mandated on organisations, compelling firms to conduct their operations in ways that point to greater efficiency in this regard (Agyabeng-Mensah et al., 2022). This situation aims to achieve a balance where revenues are generated without compromising the environment. Therefore, operational performance influences environmental performance in a positive way only if it is applied with efficiency. Furthermore, developments and strategies taken up at the organisational level must be congruent with those that are conducted at the supply chain level. Therefore, adjustments made in operations must be coherent across all supply chain members if there is to be stability and an increase in supply chain performance. In accordance with this literature, the following hypotheses were therefore developed:

These days, effective environmental performance is expected from an organisation. It is believed that good environmental performance results in a competitive advantage that derives from cost savings and corporate reputation (Uddin, 2021). According to Kim et al. (2019), environmental performance includes the accomplishments emanating from improving the natural and social environment through the support of management tools. It entails a process that requires the involvement of a firm’s resources, including the commitment of top management, incorporating business strategy with environmental objectives and adhering to environmental management accounting. Performance indicators are used to evaluate the process and are listed by GRI in terms of materials, energy, water, biodiversity, emissions, effluents and waste, products and services, compliance, transport, total expenditure and investments in environmental initiatives, supplier environment assessment and environmental grievance mechanism (GRI, 2014).

This study examines environmental performance as a unidimensional construct (refer to Figure 1) and defines it as a firm’s awareness of its liability and accountability towards sustainable development (Cho et al., 2012). The study further hypothesises that environmental performance influences supply chain performance in a positive way. The reasoning here is that environmental performance has been found to lead to a competitive advantage (Uddin, 2021), which results in an increase in supply chain performance. According to the resource-based view, firm-specific organisational resources and capabilities, such as environmental technologies, can increase efficiency through productivity while minimising environmental impairment simultaneously (Arda et al., 2023; Lubis, 2022). Organisational efforts towards becoming environmentally sound could further help to stimulate environmental innovation. This can give rise not only to improved environmental performance but also to an increase in revenue as well as cost reduction. The general understanding, therefore, is that the initiative of a firm to improve its environmental performance is likely to result in a competitive advantage, and also in improved supply chain performance. On the basis of this literature, the following hypothesis was then developed:

A supply chain generally refers to the action of getting a product to the end customer (Jain et al., 2022) with performance measured by the costs, resource deployment and service delivery (Nicoletti, 2023). This study adopts a definition by Green and Inman (2005), who describe this function as the ability of a supply chain to provide quality, quick and dependable service delivery in addition to minimising total overheads for the end customer. It is measured as a unidimensional construct (refer to Figure 1) similar to a study by Mofokeng and Chinomona (2019).

This study predicts green marketing practices, green logistics practices, operational performance and environmental performance to have a salient bearing on supply chain performance. While this conceptual model can contribute to new knowledge by explaining the influence of green practices on supply chain performance with empirical evidence, the literature brings to light that supply chain performance is more dynamic than often perceived. For example, Hanaysha and Alzoubi (2022) emphasise this function as a challenging task that requires managing the relationships between each of the links in the supply chain through customer relationship management and supplier relationship management. However, these authors also highlight the significance of measuring supply chain performance. Supply Chain Performance Measurement (SCPM) functions as an indicator of how efficiently the supply chain system is operating. Using this measure allows for a deeper understanding of the supply chain, which can drive an improvement in overall performance. Over 16 SCPM methods in general practice have been identified with the SCOR Model, which has been acknowledged as the best method, comparatively speaking (cf. Dissanayake & Cross, 2018; Estampe et al., 2013; Gonzalez-Pascual et al., 2021). Created by the Supply Chain Council, the SCOR Model applies a methodology for managing supply chain activities and processes, thus generating a set of practical guidelines for analysing supply chain management practices (Nicoletti, 2023). Supply chain performance therefore is quite dynamic – and this study hopes to contribute towards a more profound understanding of the concept.

Statistics gathered following company background information check indicated that a majority of the manufacturing firms (49.6%) had been in operation between 6 and 10 years, with 39.7% of the total sample situated in the Johannesburg metro. Partnership was the form of business ownership that the majority of profiled manufacturing firms (27.4%) had and 63% of the total reported an annual turnover of between R1 million and R9 million. Only 12.3% of the firms had more than 200 employees. Table 1 provides more information about company background. Statistics about respondent profile indicated that a majority (53.4%) were designated in the operations and manufacturing department, with 42.4% of the total having between 6 and 10 years of work experience. A majority of the respondents (29.8%) were also managers and supervisors.

Alpha coefficients indicated in Table 2 ranged between 0.839 and 0.887, and therefore signified construct reliability according to Kennedy (2022). Tests for CR further yielded results of at least 0.88 CR for all research constructs, thus confirming reliability (≥ 0.7) (Chin, 1998).

In addition to testing the AVE, measurement scales were also assessed against convergent validity and discriminant validity. Table 2 illustrates that standardised regression weights (factor loadings) loaded well (≥ 0.5) on their relative constructs to indicate convergent validity, according to the seminal work of Anderson and Gerbing (1988). The variance extracted estimates for all constructs ranged from 0.56 to 0.69 and thus confirmed validity as they aligned with the ≥ 0.5 threshold advocated by Sarstedt et al. (2014). The presence of discriminate validity was confirmed when the square roots of each construct’s AVE were found to be greater than their respective correlations with other constructs (Yusoff et al., 2020). The results for discriminate validity are presented in Table 3.

Once the measurement model assessment had been concluded, the next step was to measure the model’s explanatory power by ascertaining the coefficient of determination (R²) (cf. Hair et al., 2019). To start with, a standard test for collinearity was conducted to determine if there are two or more predictor variables that are linearly correlated, which would imply that predictors within the statistical model are misconceptualised. Table 4 exhibits the results from the collinearity analysis.

Variance Inflation Factors (VIF) were ascertained and compared against the recommended threshold of ≤ 3 (Becker et al., 2015). All inner VIF values presented in Table 4 confirm collinearity as non-existent.

Given that there were no concerns for collinearity, the next step was to examine the R². The results illustrated in Figure 2 indicates operational performance (0.489), environmental performance (0.484) and supply chain performance (0.498) as having moderate explanatory power (0.25–0.50), respectively, with green logistics practices (0.238) representing a relatively weaker explanatory power (< 0.25) (Hair et al., 2011). A greater statistical power denotes that Partial Least Squares-Structural Equation Modelling (PLS-SEM) is more likely to acknowledge relationships as significant when they are present in the population (Sarstedt & Mooi, 2019).

Examining predictive performance is important for theory building and validation purposes (Sharma et al., 2022). This was performed by calculating the Q², which is a blindfolding procedure that removes single points in the data matrix, credits the removed points with the mean and estimates the model parameters (Sarstedt et al., 2014). Table 5 presents the results following the test for predictive accuracy using blindfolding. Q² indicators with a 0.220–0.336 range can be observed. This suggests that endogenous constructs exhibit a medium predictive accuracy (0.025–0.5) of the structural model (Hair et al., 2019). In other words, the constructs green logistics practices, operational performance, environmental performance and supply chain performance illustrate that this study has a medium chance of accuracy with the hypotheses developed and ultimately the theory adopted.

Partial Least Squares-Structural Equation Modelling maintains that unobserved variables directly or indirectly influence other unobserved variables and that testing results in estimations that indicate how these variables are related (Memon et al., 2021). Table 6 presents the results obtained from path analysis. A positive Beta coefficient (β) implies that there is a strong relationship between unobserved variables (Nohara et al., 2022), while t-statistics and p-values express statistical significance and power (Becker et al., 2013).

The tests in this procedure were performed on a 95% confidence interval to ensure that false positives are averted (Kock, 2016). Accordingly, p-value tests were based on the 0.05 significance level (i.e. 1% – 95%), while t-statistics were compared against a threshold of 1.96 (Kock, 2016). Nine out of the ten hypothesised relationships were levelled at a t-statistic of at least 2.95, with power identified at ≤ 0.05. These hypotheses convey that there is a significant relationship between green practices and supply chain performance within the textile and clothing industry. However, H10 was rejected, suggesting that green logistics practices are, in essence, ineffective in directly improving the supply chain performance of South African manufacturing firms in the textile and clothing industry.

The purpose of testing mediation effect is to learn, for instance, more about potential false effects or suppressor effects (O’Rourke & MacKinnon, 2018). A two-step approach was used for an inclusive mediation analysis (Yang & Liu, 2023). The test for indirect effect was performed in Step 1, followed by a test for mediating strength in Step 2. Table 7 exhibits the results from the mediating effect analysis (Step 1).

Operational performance and environmental performance were tested for mediating effects between green practices and supply chain performance. The results indicated that the four mediation tests equated to a t-statistics ratio of at least 2.91 with power identified at ≤ 0.05. This meant that all four tests for mediation were found to be significant and supported. In addition, the Variance Accounted For (VAF) values were computed in order to ascertain the strength of each mediating effect. The results are illustrated in Table 8 (Step 2).

A VAF value less than 20% suggests a near-zero mediation, and a value greater than 20%, but less than 80% can be characterised as typical partial mediation (Hair et al., 2014). As such, the results presented in Table 8 can be understood as reflecting partial mediations, given that effects conformed to the > 20% and < 80% pragmatism. The two intermediate variables, which are, operational performance and environmental performance, therefore support and complement the hypothesised relationships between the independent variables, green marketing practices and green logistics practices and the dependent variable, supply chain performance.