Operational capabilities are firm-specific sets of skills, processes and routines developed within the operations management system that regularly solve its problems by configuring its operational resources (Zimmermann & Foerstl, 2014). Operational capabilities include time-based management, manufacturing, technology, innovation, lean systems, delivery reliability, process flexibility, logistics integration, information visibility, collaboration capability, supply network flexibility (SNF) and supplier relationship management (Chin et al., 2014; Sandberg & Abrahamsson, 2011). The current study is based on four operational capabilities, namely manufacturing, technology, innovation capabilities, and SNF, and their contribution to firm competitive performance as a strategic solution to the problems hindering the effective response of supply chain operations within the FMCG manufacturing sector in SA. These operational capabilities are selected because they are critical success factors contributing to a firm’s sustained competitive performance by optimising production adaptability, ensuring cross-functional integration, improving agile capabilities and maintaining a seamless and uninterrupted flow of supplies (Chin et al., 2014; Lyu et al., 2019; Mandal et al., 2016; Sandberg & Abrahamsson, 2011). Thus, their strategic adoption cannot be understated because of their impact on firms’ resilience prospects (Lu et al., 2023).

Firms’ competitive performance may be perceived as their ability to implement strategies that reduce costs, limit market competition and exploit opportunities more efficiently and effectively than competitors (Barney, 1991; Prajogo & McDermott, 2008). It is a key element of any firm that measures and differentiates its success from its competitors. It is achieved by optimising key performance indicators (KPIs), such as on-time delivery, competitive pricing and cost, high-quality, correct quantity and flexibility, which are known to be critical measures (Quaye & Mensah, 2017).

Supply chain flexibility (SCF) is the ability of all supply chain networks to adjust to environmental unpredictability and to satisfy an increasing variety of customer demands without incurring high costs, delays, organisational disruptions or performance losses (Piprani et al., 2022). Some notable benefits of SCF include reduced backorders, optimised lead times and increased client satisfaction (Delic & Eyers, 2020).

Supply chain responsiveness (SCR) refers to integrating supply chain partners to respond quickly to changes in the business environment (Williams et al., 2013). It further entails how quickly a supply chain network can respond to changing customer and/or supplier needs. Supply chain responsiveness depends highly on integrating significant factors such as supply chain strategies, postponement and supplier relationships (Rajagopal et al., 2016). Responsive supply chains enable firms to manage market dynamics driven by high demand volatility and rising competition. It allows the sector to respond effectively to customer needs and the marketplace changes (Lurie & Andersson, 2018).

The study’s research model is presented in Figure 1. It consists of four predictors (manufacturing capability, technological capability, SNF and innovation capability), one mediator (firm competitive performance) and two outcomes (SCR and SCF).

The hypotheses depicted in Figure 1 are developed in this section.

This study is anchored on deductive reasoning, which supports the positivist paradigm, and it adopts a quantitative approach to enable generalisation of the results to other FMCG environments. A correlational research design was employed, as the study aimed to test predictions among constructs without testing causality. Some previous studies focusing on similar SCM topics used the same research design (Adeniran & Johnston, 2011; Jain et al., 2017; Nematatani & Chinomona, 2024; Protogerou et al., 2012; Zhou et al., 2019).

The study involved firms operating in the FMCG manufacturing sector in Gauteng province. The updated list of the number of registered FMCG firms was not accessible the Trade and Industrial Policy Strategies (2021) report estimates that there are about 1800 registered food processing companies in SA. Accordingly, the current study’s sample comprised 420 supply chain professionals, including managers, supervisors and specialists drawn from FMCG manufacturing firms. This sample size is consistent with previous similar studies, such as Mafini and Muposhi (2017) (n = 500); Hove-Sibanda and Pooe (2018) (n = 500); Mofokeng and Chinomona (2019) (n = 600). Previous similar studies by Adeniran and Johnston (2011), Huang et al. (2023), and Bharadwaj (2024) used purposively selected respondents with expertise in supply chain. The purposive sampling technique was employed to ensure that the targeted respondents were individuals who had the information relevant to the study.

Data collection was conducted through an online survey using a structured, self-administered questionnaire. The instrument comprised 29 items adapted from previously validated scales. Manufacturing capability was measured using a four-item scale adapted from Wang et al. (2016). A five-item scale adapted from Wang et al. (2004) was used to measure technological capability. Another five-item scale, adapted from Koste et al. (2004), was used to measure innovation capability. Additionally, a six-item scale derived from Moon et al. (2012) was employed to measure SNF, while firm competitive performance was measured using three items developed by Fawcett et al. (2007). Lastly, the SCF scale used three items adapted from Sanchez and Perez (2005), while the SCR construct was measured using three items developed by Zhou and Benton (2007). Response options for the measurement scales were presented on a 5-point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree. Appendix 1 presents the full list of adapted instruments.

Harman’s Single-Factor Test was run to determine if the data were subject to a single dominant factor. Consistent with the original factor structure of the data, the test yielded seven factors accounting for 62% of the variance, with the highest contributing an acceptable 16.241%. As such, no single factor emerged, and the first factor accounted for less than 50% of the variance, indicating that common methods bias was minimal. Additionally, further tests using inter-construct correlations (Table 3) yielded values below 0.9, further confirming that common method variance was insignificant. Procedurally, the measures for the predictor, mediator and outcome variables were adapted from different sources, reducing the possibility of bias in respondents’ views of the observed associations. Moreover, as shown in Table 1, respondents were drawn from various firms and belonged to different demographic categories. Hence, sufficient statistical and procedural measures were implemented to reduce common method variance in the study.

Ethical clearance to conduct this study was obtained from the Vaal University of Technology, Faculty Research Ethics Committee (Ref. No: FRECMS-21102020-051).

The usable questionnaires were entered into an Excel spreadsheet and converted to SPSS format to perform descriptive statistics and an exploratory factor analysis (EFA) using SPSS version 30.0. Confirmatory factor analysis (CFA) was conducted to test the psychometric properties of validity and reliability. The regression analysis to determine the predictive influence of the constructs (hypotheses testing) was conducted using the partial least squares (PLS) composite-based method (PLS-SEM).

Of the initial 700 administered questionnaires, 501 were returned, yielding 420 for the final data analysis, for a response rate of 60%. Of the participating respondents, 61.2% (n = 257) were male, while 38.8% (n = 163) were female. The highest number of respondents were holders of a university degree (n = 127; 30.2%). Additionally, regarding the employment period, the largest number (n = 159; 37.9%) had been employed in their organisations for 3 years to 5 years. Lastly, in terms of professional field, those in the SCM category had the highest number (n = 195; 46.4%). Table 1 presents the demographic profile of the participants.

The CFA was used to test the psychometric properties of the measuring scales. The results are reported in Table 2.

In this study, validity was measured using content and construct validity. A panel of senior lecturers with expertise in SCM ascertained the former. The latter was determined using two aspects: Convergent and discriminant validity.

Convergent validity was measured using the average variance extracted (AVE), CR, and constructs’ factor loadings, in which the cut-off values of the constructs must be equal to or greater than 0.5 as per Fornell and Larcker’s (1981) recommendation. The AVE values of constructs are as follows: 0.716 for MC; 0.718 for TC, 0.905 for IC; 0.789 for SNF (SNF); 0.715 for CP; 0.718 for SCR and 0.698 for SCF. The CR values of the constructs are as follows: For MC, 0.881; for TC, 0.927; for IC, 0.966; for SNF, 0.957; for CP, 0.883; for SCR, 0.884; for SCF, 0.873. Lastly, MC factor loadings ranged from 0.760 to 0.989; TC ranged from 0.803 to 0.889; IC ranged from 0.939 to 0.964; SNF (SNF) ranged from 0.847 to 0.930; CP ranged from 0.828 to 0.879; SCR ranged from 0.826 to 0.877; and SCF ranged from 0.706 to 0.924, respectively. Therefore, convergent validity between the constructs is confirmed. Besides, the study used the Fornell-Larcker criterion to assess discriminant validity, which requires that the AVEs of the constructs be less than 0.90 to confirm discriminant validity (Ab-Hamid et al., 2017). Table 3 presents discriminant validity results using the Fornell-Larcker criterion.

The study’s hypotheses (path analysis) were analysed through SEM using the composite-based technique (PLS-SEM). The path analysis was determined by assessing the path coefficients through the beta (β) score to support or reject the study’s hypotheses. Figure 2 presents the path analysis model.

Figure 2 shows that four predictor constructs (MC, TC, SNF, IC) accounted for 76.4% of the variance in CP (r² = 0.764). Moreover, CP contributed only 2.6% (r² = 0.026) of the variance in SCF and 6.8% (r² = 0.068) of SCR. The remaining variances in each case are therefore attributed to factors outside the scope of the current study. Table 4 presents the hypothesis-testing results (the beta coefficient and its significance).

All six hypotheses put forward in the study were accepted based on significance levels (t ≥ 1.96; p ≤ 0.05). The outlier among the results is the inverse association observed between MC and CP (β = −0.265; t = 3.013; p = 0.003), indicating that CP decreases as MC improves.

In assessing the association between the predictor and mediator constructs, all four hypotheses were supported because of the reported significance. First of all, MC exerted a significant negative effect on CP (β = −0.265; t = 3.013; p = 0.003). This interesting result contrasts with most previous research (e.g. Bag et al., 2021; Cheng & Lu, 2022; Jin et al., 2014; Narenji et al., 2025; Yam et al., 2010), which has established positive associations between the two constructs. A possible reason for this uncommon result is the availability of alternative products on the largely monopolistically competitive FMCG market in South Africa. As suggested by Hweshure (2022) and Hirsch et al. (2024), the FMCG sector in SA is one of the most saturated markets, characterised by fierce competition among numerous brands that offer affordable options for consumers. This high congestion underlines the lower barriers to entry in this market, such that even small players are also introducing their own products. A common example is flooding the market with cheaper Chinese imports that appeal to price-sensitive consumers, who are the majority in SA, a country facing various economic hardships (Gupta, 2016). As a result, under such conditions, improvements in a firm’s MC may not necessarily lead to enhanced competitiveness, as observed in the current study.

Unlike MC, the three other predictors (TC, IC and SNF) significantly influenced CP, indicating a positive connection. These were H2 (TC and CP [β= 0.500; t = 5.079; p = 0.000]), H3 (IC and CP [β= 0.605; t = 8.898; p = 0.000)]) and H4 (SNF and CP [β= 0.299; t = 25.190; p = 0.000]). All three hypotheses were accepted. The results are consistent with previous studies (Brusset, 2016; Chang, 2011; Kim et al., 2020; Palandeng et al., 2018; Reklitis et al., 2021; Shou et al., 2017; Tiwari et al., 2015), which found that IC, TC, as well as SNF, are critical input resources for sustained competitive success.

These results establish innovation, technology and SNF as critical driving factors of CP, while singling out IC as the most potent input factor among the three. This view is particularly important given the strategic role that innovation capabilities play in enhancing operational excellence in today’s technology-dominated environments (Chirra & Raut, 2025; Fatoki, 2019; Kusi-Sarpong et al., 2021; Tiwari et al., 2015). As such, FMCG manufacturing firms should align their innovation strategies with evolving consumer trends and market opportunities. This strategy is possible through significant investment in research and development (R&D) and redesigning operational models. Such shifts will facilitate improved innovation during product development (organic or sustainable product offerings), leading to better products tailored to meet unpredictable customer needs.

Since TC also emerged as a positive enabler of CP, the results suggest that adopting recent technologies could boost the competitive positions of FMCG manufacturing firms. For example, embracing IT systems such as Oracle and SAP could enable the automation and integration of data and information across value and supply chain networks of FMCG firms in SA. The use of such systems could enhance the visibility, monitoring, alertness and control of their supply chain functions, optimising sound decision-making and facilitating rapid responses to market disturbances.

With respect to SNF and its impact on CP, the results suggest that optimising the SNF, such as through adopting multiple sourcing strategies, could enable firms to withstand macro-environmental shocks (Mwikali & Kavale, 2012). Such a multifaceted sourcing approach minimises the reliance on a single supplier and provides the necessary flexibility to ensure a seamless supply across a firm’s supply chain ecosystems. Firms within the FMCG sector may then be able to meet their customers’ expectations and maintain their competitive edge. Hence, diversification of their supply bases is a viable strategic option for firms in the FMCG sector to succeed in their supply chains.

The results for H5 revealed that CP significantly influences SCR (β= 0.261; t = 4.875; p = 0.000). This finding, therefore, implies that the competitive performance of the FMCG supply chain depends, among other things, on its responsiveness. The results support previous studies (Aslam & Li, 2025; Brusset, 2016; Gupta et al., 2019; Lee, 2004; Mee-ngoen et al., 2020) that found that competitiveness positively influences responsiveness. These results suggest that the competitiveness of FMCG firms builds their agile capabilities, facilitating improved responsiveness to market trends and variations in consumer demand. As such, FMCG firms should endeavour to boost their responsiveness by emphasising competitive excellence, achievable through approaches such as the use of state-of-the-art IT tools and systems, diversifying their distribution networks through omnichannel strategies and fostering strategic partnership agreements with key players.

Similarly, CP was positively associated with SCF (H6) (r = 0.162; t = 3.640; p = 0.000). This finding further denotes the role that firms’ competitive success plays in securing flexible supply chains. Consistently, previous studies by Gligor and Holcomb (2012) and Yu et al. (2014) found a similar connection between these two factors. By implication, highly competitive FMCG firms are capable of handling sudden changes in market and consumer demands. Typically, this is because highly competitive firms can redesign, readjust and reconfigure their production, distribution and product offerings to align with customers’ and consumers’ evolving demands.

The study also tested for the mediating effects of CP on the association between operational capabilities and SCF and SCR. The results are presented in Table 5.

The results of the mediation analysis show that CP mediates the association between all four operational capabilities and the two outcome variables, SCR and SCF. For MC, the significant mediating effect is inverse, indicating that CP decreases the influence of MC on SCF (β = −0.052; t = 2.930; p = 0.004) and SCR (β = −0.083; t = 3.695; p = 0.000). However, the margins of decrease, 5.2% and 8.8%, respectively, are quite marginal and almost inconsequential. The rest of the significant mediating influences of CP are positive, with the highest occurring in the path between SNF -> CP -> SCR (β = 0.155; t = 4.629; p = 0.000). This indicates that CP improves the influence of SNF on SCR by 15.5%. Overall, and in practice, the results illustrate that strengthening CP is essential when using the three operational capabilities (TC, IC, SND) to improve SCF and SCR in the FMCG manufacturing sector.

The mediation model presented in Figure 3 indicates the results of the analysis.

The study aimed to investigate the relationships among operational capabilities, firm competitive performance, supply chain responsiveness and flexibility in the FMCG manufacturing sector in SA. This study reveals that operational capability practices, particularly innovation, technology and SNF, are important drivers of CP in the sector. However, the influence of manufacturing capability on competitive performance is negative, indicating that the two factors move in opposite directions. The study further shows that how a firm in the FMCG manufacturing sector performs competitively affects the supply chain’s flexibility and responsiveness. Additionally, competitive performance mediates the associations among technological capability, innovation capability, and SNF with SCF and SCR, demonstrating a positive stimulus effect of competitive performance on these connections.

The study has some theoretical and practical contributions. Theoretically, the study broadens the SCM literature by articulating the roles and importance of operational capabilities, their links to competitiveness, and SCF and SCR. Furthermore, the study tests the research model in the FMCG manufacturing sector in SA, where there is limited empirical evidence of such nature. This makes the study a groundbreaking attempt to extend this model to this vital economic sector within SA.

In practice,. these findings identify key success factors that must be adopted by SCM professionals operating in the SA FMCG manufacturing sector to optimise their firms’ competitive performance, supply chain flexibility and responsiveness. The study further shows that an emphasis on operational capabilities, such as innovation, technology and supply diversity, is a critical catalyst for achieving their competitive aspirations. Nurturing competitiveness is also essential to achieve an agile FMCG supply chain that is both flexible and responsive to market dynamics.

The quantitative data used in the study were collected in one SA province, Gauteng, which limits the generalisability of the findings. Future studies could therefore broaden the scope of the study by including FMCG manufacturing firms from other provinces in the country. This will effectively confirm whether the same positive connections found in most of the hypotheses in the current study are consistent across various regions of SA. The study is further limited by the lack of direct relationships between the four operational capabilities (manufacturing capability, technological capability, supply chain network flexibility and innovation capability) and firm competitive performance. As indicated in the mediation model (Figure 3), future studies can then expand the current results by testing these direct relationships as well. Future studies could also consider the moderating effects of demographic factors, such as firm size or turnover, to determine how these factors influence the associations examined in the study. Additionally, the model could be expanded by including other operational capabilities, such as lean systems, delivery reliability, process flexibility, logistics integration, information visibility, collaboration capability and supplier relationship management, that were excluded in the current study.