Impact of Business Model Innovation on Sustainable Performance of Processed Marine Food Product SMEs in Thailand—A PLS-SEM Approach

Abstract

This study aims to develop four conceptual higher order models for assessing the causal relationship between the environmental turbulence (ET), sustainable competitive advantage (SCA), business model innovation (BMI), and sustainable performance (SP) of small and medium sized enterprises (SMEs). The conceptual models were developed through literature review and tested with 91 entrepreneurs and managers from processed marine food product SMEs in Thailand. The higher order models were tested with partial least square structural equation modeling using seminr package in R. The results reveal that SCA mediates the relationship between BMI and SP. Further, the study found a serial mediation effect of BMI and SCA in the relationship between ET and SP. Thus, this study is novel in its approach of using ET as an antecedent and moderator and SCA as a mediator in assessing the relationship between BMI and SP. The study also found that the effects of ET are balanced when the SMEs incorporate BMI, which further leads to the achievement of SCA and SP. Thus, the findings extend an increasingly complex literature in the assessment of SCA’s role in SMEs’ SP. Further, the scale and the models can be used to assess how the SMEs respond to ET and modify their BMI to attain SCA and SP.

1. Introduction

The marine food processing sector is a crucial sector in Thailand, contributing significantly to the national economy. In terms of international trade, Thailand is the fifth largest processed seafood exporter in the world. Several factors influence Thailand’s competitive advantage over other ASEAN countries in the marine food processing sector. The top three significant factors are its (a) geographical location, (b) the availability of its skilled labor, and (c) its well-developed industrial cluster with an active role for small and medium sized enterprises (SMEs) [1,2]. The structure of the seafood processing sector in Thailand is divided into “chilled and frozen seafood processing and canned seafood, prepared and preserved seafood” [2] and is concentrated in the Samut Sakhon and Samut Songkhram provinces of central Thailand [3]. Overall, SMEs play a significant role in the economy and contributed 35.3 percent to the national GDP in 2019 [4]. This pattern of SMEs’ reasonable share in the sectoral contribution is also reflected in the marine food processing industry. The next crucial factor for the sustained success of this sector is the support extended by the government through the training for the adoption of advanced manufacturing technologies, improving food quality standards, research, and food innovation [5]. So, all these aspects support SMEs in meeting domestic and global demand, resulting in increased competitiveness. Thus, the Thai seafood processing sector has undergone various changes in the past decade through adopting product, process, and positioning innovations [6,7]. Currently, the seafood processing sector in Thailand has a substantial value in production, distribution, and consumption, with a significant market share domestically and globally. However, the SMEs in this sector lack effective implementation of technology and innovation as their business planning process is limited to immediate customer needs [6]. Therefore, the SMEs operating in the dynamic business environment are unprepared for crises or industrial transitions, making it difficult to expand their business operations internationally and survive in global competition.

The National Economic and Social Development Board (NESDB) has emphasized that entrepreneurs must inculcate business skills and knowledge for increased business cooperation with neighboring countries. Further, the NESDB, in their Twelfth National Economic and Social Development Plan (2017–2021), highlighted the need for sustainable improvement in SMEs’ products and services through technology adoption and innovation [8]. This is because if the SMEs fail to adopt the technology and innovation, they may fall behind the competition and fail to meet international standards [9]. Hence, SMEs must embrace science, technology, innovation, and research to add value to their existing products [8]. Further, SMEs should focus on improving their business model to utilize the opportunities and create added value to meet the supply and demand conditions [9]. The core functions of the business model are creating, delivering, and capturing values from/for the market [10,11,12,13]. Therefore, apart from other entrepreneurial skills, it is essential for entrepreneurs/SMEs to constantly focus on innovating their business model to prevent the risks associated with the ‘changing needs’ of customers, changes in the business environment, and industrial and technological advancements. However, business model innovation should not be understood as technological innovation; it is the “reinvention of the business” itself. Thus, SMEs can also internationalize their business operations by focusing on business model innovation (BMI) [10,13,14]. This is also considered a crucial factor in business operations for delivering and capturing value [11]. It can make radical transitions in the existing process of the firms and enables the firm to stay competitive and perform efficiently in the long term. Hence BMI is an essential business strategy for processed marine food product SMEs, one which can facilitate their achievement of sustainable competitive advantage (SCA) and sustainable performance (SP).

Environmental turbulence (ET) may or may not positively or negatively impact firm performance [15]. However, this depends on the firm’s response to dynamic environmental conditions, which may lead them to achieve SCA and SP. Moreover, SMEs are forced to focus on innovating their business models frequently due to turbulent external environmental conditions, which may affect their BMI and performance.

There are prior studies that have focused on measuring the impact of BMI and firm performance (FP) [16,17,18], the relationship between BMI and SP [19], ET and SCA [20,21], and SCA and SP [22]. However, to the best of the author’s knowledge, the unified approach to the assessment of the relationship between ET, BMI, SCA, and SP has not been tested so far. Further, the scales in the present literature for assessing sustainable performance are also limited to either financial or operational performance or both. Nevertheless, there is no unified approach to the use of economic, environmental, social, and operational dimensions to assess the sustainable performance of marine product SMEs. Hence there is a gap in the literature in regard to the adoption of a unified, comprehensive framework for measuring BMI and SP in the context of marine product SMEs through all sustainability dimensions of UNSDG. Also, in recent years SMEs have been suggested to inculcate sustainability within their business practices [23] due to (1) the sustainability policy initiatives of the government, (2) the internal and external pressures from stakeholders, (3) the increasing customer pressures for environmentally friendly products, and (4) the rising opportunities for innovation and growth [23,24]. However, a paucity of studies have focused on investigating the SP of SMEs with four dimensions of sustainability (economic, environmental, social, and operational). Even though there are studies which include the word ‘sustainability’ to measure the firm performance, they focus only on financial and non-financial aspects in the long run [25] and the firm’s economic performance [26,27]. Therefore, the environmental and social performance of SMEs is largely understudied. With due consideration to the importance of studying sustainability through the UNSDG dimensions, this study proposes an empirical investigation for measuring the impact of BMI on SP. Also, from the above discussion, this study draws an equation for measuring the causal relationship between BMI, SCA, SP, and ET using a unified and comprehensive approach, which has not been investigated by prior studies. Hence, this research tests the causal effects of ET as a moderating variable (in the causal relationship between BMI → SCA, and BMI → SP) and antecedent of BMI; and further tests the causal effects of BMI as an independent variable (for measuring it’s causal effects on SCA and SP), and mediating variable (between ET → SCA, and ET → SP).

Thus, the major objectives of this study are:

(O.1) to develop and validate the measurement scale and test the hypothesized models for measuring BMI, SCA, SP, and ET from the literature,

(O.2) to analyze the impact of BMI on SCA and SP of processed marine food product SMEs in Thailand,

(O.3) to identify the effects of ET on BMI, SCA and SP, and

(O.4) finally, to analyze the serial mediation effects in the relationship between ET and SP.

Considering the research gap in the literature for adopting a unified approach in the context of SMEs, this research is directed to address the following research questions (RQ):

RQ1: Does BMI impact SCA?

RQ2: Does BMI impact the SP of SMEs?

RQ3: Does ET moderate the causal relationship between BMI and SCA?

RQ4: Does ET moderate the causal relationship between BMI and SP?

RQ5: Does ET have an impact on BMI and SCA?

RQ6: Are there any serial mediation effects in the relationship between ET and SP?

Based on the research questions and research objectives, the remaining sections of the manuscript are organized as follows: literature review, methodology, data analysis and interpretation, findings and discussion, limitations and directions for future research, implications, and conclusion.

2. Literature Review

The literature review section is divided into two sub-sections. The first sub-section focusing on the latent variables (BMI, SCA, SP, and ET) and the second sub-section focusing on the causal relationship between them.

2.1. Latent Variables

2.1.1. Business Model Innovation (BMI)

BMI has gained more attention from researchers, practitioners and policymakers and has grown in the past two decades [28]. A critical review of fifteen years research on BMI [29] found that it has several definitions and frameworks. However, it has insufficient theoretical foundations, the preciseness of the construct and its influencing antecedents, moderators, and outcome variables have, particularly, not been linked effectively. Therefore, there is a literature gap in the distinct theory of BMI and its causal relationship towards antecedents and subsequent factors [29]. Though there are numerous definitions and conceptualizations of BMI, there is no agreement among scholars on the concept of BMI itself. Mostly these definitions are adopted based on study purpose [30]. Thus, there are at least three approaches for defining the business model (BM). They are the (1) system, (2) business process, and (3) business management approaches [31,32,33]. In the system approach, BM is conceptualized as a “system of interdependent activities” [31]. In the business process approach, it is defined as the combination of people (who), place (where), time (when), product (what), preferences (why), transaction (how), and transaction volume (how much) required and involved in offering end-users and customers with the services/products [32]. In the business management approach, BM is conceptualized as the scientific process of decision making through hypotheses testing and informed decision making [33]. In any case, BM leads to innovation in terms of organization process, new product development, business process improvements, new collaborations for business efficiency and cooperation for business development [30]. However, there is a conceptual difference between defining BM and BMI, because BMI is limited to non-trivial changes [29]. Therefore, it is not about introducing new technologies in production or expanding to new markets. It is all about better delivery of the available products/services to the currently served market with the available technology, which is hard to replicate by the other firms in the same industry [34]. So, BMI is referred to as improving an element in the business model to enhance the company’s performance. It is also referred to as replacing the business model elements by offering products and services to the customers that were not available before [32]. Thus, the conceptualizations and definitions of business model innovation vary based on the approaches and objectives of the study. However, the definition of Zott and Amit [31], which conceptualizes the BMI based on ‘content’, ‘structure’, and ‘governance’ of transactions, was found more suitable for this study because it is defined based on the notion of entrepreneurial firms. Based on the several definitions, it is understood that BMI is something beyond the product, process, technological, and organizational innovation. It also involves finding new ways of innovating the business model itself. Thus, it is evident that BMI involves more knowledge and constant improvements than investments in technology and research. In the simplest terms, BMI is a new way of doing business that creates, delivers, and captures value [10,11,12,13], which is hard to replicate. Hence it is essential for the firms to review and modify BMI frequently to sustain themselves in competitive and hard economic times because BMI is the most influential source of competitive advantage [33] and it improves the performance of SMEs [35]. The two critical design themes of BM were identified and proposed by Zott and Amit [31], namely novelty-centered and efficiency-centered. Novelty-centered BM design focuses on conceptualizing and adopting new ways to carry out economic activities by doing things that were not done before. For instance, connecting with new parties, new ways to link participants, and devising new transaction mechanisms. On the other hand, efficiency-centered BMI reduces the transaction costs of the participants. Further, several studies [12,31,36,37,38,39] have suggested that BMI dimensions should be centered on value propositions, novelty, processes, transaction structure, resource structure, and interactions of the firm with its key stakeholders. The value proposition element positioned on novelty is measured based on its strength and the extent it is maintained under different business environments and conditions [16,31,39]. The novelty dimension in this study not only focuses on value creation through new products, services, markets, and partners [12,35,38] but also on ideas, innovations, processes, and actions to retain the customer base. The interactions of the firms with their stakeholders are measured in terms of how the SMEs maintain customer relationships and links with suppliers and partners who were not connected before [12]. Further, the interaction structure measures SMEs’ transactions and resource structure [39]. Therefore, the transaction structure dimension explains the flow of materials, goods, services, money, and information exchange [12,38] in the entire supply chain and its scope for improvement from time to time. However, the resource structure of BMI in this study explicitly points out the knowledge and capabilities of the firm to react to the market changes rather than the tangible resources. In SMEs, transaction and resource structures interact to create and capture value from a primary commercial opportunity [37]. Further study also included the pricing strategy, a major element for assessing the total value captured [36,37,39], and reviewing cost structure to adjust to the market environment changes for profitability [39]. Furthermore, SMEs can adopt novel and efficient business models simultaneously [31] or may use an ambidextrous approach. Thus, the BMI construct adopted in this study was based on previous studies [16,31,35,38] and is centered on novelty (value propositions), and it imparts the efficiency aspects of the business operations (interactions, transactions, resources, pricing, and cost) for creating, delivering, and capturing value; these aspects have evidence of strong theoretical foundations for their measurability in the context of SMEs.

2.1.2. Sustainable Competitive Advantage (SCA)

SCA, introduced by Michael E. Porter and Jay Barney [40,41], is one of the most cited theories in strategic management literature [41]. The competitive advantage is crucial for the firms to achieve profits. The firms can achieve it by adopting a combination of generic ‘competitive strategies’ [40], namely ‘differentiation’, ‘cost leadership’, and ‘focus’. Such strategies determine the firm’s position within an industry based on its level of profitability compared with the industry average [40]. However, sustaining a competitive advantage is challenging because achieving long-term success is a major goal of any firm in the world [42,43]. A firm might achieve SCA as a result of implementing a value creation strategy not duplicatable by the potential competitors. The sustained competitive advantage does not mean it will “lasts forever”; again, it depends on an industry’s economic structure [41]. On the other hand, Porter referred to SCA as a firm’s “above-average performance” compared with the industry average over time. However, the fact is that if the generic strategy is not sustainable, it will not lead to the attainment of above-average performance in the industry [40] because if the competitive advantage is based only on the profitability and those generic strategies of the firm are easily imitable by the competitors, then all the firms in an industry may be able to attain above-average performance. So, the firms must make efforts to develop a strategy that cannot be imitated by their rivals and sustain their position over time [40,42]. Such efforts require firms to invest in knowledge, internal resources, and capabilities [42]. Thus, firms must look into sources of SCA within the internal environment to exploit opportunities and compete in the external environment. In congruence with a resource-based view, the SCA is obtained by creating or developing the resources and capabilities of a firm in response to the dynamic market environment [42,44]. So, it is important to find out the firm’s internal capabilities and resources that act as a source of SCA. Also, it does not imply that all the potential firm resources would assist the firm in obtaining SCA [41]. Thus, the entrepreneurs or managers should discover their appropriate resources and capabilities within their firm and analyze those resources in terms of ‘value’, ‘rareness’, ‘imitability’, and then exploit them through their ‘organization’ (VRIO) [41,44]. Several studies [22,43,45,46] have also emphasized and discussed the importance of ‘value’, ‘rareness’, ‘inimitability’, and ‘organization’ in measuring sustainable competitive advantage and have adopted them in their study. Therefore, this study was constructed with a focus on the attributes of value, rareness, inimitability, and organization for measuring the SCA of SMEs. In addition, this study also included location as an additional variable, which can ease the manufacturing (coastal areas and aquaculture farms), and logistics aspects (business cluster, port infrastructure, port accessibility, and logistics cost reduction) of the firms [40].

2.1.3. Sustainable Performance (SP)

In recent times, various academic fields have witnessed myriad research interests in sustainability and its dimensions. Sustainability has become a keen interest among researchers and practitioners, and sustainability research has dominated the entrepreneurship and strategic management literature [26].

Various dimensions have been utilized for measuring the sustainable performance of a firm, with a particular interest in measuring financial and non-financial aspects [25]. SP is defined as “the extent to which SME entrepreneurs can sustain the economic and social performance in the long run” [47]. Even though SMEs have the capability and potential to adopt sustainable practices, they lack the knowledge to capture long-term opportunities [48]. Earlier studies identified how social and environmental impacts are not realized by the SMEs due to a lack of resources, time [49] and knowledge [48]. However, in recent years, SMEs have been suggested to inculcate sustainability within their business practices [23]. Also, in the past, many studies have addressed only a firm’s economic performance, while there is a paucity of studies addressing social and environmental performance [26,27]. The performance outcomes of the firm have also been studied from three dimensions, namely ‘environmental’, ‘economic’, and ‘operational’ [24,50]. However, those studies also have limitations because the social dimension was not included. Only limited studies have agreed on using the ‘economic’, ‘environmental’ and ‘social’ dimensions of sustainability in the context of entrepreneurship and SMEs [26,51,52], which place more weight on moral and ethical issues in business practices [53]. From the above literature, it is clear that there is no consensus among scholars on the dimensions for measuring the SP of SMEs. Thus, this study emphasizes the sustainability dimensions of SDG 12 for measuring the sustainable performance of SMEs. The three sustainability dimensions ‘economic’, ‘social’, and ‘environmental’ were considered for measuring the SP of SMEs. Also, the ‘operational performance’ [7,50,54] dimension was included as the fourth dimension, along with the three major dimensions for measuring sustainable performance [27]. Various items were included for the four dimensions of SP based on SDG 12, a literature review, and processed marine food product sectoral specifications (SMEs context). Based on the above literature, the variables for economic, social, environmental, and operational performance were selected as follows.

The variables for economic performance [26,47,50,52,55,56] included in this study were ‘decrease in the cost of spending on resources’, ‘efficient utilization of resources’, ‘employment’, ‘return on investment (ROI)’, ‘income stability’, ‘sales growth’, ‘cash flow’, and ‘market share’. The variables for social performance [26,52,55] included in this study were ‘labor practices’, ‘well-being of workers and the local community’, ‘reducing negative business impacts’, and ‘building brand image in the local community’. The variables included in the environmental performance [26,50,52,55] were ‘reduction in air emissions’, ‘wastewater reduction’, ‘solid waste management’, ‘environmentally friendly usage of energy and water’, ‘reduction in accidents’, and ‘maintaining hygiene’. The variables included in the operational performance [7,47,50,56] were an increase in-‘on-time delivery of goods’, ‘product line’, and ‘customer satisfaction’; a decrease in-‘inventory levels’, ‘scrap rate’, and ‘order processing time’, and improvements in–‘capacity utilization’, ‘supplier and customer relationship’, and ‘customer retention’.

2.1.4. Environmental Turbulence (ET)

Several environmental conditions affect business performance. SMEs face ET similar to large firms, which affects their major decisions [57]. Since the term ‘environmental conditions’ implies a broader meaning, this study chose the set of well-defined ‘environmental conditions’ of ‘market turbulence’, ‘technological turbulence’, and ‘competitive intensity’ [58,59] from previous studies. Among those, the first aspect, ‘market turbulence’, refers to the proportion of changes in customer expectations in terms of their changing needs, tastes, and preferences. The second aspect, ‘competitive intensity,’ refers to the intensity of rivalry and the ability and actions of rival firms within a particular industry. Finally, the third aspect, ‘technological turbulence,’ implies that firms face high pressures within their industry to adopt burgeoning technologies. Thus, environmental turbulence indicates that the degree of changes arises from the external environment, which is unpredictable and dynamic and affects the overall business performance [60]. Environmental turbulence affects a firm’s performance both negatively and positively. The negative effects are ‘pressures from customer markets’, ‘competitor markets’, and ‘new technologies’; and the positive effects can be realized with ‘new opportunities’ (new products, services, processes, technology, business models, and competitive advantage) [15]. Firms should therefore plan effectively to respond to this turbulence.

Among these three aspects, a prior study [60] used ‘market turbulence’ and ‘technological turbulence’ to explain the effects of ET in less competitive oligopolistic sectors such as telecommunication. However, all three of the dimensions (‘market turbulence’, ‘competitive intensity’ and ‘technology turbulence’) have been found to be suitable for measuring environmental turbulence in highly competitive markets with perfect competition, something which has been proven by various studies in SMEs contexts [61,62,63,64]. Therefore, this study adopts all three dimensions mentioned above for assessing the ET of processed marine food product SMEs in Thailand.

2.2. Concept Models and Hypotheses Development

The concept models (Figure 1, Figure 2, Figure 3 and Figure 4) and hypotheses were developed in accordance with the study’s objectives and previous theoretical postulations. Four concept models were developed to test the propositions. The major latent variables used for the study are BMI, SP, SCA, and ET. The variables BMI, SP, and SCA, were used as independent, dependent, and mediating variables to address research questions 1 and 2. Further, ET was used as a moderating and independent variable to test its effects under different conditions and answer the research questions 3, 4, 5, and 6. Thus, the four concept models used in this study are presented in Figure 1, Figure 2, Figure 3 and Figure 4, and the literature on the causal relationships is provided in this sub-section.

2.2.1. The Causal Relationship between BMI and SP

BMI is a valuable source that can enhance SP [65]; the positive causal relationship between BMI and FP is well established [16,38,66]. However, a paucity of studies investigated the SP of SMEs with four dimensions of sustainability. Hence, this study proposes empirically investigating the impact of BMI on SP. It has already been proved that the BMI of the firms based on novelty and efficiency has a positive impact on (start-up enterprises, SMEs, and large firms) SP [19]. However, the variables utilized for measuring SP covered limited aspects (‘sales & profit’; ‘cost & cost reduction’; ‘risk & risk reduction’; ‘attractiveness as employer’; and ‘reputation & brand value’). Hence, this study measures the impact of BMI and SP as a whole, which would facilitate SMEs to survive in the competitive scenario by achieving various economic, social, environmental, and operational benefits. From the literature review, it is evident that there are very limited studies measuring SMEs’ BMI and their SP. Subsequently, the below hypothesis was framed.

Hypothesis 1:

Business model innovation (BMI) of processed marine food product SMEs has a significant and positive impact on sustainable performance (SP).

2.2.2. The Causal Relationship between BMI and SCA

The positive impact of BMI in attaining competitive advantage has been widely reported [16,32]. Nevertheless, only limited studies have empirically investigated how BMI facilitates the firms in sustaining their competitive advantage because BMI involves value creation and implementing such value creation leads to achieving SCA [41]. To the best of the authors’ knowledge, only one study [20] has empirically investigated and reported the positive causal effects of BMI on SCA in SMEs. Also, only one study [46] has investigated the SCA using VRIO in food sector SMEs. Thus, the conceptualizations and empirical investigations of the causal relationship between BMI and SCA in the context of processed marine food product SMEs are lacking. Subsequently, the below hypothesis was framed.

Hypothesis 2:

Business model innovation (BMI) of processed marine food product SMEs has a significant and positive impact on sustainable competitive advantage (SCA).

2.2.3. The Causal Relationship between SCA and SP

The theory on the relationship and impacts of competitive advantage on firm performance is well established in the literature under various contexts [16,40,67]. Nevertheless, limited studies have empirically examined the casual impact of SCA on SMEs’ SP and proved the positive and significant impact of SCA on SP [22]. However, the study focused only on the financial aspects of SP. Thus, this research conceptualized the following hypothesis covering the VRIO aspects of SCA and four dimensions of SP (economic, environmental, social, and operational).

Hypothesis 3:

Sustainable competitive advantage (SCA) of processed marine food product SMEs has a significant and positive impact on sustainable performance (SP).

2.2.4. The Mediating Role of SCA in the Causal Relationship between (a) BMI and SP; and (b) ET and SP

The mediation effects of competitive advantage in the relationship between BMI and firm performance have already been studied [16]. The study results indicate partial mediation of competitive advantage in the causal relationship between BMI and firm performance. Some studies investigated the causal relationship between BMI → SP [19], ET →SCA [20,21], and SCA → SP [22]. Therefore, this study draws an equation (refer to concept model C) for measuring the relationship between BMI, SCA, and SP and ET, SCA, and SP using a unified and comprehensive approach; one which has never been investigated by prior studies. Thus, the following hypotheses were constructed to analyze the mediating role of SCA in the causal relationship between BMI → SP; and ET → SP.

Hypothesis 4:

Sustainable competitive advantage (SCA) of processed marine food product SMEs mediates the relationship between their business model innovation (BMI) and sustainable performance (SP).

Hypothesis 5:

Sustainable competitive advantage (SCA) of processed marine food product SMEs mediates the causal relationship between environmental turbulence (ET) and sustainable performance (SP).

2.2.5. The Moderating Role of ET in the Causal Relationship between (a) BMI & SP and (b) BMI and SCA

The moderation effects of ET in the causal relationship between the latent constructs have been well established and documented in the literature [60,62,68,69,70]. The motivation for adopting ET as a moderating variable for studying the relationship between BMI and SP was obtained from the recommendations of previous studies [31,62]. The prior study findings indicate that the factor ET does not moderate the relationship between BMI and firm performance. However, the moderating role of ET in the relationship between BMI and SCA and BMI and SP has not been largely investigated. Thus, the subsequent hypotheses were framed to test the moderation effects in both relationships.

Hypothesis 6:

Environmental turbulence (ET) significantly moderates the causal relationship between business model innovation (BMI) and sustainable performance (SP) of processed marine food product SMEs.

Hypothesis 7:

Environmental turbulence (ET) significantly moderates the causal relationship between business model innovation (BMI) and sustainable competitive advantage (SCA) of processed marine food product SMEs.

2.2.6. The Causal Relationship between ET and BMI

ET is the major factor affecting SMEs’ BMI positively and negatively [15]. Recently, the positive impact of the external environment (competitive intensity and market turbulence) on BMI in the context of MSMEs was studied [71], and a significant positive impact of ET on BMI was also reported [20]. Thus, the following hypothesis was conceptualized to study Thailand’s processed marine food product SMEs.

Hypothesis 8:

Environmental turbulence (ET) positively impacts on business model innovation (BMI) of processed marine food product SMEs.

2.2.7. The Causal Relationship between ET and SCA

ET happens when there are rapid changes in the market, competition, and technology [58]. There are contradicting results in terms of the causal relationship between ET and SCA. There are studies that have reported a significant positive relationship [20] and some that have reported an insignificant relationship [21] between ET and SCA. However, they did not adopt the unified approach framed in this research. Hence, the following hypothesis was framed.

Hypothesis 9:

Environmental turbulence (ET) positively impacts the sustainable competitive advantage (SCA) of processed marine food product SMEs.

2.2.8. The Mediating Role of Business Model Innovation (BMI) in the Causal Relationship between (a) Environmental Turbulence (ET) and Sustainable Competitive Advantage (SCA); and (b) Environmental Turbulence (ET) and Sustainable Performance (SP)

A few studies have utilized BMI as a mediator for assessing the causal effects on SCA [20,72] and firm performance [73,74]. Hence, this study developed the concept model of BMI with the antecedent, moderator, and outcome variables [29]. It also adopted the conceptualization of BMI [20] as a mediator for testing the relationship between industry turbulence (antecedent) and SCA (outcome). Because the outcome of the study [20] confirmed the impact of industry turbulence on BMI and the impact of BMI on SCA, it further proved the partial mediation effects of BMI in the causal relationship between industry turbulence and SCA. Furthermore, BMI was also conceptualized [75] as a mediating variable to measure the impact of ET on firm performance. Thus, this research alternatively tested BMI as a mediator, assigned ET as the antecedent (refer to concept model D), and further assigned the outcome variables as SCA and SP.

Hypothesis 10:

Business model innovation (BMI) of processed marine food product SMEs mediates the causal relationship between environmental turbulence (ET) and their sustainable competitive advantage (SCA).

Hypothesis 11:

Business model innovation (BMI) of processed marine food product SMEs mediates the causal relationship between environmental turbulence (ET) and their sustainable performance (SP).

2.2.9. The Causal Relationship between ET and SP

The conceptualization of the causal relationship between ET, BMI, and firm performance is evident from a prior study [75]. Also, the impact of ET on firm performance has been well documented in previous studies [62,76]. Further, the effects of mediating variables in the causal relationship between ET and the firm’s performance have been well documented [62]. However, the impact of ET on SP has not been tested in the context of SMEs. Furthermore, since ET and SP are proven to be the antecedents and outcome variables between BMI and SCA, there are possibilities of the serial mediation of ET on SP through BMI and SCA, which has never been empirically investigated. To the best of the author’s knowledge, only one study [20] has investigated the effects of ET on BMI and BMI on SCA. However, that study was limited to the service sector. At the same time, another study [22] recorded the impact of SCA on SP. However, the SP of that study was limited to the financial aspects of SMEs. Subsequently, the below hypotheses were proposed to measure both the direct and indirect effects of ET → SP, using BMI & SCA as serial mediators.

Hypothesis 12:

Environmental turbulence (ET) positively impacts the sustainable performance (SP) of processed marine food product SMEs.

Hypothesis 13:

Environmental turbulence (ET) significantly impacts sustainable performance (SP) through business model innovation (BMI) and sustainable performance (SP); thus, the serial mediation effects exist in the relationship between ET and SP.

3. Methodology

3.1. Target Population and Sample

The target population of the study were the entrepreneurs/managers of processed marine food product SMEs operating in the ten provinces of Thailand, namely: Samut Sakhon, Bangkok, Samut Prakan, Samut Songkhram, Rayong, Chachoengsao, Chonburi, Ranong, Chumphon, and Ratchaburi provinces. The ten provinces were chosen based on their geographical nature (coastal provinces in the Gulf of Thailand) and business concentration (fisheries, aquaculture farming, and food processing companies). The population frame was generated based on the directory list received from The Office of Small and Medium Enterprises Promotion (OSMEP), Thailand.

The list was carefully reviewed, and the 206 registered, processed marine food product SMEs currently under operation were identified and included in the study. Further, the sample frame was drawn from the list based on their accessibility. It was ensured that the SMEs have at least one of the communication modes such as phone number, email, Facebook, line application, and website, along with their postal address. Because there were severe inter-provincial COVID-19 travel restrictions during the study period, the researchers planned to collect data through online and digital methods. Thus, the final sample frame consisted of 186 SMEs which had at least any one of the communication modes. The entrepreneurs/managers of registered marine food processing companies were opted as inclusion criteria for selecting the sample. Hence, the exclusion criterion was “The entrepreneurs/managers working in large enterprises”. After obtaining ethical committee approval from the institution (June 2021), the survey form (in both Thai and English) was sent to all 186 processed marine food product (PMFP) SMEs and received 97 responses. The data were collected during the time period of July 2021–March 2022. However, six respondents indicated they were not interested in participating in the survey.

Further, the power and effect sizes were calculated to estimate the partial least squares (PLS) structural model [77] and determine the sample size to mitigate type-1 and type-2 error issues [78,79]. The study’s sample was determined using G*Power software (version 3.1.9.7) based on the minimum power and effect sizes recommended for social sciences [78,80] by using F-tests with medium effect size f² = 0.15, error probability α = 0.05, and power = 0.85, for the predictors included in the study. Thus, these estimations suggested that the total required minimum sample size should be 87. The final sample of this study consisted of 91 responses, which was sufficient. Further, the response rate was close to 50 percent, which is greater than the 30 percent acceptable level for online surveys [81]. Therefore, the sample size was sufficient for further testing the hypotheses. Thus, the collected data were subjected to validity, reliability, and hypotheses testing.

3.2. Measurement Scale Constructs

The measurement scale constructs of the study were framed based on the literature review. This study adopted, modified, and translated the scale constructs from English to the local language (Thai) and from the local language to English to verify the language efficiency, appropriateness, and comprehensibility. The ten items of business model innovation (BMI) constructs were adopted from various studies [16,31,35,38,39,66,82,83]. The items BMI1 “We offer a new combination of products and services through our business model”, BMI2 “We have expanded our products and services to new markets”, BMI3 “The business model brings new suppliers and channel partners”, BMI6 “We often introduce new operational processes, routines, and norms into our business model”, and BMI7 “We frequently review our business model and introduce new ideas and innovations to meet the current market demand” were adopted/slightly modified from prior studies [31,35,38]. The items BMI4 “The value propositions offered through our products/services now are not the same as offered two years ago”, BMI5 “We made new arrangements for information exchange throughout the supply chain in the past two years”, BMI8 “We regularly consider innovative opportunities for changing our existing pricing models” were adopted and slightly modified from the studies [16,39]. The two items BMI9, “Our production costs are constantly examined and if necessary improved in relation to market prices”, and BMI10, “ We emphasize innovative/modern actions to increase customer retention (e.g., CRM)”, were adopted from previous studies [66,82]. After careful examination of BMI theories from the literature, the following item was framed [31] and slightly modified, BMI11 “We always look forward to improving the flow of materials, products, services, information, and money throughout the supply chain”. The items BMI1, BMI2, BMI3, BMI4, BMI6, and BMI7 are novelty-centered, while the items BMI5, BMI8, BMI9, BMI10, and BMI11 are centered on novelty and efficiency which focus on cost reduction by avoiding the additional investment of time, efforts, and money. Altogether, the final scale constructs of BMI consisted of eleven items.

The sustainable competitive advantage (SCA) construct’s notion based on the VRIO framework was developed from the literature [40,41,44]. The items of SCA scale constructs were adopted from prior studies [22,46]; however, the items were modified and developed accordingly to suit the processed marine food product SMEs. The items included for SCA construct are SCA1 “We focus on transforming ideas into new services, processes, and procedures”, SCA2 “We offer superior products and services than other competitors”, SCA3 “We use the best available raw materials in our production process”, SCA4 “We maintain authentic flavor in our products without any artificial enhancers (coloring agents, artificial flavors, etc.)”, SCA5 “We use more efficient methods and techniques to maintain the shelf-life of our products which is difficult and costly to imitate”, SCA6 “The machineries we use is not available with the competitors with same quality and facilities”, SCA7 “Our manufacturing location adds more value to the product”, SCA8 “The resources used by us to maintain a long-shelf life will differentiate our products from competitors’ products”, SCA9 “Our access to superior limited resources may contribute to our competitive advantage in the marketplace”, and SCA10 “Our approaches for exploiting new resources may contribute to our competitive advantage in the marketplace”. Altogether, the final scale constructs of SCA consisted of ten items.

The environmental turbulence (ET) items were adopted from prior studies [59,66]. The ET scale consisted of five items, namely, ET1 “The actions of competitors in our major markets have been changing quite rapidly”, ET2 “Technological changes in our industry were rapid and unpredictable”, ET3 “The competitive market conditions were highly unpredictable”, ET4 “Customers’ product preferences changed quite rapidly”, and ET5 “Changes in customers’ needs were quite unpredictable”.

The sustainable performance (SP) scale construct was adopted from several studies [26,50,52,55,84] The economic performance (EP) items adopted from the prior studies [7,26,47,50,52,55,56,84,85,86] included eleven items, EP1 “Decrease in energy consumption cost”, EP2 “Decrease in material procurement cost”, EP3 “Decrease in waste treatment/waste disposal cost”, EP4 “Decrease in transportation cost”, EP5 “Decrease in total cashflow time”, EP6 “Efficiency in resources utilization”, EP7 “Employment opportunities for local people”, EP8 “Return on investment”, EP9 “Income stability”, EP10 “Sales growth”, and EP11 “Market share”. The social performance (SOP) included five items which were adopted from the previous studies [26,52,55,84,87]. SOP1 “Fair Labor Practices (no child labour use)”, SOP2 “Local community health and safety”, SOP3 “Improvements in the work environment (Employee’s health and Safety)”, SOP4 “Reduction in the impact of products, services, and activities on the local community”, and SOP5 “Positive brand image in the local community”. The environmental performance (ENP) consisted of six items from the previous studies [26,50,52,55,84], namely, ENP1 “Reduction of air emissions in our activities”, ENP2 “Improvements in solid waste management”, ENP3 “Reduction of wastewater disposal”, ENP4 “Environmentally friendly usage of utilities (e.g., energy and water)”, ENP5 “Improvements in maintaining the hygiene factors”, and ENP6 “Reduction in organizational and environmental accidents”. The operational performance (OP) included ten items, where six items were adopted from [50], i.e., OP1 “Increase in the amount of goods delivered on time”, OP2 “Decrease in inventory levels”, OP3 “Decrease in scrap rate”, OP4 “Promotes product quality”, OP5 “Increase in product line”, and OP6 “Improvements in capacity utilization”. The four items were referred to and modified to suit the study context [7,47,56,87], i.e., OP7 “Increase in customer satisfaction”, OP8 “Improvements in customer retention”, OP9 “Improvements in supplier and customer relationship”, and OP10 “Reduction in order processing time”. Altogether, the final scale constructs of SP consisted of 32 items.

The final latent scale constructs, i.e., business model innovation (BMI) consisted of eleven items, sustainable competitive advantage (SCA) consisted of ten items, environmental turbulence (ET) consisted of five items, and sustainable performance (SP) included four dimensions namely, economic performance (eleven items), social performance (five items), environmental performance (six items), and operational performance (ten items). The extent of respondents’ agreement or disagreement towards the scale constructs was measured using a five-point Likert scale ranging from 1—strongly disagree—to 5—strongly agree.

3.3. Content Validity and Pilot Study

After obtaining the ethical committee approval, the pilot study was conducted from the last week of June 2021 to mid of July 2021 with twelve respondents [88], and the results indicated the scale was precise, appropriate, and relevant to the field of study. The content validity was assessed by selecting three experts in the domain [81]. The three experts were asked to measure the theme, importance, appropriateness, and language clarity of the measurement scale in the field of study. The suggestions to include the term “supply chain” in the items BMI5 and BMI11 and the term “brand image” in the sustainable performance social dimension (SPS5) were incorporated.

4. Data Analysis and Interpretation

The data analysis was conducted using the seminr package [89] and R statistical programming language [90]. There are two rationales for adopting PLS-SEM in this research. The first rationale is the growing interest among researchers in using PLS-SEM in strategic management; it facilitates researchers in achieving new insights [91]. The second rationale is that the partial least square approach is widely used to confirm the theory and models to analyze the causal relationships. This is because the partial least square technique is appropriate under the conditions where the research questions are new or shifting, the theoretical model is not well established, the model with a substantial number of indicators, to test the epistemic relationships using formative and reflective measurements [92]. Therefore, this study explores the relationships between the latent constructs using four conceptual models under different causal conditions with PLS-SEM. Even though the models are new, this study adopted the reflective measurements because the notion of the individual constructs is established in the literature under various contexts. This study adopted a two-stage approach to test the hypotheses. Hence, in the first stage, the lower order model was tested for factor loadings, composite reliability, convergent validity, and discriminant validity and presented in the next section. In the second stage, all four structural models were tested for their reliability, validity, VIF, R², F² for the model suitability. Further, the bootstrapped paths (n = 1000) of the four conceptual structural models developed through the literature review were tested to find the causal relationship between the latent constructs under certain scenarios. The results of the structural model are also presented in this section.

Further, the measurement model was assessed using the outer loadings of lower constructs because sustainable performance is a higher-order construct. Subsequently, all the models were assessed for factor loadings, composite reliability, convergent validity, and discriminant validity. Finally, the conceptual models and hypotheses proposed in the study were tested using the bootstrapping approach with one thousand iterations (n = 1000). The results are presented in this section.

4.1. Measurement Scale Assessment (Stage-1)

The measurement scale was assessed for reliability (indicator and internal consistency) and validity (convergent and discriminant) [77].

4.1.1. Reliability

The initial step in the assessment process is measuring the reliability of the indicators and internal consistency. The measurement scale must measure the intended concept, which indicates the internal scale consistency [81].

Step-1

In the first step, the indicator reliability of the constructs is measured based on the value of the outer loadings of the constructs. The loadings of the indicators should be above the value of 0.708 [77]. However, in social sciences research, there may be possibilities for lower loadings, which is evident from studies in strategic management [91]. If the indicator loadings are less than 0.4, they should be removed from the study as they may affect the relationship of the proposed study constructs [77,91,93]. However, the items with low indicator loadings (BMI9, SCA4, SCA6, EP7, and OP5) between 0.40 and 0.708 were carefully examined and dropped from the study, which led to a substantial internal consistency and convergent validity [77]. The loading of the item SCA3 was low at 0.584, but the removal of the item did not impact the overall internal consistency of the scale; thus, the item was retained. Further, other items loaded above 0.6 were considered satisfactory with practical significance [94] and retained in the study. Majority of the outer loadings were high, or greater than 0.70 [77,93], indicating a high indicator reliability. The indicator loadings are presented in

Table 1. Reliability.

Constructs	Items	Indicator Loadings	Cronbach’s Alpha	Composite Reliability Rhoc	Rhoa	AVE
BMI	BMI1	0.641	0.892	0.912	0.898	0.509
	BMI2	0.708
	BMI3	0.682
	BMI4	0.779
	BMI5	0.705
	BMI6	0.661
	BMI7	0.745
	BMI8	0.679
	BMI10	0.773
	BMI11	0.746
SCA	SCA1	0.738	0.864	0.894	0.88	0.519
	SCA2	0.634
	SCA3	0.584
	SCA5	0.647
	SCA7	0.642
	SCA8	0.773
	SCA9	0.869
	SCA10	0.826
ET	ET1	0.809	0.861	0.897	0.904	0.635
	ET2	0.859
	ET3	0.822
	ET4	0.778
	ET5	0.709
EP	EP1	0.637	0.908	0.923	0.917	0.548
	EP2	0.739
	EP3	0.748
	EP4	0.755
	EP5	0.65
	EP6	0.714
	EP8	0.79
	EP9	0.811
	EP10	0.798
	EP11	0.739
SOP	SP1	0.74	0.906	0.93	0.929	0.729
	SP2	0.932
	SP3	0.916
	SP4	0.833
	SP5	0.835
ENP	ENP1	0.711	0.905	0.927	0.916	0.682
	ENP2	0.884
	ENP3	0.888
	ENP4	0.81
	ENP5	0.868
	ENP6	0.778
OP	OP1	0.661	0.903	0.921	0.913	0.565
	OP2	0.637
	OP3	0.728
	OP4	0.752
	OP6	0.728
	OP7	0.879
	OP8	0.812
	OP9	0.775
	OP10	0.764

Source: Authors’ own findings.

.

Step-2

The second step focuses on assessing the internal consistency of scale constructs. The values of Cronbach’s Alpha α, Composite reliability rho_C, and the exact reliability rho_A are presented in Table 1. The Cronbach’s alpha values ranged from 0.861 to 0.907, which is greater than 0.70 and less than 0.95, indicating a good internal consistency [77,95]. The results of the composite reliability rho_C values ranged from 0.89 to 0.93, which meets the threshold values of 0.70 to 0.90, indicating “good” reliability; however, the values above 0.90 are also acceptable as they are below 0.95. The true reliability of the constructs lies between Cronbach’s alpha and Composite reliability, where the former is conservative and the latter is too liberal. Thus, the exact reliability rho_A values ranged from 0.88 to 0.92, which lies between Cronbach’s Alpha and composite reliability [77].

Step-3

4.1.2. Convergent Validity

The convergent validity of the scale construct was assessed using the values of average variance extracted (AVE). The threshold AVE values should be greater than 0.50 [96]. With the AVE values for all the scale constructs greater than 0.50 (Refer to Table 1), the scale constructs met the threshold criteria, indicating that the scale constructs had explained 50% of the variance of its indicators [97].

Step-4

4.1.3. Discriminant Validity

The discriminant validity of the construct was analyzed using Henseler’s “heterotrait-monotrait ratio of correlations” (HTMT) [98]. Due to its efficacy and superior performance, HTMT has recently been highly preferred by researchers over Fornell and Larcker’s criterion [77,98]. According to Henseler et al., Fornell and Larcker’s criterion and cross-loadings examination are less sensitive in detecting the discriminant validity issues in PLS-SEM. Thus, the HTMT values of the constructs were also examined to assess the discriminant validity and found that the values (

Table 2. Discriminant validity.

HTMT
Constructs	BMI	SCA	ET	EP	SOP	ENP	OP
BMI	NA	NA	NA	NA	NA	NA	NA
SCA	0.775	NA	NA	NA	NA	NA	NA
ET	0.527	0.488	NA	NA	NA	NA	NA
EP	0.468	0.562	0.377	NA	NA	NA	NA
SOP	0.340	0.428	0.299	0.383	NA	NA	NA
ENP	0.285	0.419	0.379	0.593	0.663	NA	NA
OP	0.375	0.462	0.323	0.607	0.794	0.755	NA

Source: Authors’ own findings.

) are lower than the suggested threshold of 0.85 [98], indicating that the constructs are distinct and satisfy the requirements of the HTMT criterion.

4.2. Structural Models Assessment (Stage-2)

This research followed the steps of Hair et al. [77] to assess the reliability, validity, VIF, R2, and f2 of the proposed four models. The proposed hypotheses were tested under different theoretical concept models to analyze the impact of the independent, mediating, and moderating variables.

4.2.1. Concept Model A

The concept model A and hypotheses were framed based on the theory that proposed testing the direct effects of BMI on SP and the indirect effects of BMI on SP through the mediator SCA. Hypotheses 1–4 were tested in concept model A (Refer to

Table 3. Hypotheses for Concept Model A.

Hypotheses	Proposed Structural Paths
Hypothesis 1	BMI→ SP (positive impact)
Hypothesis 2	BMI→ SCA (positive impact)
Hypothesis 3	SCA→ SP (positive impact)
Hypothesis 4	BMI→ SCA→ SP (positive impact & SCA-mediator) Indirect effects

Source: Authors’ own compilation.

).

The model was tested for reliability, discriminant validity (HTMT), in-sample predictive power, collinearity and effect size before bootstrapping the model [77]. The Cronbach alpha values, rho_C, and rho_A for the variables in the model were above the threshold level of 0.7, and the AVE exceeded 0.5 (Figure 5). The constructs’ HTMT values were also lower than the suggested threshold of 0.85 [98]. The r² values for SCA = 0.489 and SP = 0.292 infer that the in-sample predictive power [77,99,100] of the model is moderate [101].

The model collinearity test results of the VIF values with SP: BMI = 1.958 and SP: SCA = 1.958 inferred no problematic collinearity issues in the concept model A. The f² values of BMI -> SCA = 0.958, BMI -> SP = 0.01, and SCA-> SP = 0.120 implies that the effect size of BMI -> SCA (f² > 0.025) and SCA-> SP (f² > 0.025) are large. Whereas the effect size of BMI -> SP (0.025 < f² > 0.01) is medium [102]. Further, the model was bootstrapped (n = 1000) to test the hypotheses, and the results are presented below.

The bootstrapped PLS-SEM results of Concept Model A were summarized in

Table 4. Concept Model A.

Model A
Paths	Original Estimates	Bootstrap Mean	Bootstrap SD	T Statistic	5% CI	95% CI	Hypothesis Support
H1: BMI -> SP	0.117	0.126	0.142	0.827	−0.118	0.352	Not Supported
H2: BMI -> SCA	0.699	0.708	0.063	11.073	0.597	0.802	Supported
H3: SCA -> SP	0.452	0.485	0.115	3.940	0.301	0.668	Supported
H4: BMI -> SCA -> SP	0.316	0.345	0.093	3.396	0.196	0.502	Supported

Source: Authors’ own findings.

and depicted in Figure 6. The results indicate that Hypothesis 1 is not supported due to its lower t-values (less than 1.96, p > 0.05) and the bootstrapped confidence intervals (−0.118 (5%) to 0.352 (95%)). Though the path estimates are positive with 0.117, BMI has no significant impact on SP. The path estimate of the impact of BMI on SCA is statistically significant with the coefficient value of 0.699, t-value of 11.073 > 1.96, and bootstrapped confidence intervals lie between the values 0.597 to 0.802. Thus, it indicates a significant and positive impact of BMI on SCA; hence, Hypothesis 2 is supported. The path estimates of SCA on SP are statistically significant with the coefficient value of 0.452, t-value of 3.940 > 1.96, and the bootstrapped confidence intervals lie between the values 0.301 to 0.668. Therefore, the positive impact of SCA on SP is noted, and Hypothesis 3 is supported. The path estimates for the paths BMI to SP through SCA is statistically significant with the coefficient value of 0.316, t-value of 3.396 > 1.96, and the bootstrapped confidence intervals lie between the values 0.196 to 0.502.

Hence, this indicates the mediating role of SCA in the relationship between BMI and SP; and also confirms the indirect effects of BMI on SP through SCA. Hypothesis 4 is supported. The results of concept model A confirm that all the hypothesized causal paths are significant except the direct path from BMI to SP.

4.2.2. Concept Model B

The Concept Model B proposes to test the direct effects of BMI on SP and the indirect effects of BMI on SP through the mediator SCA. Further, the factor ET was introduced as a moderating variable between BMI to SCA and BMI to SP. Hypotheses 1–4, 6 and 7 were tested in the Concept Model B (Refer to

Table 5. Hypotheses for Concept Model B.

Hypotheses	Proposed Structural Paths
Hypothesis 1	BMI→ SP (positive impact)
Hypothesis 2	BMI→ SCA (positive impact)
Hypothesis 3	SCA→ SP (positive impact)
Hypothesis 6	BMI*ET→ SP (positive impact & ET-moderator)
Hypothesis 7	BMI*ET→ SCA (positive impact & ET-moderator)

Source: Authors’ own compilation.

and

Table 6. Concept Model B.

Model B
Paths	Original Estimates	Bootstrap Mean	Bootstrap SD	T Statistic	5% CI	95% CI	Hypothesis Support
H1: BMI -> SP	0.159	0.145	0.13	1.219	−0.078	0.347	Not Supported
H2: BMI -> SCA	0.627	0.606	0.114	5.502	0.411	0.781	Supported
H3: SCA -> SP	0.246	0.359	0.134	1.844	0.132	0.579	Not Supported
H6: BMI*ET -> SP	0.238	0.189	0.138	1.73	−0.078	0.372	Not Supported
H7: BMI*ET -> SCA	0.221	0.142	0.171	1.293	−0.157	0.385	Not Supported

Source: Authors’ own findings.

).

Before proceeding with bootstrapping the model, the model was tested for reliability, in-sample predictive power, collinearity, and effect size. The Cronbach alpha values, rho_C, and rho_A for the variables in the model were above the threshold level of 0.7 (refer to Figure 7).

However, the AVE values for BMI, BMI*ET, and SP were found to be 0.480, 0.488, and 0.462. According to Fornell and Larcker, if the Cronbach alpha values were well above the threshold level of 0.7, the existence of convergent validity is proved even with an AVE less than 0.50 [96]. Thus, the results (Figure 7) prove that there is an existence of convergent validity. Further, the HTMT values of the constructs were also found to be lower than the suggested threshold of 0.85 [98]. So we proceeded further with the model analysis (Figure 8).The r² values for SCA = 0.585, and SP = 0.404 infer that the in-sample predictive power [77,99,100] of the model for SCA is substantial and SP is moderate [101]. The model collinearity test results of the VIF values with SCA: BMI = 1.317, SCA:ET = 1.324, SCA: BMI*ET = 1.007, SP: BMI = 2.265, SP:ET = 1.387, SP: BMI*ET = 1.188, SP: SCA = 2.410 inferred that there were no problematic issues of collinearity in the model. The f² values of BMI*ET -> SCA = 0.165, BMI*ET -> SP = 0.120, and SCA-> SP = 0.041 implies that the effect sizes of BMI*ET -> SCA (f² > 0.025), BMI*ET -> SP (f² > 0.025), and SCA-> SP (f² > 0.025) are large [102]. Further, the model was bootstrapped (n = 1000) to test the hypotheses, and the results are presented below.

The bootstrapped PLS-SEM results of Concept Model B are summarized in Table 6 and depicted in Figure 8. The results indicate that Hypotheses 1, 3, 6, and 7 are not supported due to their lower t-values (H1: 1.219, H3: 1.844, H6: 1.73, and H7: 1.293), which is less than 1.96, p > 0.05. However, the path coefficient (0.627) for the impact of BMI on SCA is positive and statistically significant with a t-value of 5.502 > 1.96; Hypothesis 2 was supported. Hence, the results of concept model B indicate the absence of moderation effects of ET in both the paths, BMI and SCA and BMI and SP. Therefore, the results are in line with the previous studies [31,62], and the moderating role of ET is not significant in this study. The results of Concept Model B confirm that the proposed paths are insignificant, except one path BMI to SCA is statistically significant.

4.2.3. Concept Model C

The Conceptual Model C proposes both direct and indirect effects of BMI and ET on SP through SCA. Hence, the model tests Hypotheses 1–5, 9, and 12 (

Table 7. Hypotheses for Concept Model C.

Hypotheses	Proposed Structural Paths
Hypothesis 1	BMI→ SP (positive impact)
Hypothesis 2	BMI→ SCA (positive impact)
Hypothesis 3	SCA→ SP (positive impact)
Hypothesis 4	BMI→ SCA→ SP (positive impact & SCA-mediator) Indirect effects
Hypothesis 9	ET→ SCA (positive impact)
Hypothesis 12	ET→ SP (positive impact)
Hypothesis 5	ET→ SCA→ SP (positive impact & SCA-mediator) Indirect effects

Source: Authors’ own compilation.

).

The model was tested for reliability, discriminant validity (HTMT), in-sample predictive power, collinearity, and effect size before proceeding with bootstrapping the model. The Cronbach alpha values, rho_C, and rho_A for the variables in the model were above the threshold level of 0.7 (Figure 9), and the AVE exceeded 0.5. The constructs’ HTMT values were also lower than the suggested threshold of 0.85 [98]. The r² values for SCA = 0.502 and SP = 0.316 infer that the in-sample predictive power [77,99,100] of the model for SCA is substantial and for SP is moderate [101]. The model collinearity test results of the VIF values with SCA: BMI = 1.319, SCA: ET = 1.319, SP: BMI = 2.124, SP: SCA = 2.007, and SP: ET = 1.355 inferred that there were no problematic issues of collinearity in the model. The f² values of BMI -> SCA = 0.598, BMI -> SP = 0, SCA-> SP = 0.112, ET -> SP = 0.035 implies that the effect size of BMI -> SCA (f² > 0.025), SCA-> SP (f² > 0.025), and ET -> SP (f² >0.025) are large. Whereas the effect size of BMI -> SP is null [102]. Further, the model was bootstrapped (n = 1000) to test the hypotheses and the results are presented below.

The bootstrapped PLS-SEM results of Concept Model C are summarized in

Table 8. Concept Model C.

Model C
Paths	Original Estimates	Bootstrap Mean	Bootstrap SD	T Statistic	5% CI	95% CI	Hypothesis Support
H1: BMI -> SP	0.034	0.031	0.156	0.216	−0.238	0.269	Not Supported
H2: BMI -> SCA	0.633	0.632	0.100	6.325	0.462	0.780	Supported
H3: SCA -> SP	0.409	0.438	0.130	3.154	0.227	0.654	Supported
H4: BMI -> SCA -> SP	0.259	0.279	0.102	2.538	0.129	0.456	Supported
H9: ET -> SCA	0.134	0.147	0.114	1.173	−0.039	0.340	Not Supported
H12: ET -> SP	0.211	0.217	0.116	1.827	0.013	0.400	Not Supported
H5: ET -> SCA -> SP	0.055	0.063	0.055	0.986	−0.016	0.164	Not Supported

Source: Authors’ own findings.

and depicted in Figure 10. The results indicate that Hypotheses 1, 5, 9, and 12 are not supported due to their lower t-values (H1: 0.216, H5: 0.986, H9: 1.173, and H12: 1.827), which is less than 1.96, p > 0.05. The path coefficients (0.633) for the impact of BMI on SCA are positive and statistically significant with a t-value of 6.325 > 1.96; thus, Hypothesis 2 was supported. The path estimate of the impact of SCA on SP is statistically significant with a coefficient value of 0.409 and a t-value of 3.154 > 1.96; therefore, Hypothesis 3 was supported. The path estimates for the paths BMI to SP through SCA are statistically significant with the coefficient value of 0.259 and a t-value of 2.538 > 1.96; thus, Hypothesis 4 was supported. Hence, this indicates the mediating role of SCA in the causal relationship between BMI and SP and also confirms the indirect effects of BMI on SP through SCA. The results of Concept Model C confirm that only the paths BMI to SCA, SCA to SP, and BMI to SP through SCA are statistically significant. However, the direct paths from BMI to SP, ET to SP, ET to SCA, and the indirect path of ET to SP through SCA are not statistically significant. Results of the path from ET to SCA do not support the prior study’s findings [20] and we note that the effects are insignificant.

4.2.4. Conceptual Model D

The Conceptual Model D proposes the serial mediation of ET on SP through BMI and SCA. Hence, the model tests Hypotheses 1–5, 8–13 (

Table 9. Hypotheses for Concept Model D.

Hypotheses	Proposed Structural Paths
Hypothesis 1	BMI→SP (positive impact)
Hypothesis 2	BMI→SCA (positive impact)
Hypothesis 3	SCA→SP (positive impact)
Hypothesis 4	BMI→SCA→SP (positive impact & SCA-mediator) Indirect effects
Hypothesis 8	ET→BMI (positive impact)
Hypothesis 9	ET→SCA (positive impact)
Hypothesis 10	ET→BMI→SCA (positive impact & BMI-mediator) Indirect effects
Hypothesis 11	ET→BMI→SP (positive impact & BMI-mediator) Indirect effects
Hypothesis 12	ET→SP (positive impact)
Hypothesis 5	ET→SCA→SP (positive impact & SCA-mediator) Indirect effects
Hypothesis 13	ET→BMI→SCA→SP (positive impact & BMI and SCA-mediator) Serial mediation/Indirect effects of ET on SP

Source: Authors’ own compilation.

and

Table 10. Concept Model D.

Model D
Paths	Original Estimate	Bootstrap Mean	Bootstrap SD	T Statistic	5% CI	95% CI	Hypothesis Support
H1: BMI -> SP	0.031	0.035	0.146	0.216	−0.220	0.257	Not Supported
H2: BMI -> SCA	0.633	0.636	0.100	6.298	0.458	0.786	Supported
H3: SCA -> SP	0.412	0.444	0.125	3.305	0.244	0.652	Supported
H4: BMI -> SCA -> SP	0.261	0.284	0.098	2.662	0.138	0.457	Supported
H8: ET -> BMI	0.496	0.506	0.090	5.503	0.355	0.651	Supported
H9: ET -> SCA	0.129	0.130	0.111	1.160	−0.043	0.318	Not Supported
H10: ET -> BMI -> SCA	0.314	0.322	0.078	4.035	0.198	0.457	Supported
H11: ET -> BMI -> SP	0.016	0.015	0.076	0.206	−0.117	0.125	Not Supported
H12: ET -> SP	0.209	0.210	0.117	1.780	0.024	0.404	Not Supported
H5: ET -> SCA -> SP	0.053	0.057	0.054	0.991	−0.02	0.152	Not Supported
H13: ET -> BMI -> SCA -> SP	0.129	0.145	0.061	2.111	0.061	0.256	Supported

Source: Authors’ own findings.

). Before bootstrapping the model, the model was tested for reliability, discriminant validity (HTMT), in-sample predictive power, collinearity, and effect size.

The Cronbach alpha values, rho_C, and rho_A for the variables in the model were above the threshold level of 0.7 (Figure 11), and the AVE exceeded 0.5. The constructs’ HTMT values were also lower than the suggested threshold of 0.85 [98]. The r² values for SCA = 0.498, SP = 0.315, and BMI = 0.246 infer that the in-sample predictive power [77,99,100] of the model for SCA and SP are moderate and for BMI is weak [101]. The model collinearity test results of the VIF values with SCA: BMI = 1.326, SCA: ET = 1.326, SP: BMI = 2.123, SP: SCA = 1.991, and SP: ET = 1.359 inferred that there were no problematic issues of collinearity in the model. The f² values of BMI -> SCA = 0.590, BMI -> SP = 0, SCA-> SP = 0.114, ET -> SP = 0.034 imply that the effect size of BMI -> SCA (f² > 0.025), SCA-> SP (f² > 0.025), and ET -> SP (f² > 0.025) are large. Whereas the effect size of BMI -> SP is null [102]. Further, the model was bootstrapped (n = 1000) to test the hypotheses and the results are presented below (Table 10 and Figure 12).

The bootstrapped PLS-SEM results of Concept Model D are summarized in Table 10 and depicted in Figure 12. The results indicate that Hypotheses 1, 5, 9, 11, and 12 (H1: BMI -> SP, H5 ET -> SCA -> SP, H9: ET -> SCA, H11: ET -> BMI -> SP, and H12: ET -> SP) are not supported due to its lower t-values (H1: 0.216, H5: 0.991, H9: 1.160, H11: 0.206, and H12: 1.780) which is less than 1.96, p > 0.05. There are no direct effects on the paths BMI to SP, ET to SCA, and ET to SP. Also, there are no indirect effects of the path ET to SP through SCA and ET to SP through BMI.

The path coefficients (0.633) for the impact of BMI on SCA are positive and statistically significant with a t-value of 6.298 > 1.96; hence Hypothesis 2 was supported. The path estimate of the impact of SCA on SP is statistically significant with a coefficient value of 0.412 and a t-value of 3.305 > 1.96, and Hypothesis 3 was supported. The path estimates for the paths BMI to SP through SCA is statistically significant with the coefficient value of 0.261 and t-value of 2.662 > 1.96; thus, Hypothesis 4 was supported, indicating the mediating role of SCA. The path coefficient (0.496) for the impact of ET on BMI is positive and statistically significant with a t-value of 5.503 > 1.96; therefore, Hypothesis 8 was supported. The impact of ET on SCA through BMI is positive with the path coefficient of 0.314 and significant with the t-value of 4.035 > 1.96; therefore, Hypothesis 10 was supported, indicating the mediating role of BMI and corroboration with the finding of a prior study [20]. Finally, the serial mediation effects were assessed for the paths (H13: ET -> BMI -> SCA -> SP) and to analyze the indirect effects of ET on SP by using multiple mediators (BMI and SCA). The path coefficient (0.129) indicates the positive impact of ET on SP through BMI and SCA and is statistically significant with the t-value of 2.111 > 1.96; therefore, Hypothesis 13 was supported. Hence, the Concept Model D answers research questions 5 and 6 and confirms the existence of serial mediation effects; ET’s effect on SP through BMI and SCA is positive and statistically significant.

5. Findings and Discussion

This research filled a knowledge gap by covering the unexplored aspects in the previous literature. The previous study by Foss and Saebi [29] argued that the BMI construct is not associated effectively with the antecedents, moderator, and outcome variables with empirical analysis. Subsequently, few studies [20,72,73,74] have identified BMI’s antecedents and outcome variables. However, there are limited studies [20] that have conceptualized and tested ET as an antecedent to BMI and SCA as an outcome variable. Nevertheless, none of the studies have studied the linkages of BMI with the outcome variables SCA and SP (with the economic, social, environmental, and operational dimensions) and undertaken empirically investigation within the context of SMEs engaged in marine food processing. Also, this research addressed another important issue, which is the verification of the strength of association of variables both as an independent variable and mediating variable (BMI), moderating variable and mediating variable (ET), and serial mediators (BMI, SCA). Thus, this research has explored the uncovered findings from previous studies. The major objectives and research questions have been answered by empirical investigation of four concept models. Since the SP is the higher order construct, initially all the lower constructs [103] utilized in this study were tested for reliability and validity, and found that the constructs met all the reliability and validity criteria. Further, all four of the models (Concept Model A, Concept Model B, Concept Model C, and Concept Model D) were tested for reliability and validity, and structural model assessments. All four of the models were tested with the proposed hypotheses under different conditions, to satisfy the objectives and answer the research questions. However, the model comparisons were not undertaken, because the models are distinct in defining the causal relationship between the latent constructs and the research objective was to test the hypotheses and study the causal relationship between the latent constructs to establish the theoretical foundations. The hypotheses support was analyzed based on the t-value at 1.96, and if the t-values for the hypothesized paths are greater than 1.96 then the proposed hypotheses were supported for the study.

The Concept Model A tested the hypotheses H1–H4 and found the significant positive impact of BMI on SCA (H2), SCA on SP (H3), and BMI on SP through SCA (H4). Thus, the three hypotheses (H2, H3 and H4) of this model were supported, while H1 for the direct effect from BMI to SP is not supported. H2 supported the findings of the prior study [20], and H1 does not support the findings of the prior study [19] due to the distinct items/dimensions adopted for the construct of sustainable performance (SP). However, H3 confirms the mediating role of SCA in the relationship between BMI and SP and the existence of complete mediation is evident. Thus, this answers the research questions (RQ) 1 and 2.

Concept Model B tested the hypotheses H1–H3, H6, and H7. This model introduced the factor environmental turbulence (ET) as a moderator and proposed the assumption that ET moderates the relationship between BMI and SP (H6) and BMI and SCA (H7). While the hypotheses BMI on SP (H1), BMI on SCA (H2), and SCA on SP (H3) were carried out from model A. The results indicate a significant positive impact of BMI on SCA (H2), and H2 was supported for the study, which is in line with the findings of Cheah et al. [20]. Whereas H1, H3, H6, and H7 were not supported for the study. However, H3 is supported in Model A but not supported in Concept Model B; this indicates the violation of hypothesis support when the new variables and new relationships are defined in the model. Also, the moderation effect of ET on the relationship between BMI and SP (H6) and BMI and SCA (H7) is not observed. Thus, the results indicate the absence of moderation effects by ET in the relationship between the latent constructs, corroborating the study of Zott and Amit [31]. Hence, this model answers the research questions (RQ) 3 and 4.

Concept Model C tested the hypotheses H1–H5, H9, and H12. Model C introduced ET as an independent latent construct and tested the effects of ET on SCA (H9), ET on SP (H12), and ET on SP through SCA (H5), along with the inclusion of BMI paths from Model A. The results indicate that the hypotheses H1 (BMI on SP), H12 (ET on SP), H9 (ET on SCA), and H5 (ET on SP through SCA) were not supported, whereas the findings for the paths BMI on SCA (H2), SCA on SP (H3), and BMI on SP through SCA (H4) are supported and similar to the results of Model A. The results of Model C indicate that the introduction of ET as an independent latent construct and its effect on the outcome variables is not statistically significant and evident. The findings of this model for the effect of ET on SCA (H9) corroborate with the prior study findings [21]. However, the newly conceptualized proposition for the effect of ET on SP does not support the study. Thus, concept model C reaffirms the research questions (RQ) 1 and 2.

The Concept Model D tested the hypotheses H1–H5, H8–H13. Model D linked the BMI and ET and measured the causal effects of ET on BMI (H8), ET on SCA through BMI (H10), ET on SP through BMI (H11), and ET on SP (H12). Further, the study also measured the serial mediation effects for the impact of ET on SP through the mediators BMI and SCA (H13). The other hypotheses, H1–H5, H9, and H12, were incorporated similar to the Concept Models A and C. The results indicate the effects of BMI on SP (H1), ET on SCA (H9), ET on SP (H12), ET on SP through SCA (H5), and ET on SP through BMI (H11) are not supported for the study. The results of H9 and H11 do not support the prior studies’ findings [20,62]. However, the effects of BMI on SCA (H2), SCA on SP (H3), BMI on SP through SCA (H4), ET on BMI (H8), ET on SCA through BMI (H10), and ET on SP through BMI and SCA were positive and statistically significant. The hypotheses H2, H8, and H10 support the previous study findings [20]; however, the complete mediation of BMI in the relationship between ET and SCA is evident in this study. The hypothesis (H3) of the effects of SCA on SP is similar to the prior study’s findings [22]. Finally, the serial mediation effects were confirmed by the hypothesis (H13), and the existence of the causal relationship between ET and SP through BMI and SCA is evident. Hence, the findings confirm that the mediators between the ET and dependent latent construct would balance the effects [62] of ET in the SP of SMEs. However, the inclusion of SCA as a mediator does not support the effects of ET on SP, but the inclusion of BMI as a mediator establishes the relationship between ET and SCA and further from ET to SP through multiple serial mediations (BMI and SCA). Hence, Concept Model D answers the research questions (RQ) 5 and 6.

Further, the mediating role of BMI and SCA is confirmed and well established from models A, C, and D, indicating a complete mediation. Also, these two constructs (BMI and SCA) performed well as serial mediators in establishing the relationship between independent and dependent constructs. It is prudent from the results that the construct BMI perform well as an independent variable and mediating variable; in addition, with the antecedent and mediating variables, the effect of BMI is positive and significant on the outcome variable. Therefore, the findings answered all the research questions proposed in this study.

The findings of this study are distinct in terms of study constructs selection for the study context and drawing the equation of relationships between the latent constructs by utilizing various concept models for the study with the inclusion of relevant factors and effects. The same is explained below.

The four Concept Models, A, B, C, and D, tested the proposed hypotheses for the study. Since every model is unique by introducing the variables with distinct characteristics (mediator, moderator, independent variable, and serial mediators) and relationships, this study tested only the proposed hypotheses and did not statistically compare the models.

Based on the hypotheses results, it is found that the Concept Models A and D supported the study and provided new theoretical insights. However, models B and C are not supported.

The Concept Model A supported the findings of Cheah et al. [20] for the path BMI to SCA, Haseeb et al. [22] for the path SCA on sustainable business performance and does not support the findings of Danarahmanto et al. [19] for the direct path BMI to SP. Therefore, it indicated that the business model innovation of SMEs leads to sustainable performance only with the presence of sustainable competitive advantage. The Concept Model A is unique as it incorporates four sustainability dimensions (economic, environmental, social, and operational), whereas the prior studies conceptualized in the framework of BMI and SCA did not emphasize more on social and environmental dimension variables in SP.

Further, Model B confirmed that ET does not moderate the causal relationship between BMI -> SP and BMI- > SCA. In addition, the findings of Model C confirmed that ET as an independent variable has no direct and indirect causal effects on SCA and SP. This is an important finding because the results of the Concept Model B support the findings of previous studies [31,62], indicating the absence of moderation effects of ET between BMI and SCA, and BMI and SP. The other important finding from the Concept Model C is that it does not support the findings of early studies [20] for the direct path ET on SCA without the mediator BMI. Thus, the results of the findings from Model B and Model C and previous literature [20,31,62] infer the need to test the relationship of ET and SP through a serial mediation of BMI and SCA.

Hence, Model D was proposed and tested through hypotheses. The results of Concept Model D supported the study hypotheses for serial mediation effects, which are distinct in measuring the relationship between ET and SP. This is another important finding from this study because early studies largely ignored the role of ET in SP. Even though there are studies that support the findings for the causal relationship between ET to BMI [20,71] and BMI to SCA [20], only a few studies [62,76] focused on measuring ET on firm performance. Hence, the findings of this study are unique in their use of serial mediation effects of BMI and SCA in the relationship between ET and SP.

Overall, it is evident from the results that SMEs cannot achieve sustainable competitive advantage without adopting business model innovation. It is obvious that large enterprises are more competitive than SMEs due to the availability of resources and competencies and their above average performance in terms of profitability. However, this study uses a broader approach of sustainability dimensions to assess sustainable performance and found that the BMI of SMEs supports the development of required skills, knowledge, resources, and competencies to sustain the competitive advantage in both domestic and global markets, particularly in the marine food processing sector. Furthermore, the SCA will lead to the attainment of sustainable performance in terms of economic, environmental, social, and operational aspects. In addition, it is also confirmed that in a turbulent environment, SMEs must focus on implementing BMI, which will lead to the attainment of SCA and SP.

The customers, external stakeholders, and technological advancements are the dynamic environmental factors contributing to the development of knowledge and capabilities of entrepreneurs and managers in implementing BMI. Therefore, this study, in line with [83,104,105,106] recent studies, expands the theory of BMI for SMEs and suggests the need for managerial/entrepreneurial developments to improve their innovation capabilities. This is because improving the innovation capabilities will enable them to respond to environmental turbulence such as market turbulence, technological turbulence and competition intensity and develop their BMI accordingly, which will further support SCA and SP.

6. Limitations and Directions for Future Research

This study is limited to the marine food processing SMEs and did not include SMEs in the entire food processing sector. However, the four concept models could be generalized to SMEs functioning in various sectors and can be further tested empirically. Future research could include SMEs in the food processing sector as a whole. In addition, a longitudinal study could provide the big picture of the SMEs and the level of innovation in their business model at the specified time intervals, efforts made by SMEs and the dimensions considered by them to attain SCA, level of knowledge and ability of SMEs to implement sustainability measures in their business operation. Also, a longitudinal study would facilitate the comprehensive examination of the turbulent environment and the SMEs’ preparedness and response to it. Further, such a study would strengthen the theory and support applications in other sectors. Besides this, a cross-country analysis would validate the models and results, and provide an in-depth understanding.

7. Implications

7.1. Theoretical Implications

This study empirically tested the constructs and models under different conditions and provided key knowledge and added value to the strategic management, entrepreneurship, and sustainability domains. Several studies have focused on the BMI of the service sector. However, only a few studies have focused on the manufacturing sector, especially the food processing sector. Thus, this study is a major contribution to the food processing sector’s ET, BMI, SCA, and SP theories. Further, this research extended the prior works [16,19,20,31,35,68] on BMI, and ET [62], by testing their causal effects on SCA and SP with the four theoretical models, which was not previously tested by other studies. Also, this research linked all three domains using an exploratory and comprehensive approach and tested the constructs BMI, SCA, SP, and ET empirically under different hypothesized causal relationships to provide an in-depth understanding of the constructs when they function as independent, moderating, and mediating variables. Therefore this study filled the gap suggested by the early work [29], and confirmed the existence of antecedents and mediators for BMI and further extended the knowledge by proving the serial mediation effects between ET and SP. Thus, the constructs and the models can be used by researchers in exploring the causal relationship between BMI, SCA, SP and ET in different regional and business contexts.

7.2. Practical Implications

The study results reveal the growing complex nature of the relationship between BMI, SCA, SP and ET. The results confirms and extend the prior work [62], hence the entrepreneurs must understand that ET supports in boosting the BMI and will further lead to attain SCA and SP. Further, SP is not limited to economic or operational performance. Hence, entrepreneurs and managers working in the marine food processing industry can adopt this scale to frequently review and innovate their business models to build capabilities for achieving sustainable competitive advantage and also to perform efficiently to meet the national goals aimed at sustainability. The higher order scale of SP developed through this study has a keen ‘eye for detail’ in assessing the SP of SMEs. Hence, the policymakers can adopt this scale to check the factors affecting the SP and expand their training and skill development programs accordingly to achieve sustainable SME performance in terms of UNSDGs.

8. Conclusions

Recently, the government has encouraged SMEs to adopt business model innovation and sustainability dimensions in their business. Thus, this study adopted an exploratory and comprehensive approach to testing the existing concepts and theory. The models were developed by duly considering the initiatives of “sustainability” in the context of Thailand’s marine food processing SMEs. This research developed and validated the scale/tested the concept models for measuring business model innovation (BMI), sustainable competitive advantage (SCA), sustainable performance (SP), and environmental turbulence (ET) from the prior literature under the present study context. Subsequently, the hypothesized causal relationships of the constructs BMI, SCA, SP and ET were proposed under four models.

Further, the constructs were measured in a conditional method using the same latent constructs as a moderator and independent variable; and an independent variable and mediator (BMI). The study confirmed the causal relationship between BMI and SCA and BMI’s indirect effects on SP. Also, the moderation effects of ET are absent in the relationship between the latent constructs. However, the ET had a significant causal impact on BMI and indirect effects on SCA through BMI. After testing the constructs’ causal relationships under different conditions, the complete mediating roles of BMI and SCA and the serial mediation effects of BMI and SCA in the relationship between ET and SP were confirmed. Thus, it is evident that the effects of ET are balanced when the SMEs incorporate BMI, which further leads to achieving SCA and SP.

Consequently, helping SMEs to support the national sustainability goals and stay competitive and sustained in the long run. The results of this study would support SMEs and policymakers in emphasizing the development of SMEs’ skills and capabilities to innovate their business model and achieve sustainable competitive advantage and sustainable performance to expand their business globally and meet societal needs. Further, the SMEs would also learn to respond during the high turbulence in the industry.