Determinants of Superior Long-Term Business Performance in Thai Small and Medium-Sized Enterprises: An Integrated Analysis Using Fuzzy Rough Set Theory and Second Order Confirmatory Factor Analysis

Abstract

This study investigates the determinants of superior long-term business performance in Thai digital entrepreneurship through an innovative mixed-method approach combining Rough Set Fuzzy Theory and Second-order Confirmatory Factor Analysis. This research addresses a significant gap in the existing literature by incorporating business strategies, product innovation, social media adoption, and entrepreneurial orientation into a comprehensive framework, extending beyond traditional Technology–Organization–Environment (TOE) models. This study analyzes seven key factors that influence digital business success: technology, organization, external environment, social media adoption, business strategy, product innovation, and entrepreneurial orientation. The methodological approach employed for this study utilized expert consensus validation and model verification techniques to develop a novel integrated model specifically tailored for Thailand’s digital SME context. The findings reveal that business strategy and entrepreneurial orientation are primary drivers of business success. This research provides valuable insights for practitioners in the Thai digital entrepreneurship ecosystem, offering a structured approach to achieving sustainable long-term business success.

1. Introduction

Small and medium-sized enterprises (SMEs) face critical survival challenges in the global economy, with 67% of SME executives worldwide citing survival and expansion as their primary concern [1,2]. Despite generating approximately 70% of global jobs and GDP, they struggle with fundamental issues, including low profit margins, scaling difficulties, and market expansion challenges [3]. These challenges are particularly acute in Thailand, where SMEs have rapidly embraced digital transformation to meet evolving customer demands [4]. The digital transformation landscape presents both opportunities and challenges for Thai SMEs. While Thai enterprises have shown remarkable adoption rates of technology, with 71% accelerating investments in customer service technologies compared to 41% globally, many still face difficulties in maintaining sustainable long-term performance, defined as sustained growth in revenue, market share, and stakeholder value over a three-year period [5]. This challenge is particularly notable when compared to neighboring countries like Singapore and Malaysia, where digital transformation success rates are 15% and 10% higher, respectively [6]. In Thailand’s context, where 90% of SMEs have embraced digitalization under the Thailand 4.0 policy initiative, organizations must leverage digital technologies for both operational efficiency and environmental stewardship [7]. Research conducted between 2022 and 2024 indicates that Thai SMEs implementing sustainable digital practices have experienced a 35% increase in operational efficiency and a 28% improvement in stakeholder satisfaction [8]. Recent studies demonstrate that successful digital transformation, when aligned with sustainability initiatives, creates significant competitive advantages through improved operational efficiency and enhanced ESG performance, leading to an average 25% increase in market share and stronger stakeholder trust [9]. The integration of technology, organizational structure, and external environment factors has become crucial for business survival and growth in the digital age [10]. Previous research has established the Technology–Organization–Environment (TOE) framework as a foundational theoretical model, which, when integrated with social media adoption, significantly influences business performance through both financial metrics and non-financial outcomes [11,12]. Studies specific to Thai SMEs reveal that successful digital transformation requires more than technological adoption alone, with integrated approaches combining technology and sustainability initiatives achieving 40% higher performance outcomes than single-factor approaches [13]. Thai SMEs that exhibit a strong entrepreneurial orientation demonstrate superior performance in digital markets, with evidence showing a 30% higher survival rate during market disruptions compared to regional counterparts [14]. The current literature reveals several critical gaps in understanding how Thai SMEs maintain long-term performance post-digital adoption, integrate sustainability practices, and adapt frameworks to the Thai business context. The interplay between business strategy, product innovation, and entrepreneurial orientation has become increasingly critical for Thai SMEs’ sustainable growth, supported by government initiatives such as the Digital SME Development Program and SME Go Digital measures [15,16]. This research addresses these gaps by investigating the following question: “What are the key determinants influencing the superior long-term business performance of digital SMEs in Thailand”? Through a mixed-method approach combining quantitative analysis of performance metrics and qualitative assessment of successful cases, this study aims to develop and validate a comprehensive framework integrating TOE factors with social media adoption, business strategy, product innovation, and entrepreneurial orientation for Thai digital SMEs, and to identify critical success factors that drive superior long-term business performance in Thailand’s digital business ecosystem. This research contributes to both theoretical understanding and practical applications, ultimately supporting the development of more resilient and successful digital businesses in Thailand’s evolving business landscape, where integrated approaches have been shown to deliver higher long-term performance outcomes compared to traditional approaches. The findings will benefit multiple stakeholders, including SME owners, policy makers, and business development agencies, ultimately contributing to the sustainable growth of Thailand’s digital economy.

2. Theoretical Framework

This research investigates the determinants of business performance in Thai digital SMEs by examining seven critical elements through an innovative Fuzzy Rough Set analysis approach. The study integrates the traditional Technology–Organization–Environment (TOE) framework with four additional factors: business strategy, product innovation, entrepreneurial orientation, and social media adoption. Through seven hypotheses, we propose that each element positively influences superior long-term business performance, as illustrated in our conceptual framework (Figure 1). This comprehensive model combines Zadeh’s Fuzzy Logic and Pawlak’s Rough Set Theory to provide a sophisticated understanding of the factors driving success in Thai digital SMEs, detailed as follows.

2.1. TOE

The technology–organization–environment (TOE) framework, introduced by Tornatzky and Fleischer in 1990 [17], explains organizational technology adoption through three key contexts: technological (internal tech capabilities and readiness), organizational (firm’s characteristics and resources), and environmental (external pressures, customer pressures, and external pressure). This comprehensive framework has proven to be valuable across various technological innovations, and it aligns well with social media adoption theories [18].

2.1.1. Technology Context (TE)

The technological context within the TOE framework encompasses both internal technological capabilities and infrastructure readiness that organizations must possess for successful technology adoption. This context specifically emphasizes the importance of skilled personnel and modern equipment as critical components for implementation. Organizations require both technical expertise in their workforce and up-to-date technological infrastructure to effectively adopt and implement new technologies [19]. Technological readiness includes the availability of skilled employees who can understand and utilize new technologies and appropriate technological tools and equipment that enable successful adoption processes [20]. Research has shown that organizations with strong internal technical expertise and modern infrastructure demonstrate higher success rates in technology adoption and implementation, particularly when they maintain current practices and equipment internal to the firm while remaining aware of available, external technologies [21].

2.1.2. Organization Context (OR)

The organizational context within the TOE framework emphasizes several critical internal factors that influence technology adoption. Organizations that maintain dedicated technology departments and demonstrate a clear technological vision show higher success rates in digital transformation initiatives [22]. These organizations typically emphasize continuous technological development through regular employee training programs and effective human resource planning [23]. The organizational culture plays a pivotal role, particularly when it promotes consistent technology usage and maintains open communication channels between management and employees. This cultural aspect, combined with strong leadership support and clear organizational vision, creates an environment conducive to successful technology implementation [24]. Furthermore, organizations that establish effective communication channels between employees and management, while maintaining a technology-oriented culture, demonstrate better adaptation to technological changes and higher levels of innovation adoption [25].

2.1.3. External Environment Context (EX)

The environmental context within the TOE framework emphasizes the competitive landscape and customer dynamics that organizations face in their market operations. In today’s rapidly evolving business environment, organizations experience increasing market competition and shifting customer demands. Companies face growing numbers of competitors who employ diverse strategies to attract customers through varied product and service offerings [26]. The competitive pressure intensifies as customers gain more purchasing options and leverage in negotiations, leading to higher bargaining power [27]. This shift in market dynamics requires organizations to continuously adapt their strategies and offerings to meet evolving customer preferences and demands [28]. Customer bargaining power has become particularly significant as buyers have access to more information and alternatives, enabling them to demand better value propositions and competitive pricing [29].

2.2. Business Strategy (BU)

Business strategy in modern organizations encompasses dual approaches of cost leadership and differentiation to maintain competitive advantage. Companies implement strategic pricing policies that enable them to offer products at lower prices than competitors while optimizing resource utilization for maximum efficiency and value creation [30]. Organizations achieve cost advantages through systematic reduction in operational expenses and efficient resource management, allowing them to maintain lower costs than competitors. Simultaneously, firms focus on building distinctive brand images and creating differentiated product offerings that stand out in the marketplace [31]. Companies enhance their competitive position by developing unique and innovative products or services, expanding their product portfolio beyond traditional offerings, and ensuring faster time-to-market compared to competitors [32].

2.3. Product Innovation (IN)

Product innovation involves a comprehensive approach to continuous development and improvement to maintain competitive advantage in the market. Organizations consistently develop new products and services to meet evolving customer needs while focusing on quality improvements that exceed competitor offerings [33]. Companies engage in ongoing research and design processes to bring innovative products to market, emphasizing both incremental improvements and radical innovations [34]. The innovation process includes continuous service process development and the adoption of modern equipment for production and service delivery [35]. This dual approach of market-driven development and technological advancement ensures that organizations can effectively respond to customer demands while maintaining operational excellence.

2.4. Entrepreneur Orientation (EN)

The entrepreneurial mindset represents a complex interplay of proactive innovation, strategic vision, and calculated risk-taking capabilities that drive business success. Entrepreneurs demonstrate leadership by identifying and pursuing new market opportunities before competitors, consistently seeking innovative approaches to establish market dominance [36]. They combine strong business expertise with ethical governance principles, ensuring high-quality service delivery while maintaining robust risk management strategies [37]. These leaders exhibit a balanced approach between pioneering new initiatives and strategic planning, often being first-to-market with innovative services despite uncertainties [38]. Their success stems from deep industry knowledge, continuous opportunity recognition, and an unwavering commitment to achieving market leadership through ethical business practices and quality-focused operations.

2.5. Social Media Adoption (SO)

Social media adoption has become a fundamental strategic imperative for modern business operations, transforming how organizations connect with and serve their customers [39]. Organizations leverage social platforms for rapid product promotion and marketing channel expansion, enabling them to reach broader audiences more effectively [40]. These digital platforms serve as powerful tools for enhancing customer engagement, facilitating improved customer relationships, and driving repeat purchase behavior. Through social media, businesses can develop personalized interactions, create meaningful customer experiences, and maintain ongoing relationships that foster customer loyalty [41]. The strategic use of various social platforms allows companies to establish a stronger market presence while simultaneously building and maintaining valuable customer relationships through consistent digital engagement.

2.6. Fuzzy Rough Set Theory

The integration of Fuzzy Set Theory and Rough Set Theory represents a powerful advancement in data analysis methodologies [42]. While Pawlak’s Rough Set Theory (1981) excels in handling imperfect data through set approximations and requires no preliminary information [43], Zadeh’s Fuzzy Logic (1965) manages imprecise information using partial truth values between 0 and 1 [44]. Their combination creates a more robust framework for dealing with real-world data complexity [45]. The hybrid Fuzzy–Rough Set approach leverages complementary strengths—Rough Set’s ability to work with inconsistent data and Fuzzy Logic’s capability to handle gradual transitions [46]. This synergistic combination enhances uncertainty handling, improves pattern recognition, and provides more sophisticated data analysis capabilities, particularly valuable in artificial intelligence, machine learning, and decision support systems where traditional single-theory approaches may fall short [47].

3. Research Methodology

This research employed a comprehensive mixed-methods approach to develop a new framework for superior long-term business performance of Thai digital SMEs. The research methodology framework, illustrated in Figure 2, is elaborated in detail below, The methodology proceeded in two distinct phases. In the first phase, this research utilized qualitative methods, beginning with an extensive literature review to identify potential factors, followed by the application of Rough Set Theory [48] and the Fuzzy Logic technique [49,50] to validate and refine these factors through expert opinions. This qualitative phase helped establish the initial theoretical framework, validate the constructs through expert consensus, and validate the questionnaire. This research was granted approval by the Ethics Review Board of Rangsit University (COA. NO. RSUERB2024-083). The second phase employed quantitative methods, starting with Exploratory Factor Analysis (EFA) [51] to identify the underlying factor structure of the constructs. Subsequently, Confirmatory Factor Analysis (CFA) [52] was conducted to validate each individual construct’s measurement model. The analysis then progressed to first-order CFA to examine the relationships between observed variables and their respective latent constructs, followed by second-order CFA to validate the hierarchical structure of the theoretical model. This systematic approach ensured both theoretical soundness through expert validation and statistical rigor through empirical testing, resulting in a comprehensive framework for understanding Thai digital SMEs’ long-term performance.

3.1. Qualitative Research

This research employed a qualitative methodological framework integrating the Rough Set and Fuzzy Logic techniques to examine expert consensus on factors driving superior long-term business performance. Following a comprehensive literature review, this study engaged 21 carefully selected experts representing three distinct domains of digital entrepreneurship expertise, government officials and university lecturers specializing in entrepreneurship development. Data collection proceeded through a structured electronic Delphi technique comprising three sequential rounds: initial open-ended inquiries, systematic variable extraction, and closed-ended confirmation questionnaires. The qualitative data gathered from this expert panel underwent rigorous analysis using an integrated Rough Set and Fuzzy Logic approach, enabling the synthesis and elucidation of consensus findings regarding critical success factors in digital entrepreneurship.

3.1.1. Population and Sampling

This study encompasses pertinent demographic groups and influences data analysis across multiple stakeholders. Following Macmillan’s (1971) theory [53], which states that a minimum of 17 experts in qualitative research yields a minimal error rate reduction, this study employed 21 experts to achieve an error rate of less than 0.02. The researcher stratified the participants into three distinct cohorts: Group 1 comprised 7 representatives from governmental institutions, Group 2 consisted of 7 digital entrepreneurs, and Group 3 included 7 university faculty members specializing in entrepreneurship development. The total sample size of 21 participants ensures statistical reliability while maintaining balanced representation across stakeholder groups.

3.1.2. Research Instruments

The researcher developed a comprehensive data collection instrument, an online questionnaire system that proceeded in two distinct phases. The first phase employed open-ended questions to identify and extract key thematic elements from expert insights. The second phase utilized a structured questionnaire with a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree) [54] to enable experts to systematically evaluate and rate the identified factors the Step of the quantitative research present in Figure 3. This dual-phase approach facilitated both qualitative exploration and quantitative assessment of expert opinions while ensuring methodological rigor in the data collection process.

3.1.3. Data Collection

The data collection process utilized a systematic three-round electronic Delphi survey administered via email over a three-month timeframe spanning from June to August 2024. The structured iterative protocol employed sequential questionnaires, with each subsequent round building upon and refining the insights gathered from previous responses. This methodological approach ensured comprehensive expert input while maintaining analytical rigor throughout the survey administration, facilitating the progressive development and validation of expert consensus through a carefully designed framework that maximized the depth and quality of expert contributions.

3.1.4. Data Analysis

This research implemented a hybrid analytical methodology integrating Rough Set Theory and Fuzzy Logic techniques to evaluate expert consensus data. This innovative approach was designed to enhance the precision of expert opinion analysis while addressing the inherent uncertainty in decision-making processes. A 4-step analysis, as detailed below, was conducted to evaluate the expert consensus data. The integration of these complementary techniques enhanced the precision and reliability of expert opinion analysis.

• Step 1. The analysis began with Rough Set Theory [55] the rough set theory conceptual farmwork shown in Figure 4 by implementation, utilizing boundary region concepts through the following:
• Lower Approximation = {x ∈ U: [x]R ⊆ X};
• Upper Approximation = {x ∈ U: [x]R ∩ X ≠ ∅}.

The decision-making process follows two distinct rules based on suitability and practicality attributes: full agreement (d = 1), when both attributes score 5–7, considered lower approximation, and complete disagreement (d = 0), when both score 1–4, considered upper approximation, to systematically process expert opinions with the rough set condition rule shown in

Table 1. Rough Set condition rules (Source: Author).

Decision Type	Condition
Full Agreement (d = 1)	Both suitability and practicality ratings: 5–7
Complete Disagreement (d = 0)	Both attribute ratings: 1–4

. This mathematical framework provided a foundation for handling uncertainty in expert responses.

• Step 2. Expert evaluations, collected through a 7-point Likert scale, were classified using binary decision rules.

• Step 3. The secondary phase employed the Fuzzy Logic technique [56], which extends the traditional Delphi by incorporating Fuzzy Set Theory to handle the fuzziness in expert judgments by using a triangle of membership function, as shown and calculated below in Figure 5.

F = (L, M, U) L = the smallest numerical value of a member of Fuzzy; M = the largest numerical value of a member of Fuzzy; and U = Fuzzy’s greatest membership fee.

The process transformed the Rough Set binary outputs as Figure 6 into triangular Fuzzy numbers (TFN) using the following parameters in

Table 2. Triangular Fuzzy number (TFN) parameters (Source: Author).

Rough Result Output	Lower (L)	Middle (M)	Upper (U)
Agree = 1	0.50	0.75	1.00
Disagree = 0	0.00	0.25	0.50

.

• Step 4. The final consensus evaluation employed a defuzzification process using the following arithmetic mean formula: Average = (L + M + U)/3, where L, M, and U represent the lower and upper bounds of the triangular fuzzy numbers, respectively. The threshold value was established at 0.65, with variables accepted when exceeding this threshold and rejected when falling below.

3.2. Quantitative Research

This research employed a quantitative methodology to develop and validate a comprehensive model of digital, superior long-term business performance, building upon the qualitative findings from expert consensus. This study transformed the previously identified 7 factors and 42 variables into a structured online questionnaire, which was administered to 852 digital entrepreneurs. Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) were implemented to validate the construct validity, utilizing both first-order and second-order analyses to assess model fit indices. This rigorous analytical approach enabled the systematic evaluation and refinement of the theoretical framework, ensuring robust validation of the proposed model’s structural relationships and dimensional integrity within the context of digital entrepreneurship performance.

3.2.1. Population and Sampling

The study targeted Thailand’s SME population of approximately 3,225,743 enterprises (as of July 2024), focusing on online business operators. Following established statistical guidelines requiring 10–20 participants per observed variable and a minimum threshold of 200 participants for model validity, this study’s 42 observable variables necessitated a minimum sample size of 420 participants. The final dataset comprised 852 valid responses, exceeding the minimum requirement and ensuring robust statistical analysis for the causal relationship model examination (See

Table 3. Descriptive statistics of categorical and continuous demographic variables.

Variable	Frequency	Percentage (%)
Gender
Male	443	52
Female	409	48
Education Level
Below Bachelor’s Degree	46	5.4
Bachelor’s Degree	622	73
Master’s Degree	166	19.5
Doctoral Degree	18	2.1
Company Revenue
<50 million THB	376	44.1
51–100 million THB	306	35.9
100–200 million THB	139	16.3
>200 million THB	31	3.6
No. of Employees
<5 people	250	29.3
6–50 people	337	39.6
51–100 people	203	23.8
101–200 people	62	7.3

).

3.2.2. Research Instruments

The research instrument consisted of a structured online questionnaire developed from expert consensus findings, incorporating 7 validated factors with 42 variables. The measurement system employed a 7-point Likert scale where respondents rated their agreement level from 1 (Strongly Disagree) to 7 (Strongly Agree). This comprehensive measurement tool enabled systematic quantitative data collection while maintaining alignment with the established theoretical framework.

3.2.3. Data Collection

Data collection was executed through a comprehensive online questionnaire distributed across multiple channels, including IT events, social media platforms, and direct email outreach, during a three-month period from August to October 2024. This study employed a single screening criterion to verify participants’ status as digital business entrepreneurs of SMEs in Thailand, resulting in 852 valid and complete questionnaires that were retained for analysis after rigorous data validation. This sampling approach ensured a representative cross-section of Thailand’s digital entrepreneurship ecosystem, providing a robust foundation for subsequent statistical analysis.

3.2.4. Data Analysis

The data analysis process followed a comprehensive four-step methodology, beginning with Exploratory Factor Analysis (EFA) [57] that employed the Kaiser–Meyer–Olkin (KMO) measure requiring values exceeding 0.6 [58], followed by Confirmatory Factor Analysis (CFA) [59] to evaluate the measurement model fit. The CFA phase involved a systematic examination of individual constructs, progressing through first- and second-order factor analyses, while utilizing multiple fit indices for validation. These indices included CMIN/DF (≤3.0) [60], along with AGFI, GFI, NFI [61], CFI and TLI [62] (all requiring values greater than or equal to 0.9), RMSEA, and RMR (both needing to be less than or equal to 0.08) [63], as well as the statistical value factor loading (≥0.70) [64], CR (≥0.70) [65], AVE (≥0.50) [65], CA (≥0.70) [66], and R² (≥0.20) [67] as the established acceptance criteria.

4. Results

This study analyzed qualitative data using Rough Set and Fuzzy Logic Theory to examine expert-provided information. The quantitative analysis employed Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to evaluate model fit. Subsequently, first-order and second-order factor analyses were conducted to assess the overall model’s goodness of fit. The results are presented below.

4.1. Qualitative Result

The validation process employed expert evaluation using CRIP (Composite Reliability Index for Panels) analysis, with a threshold value of 0.65 determining acceptability detail of Expert Consensus result shown in

Table 4. Expert consensus results (Source: Author).

No.	Question	Average			Crisp	Result
Technology Context (TE)		L	M	U
TE1	Your company always promotes the use of modern technology in operations.	0.50	0.75	1.00	0.750	Acceptable
TE2	Your company provides modern computer equipment for use within the organization.	0.48	0.73	0.98	0.726	Acceptable
TE3	Your company has employees who consistently possess skills in using new technologies.	0.48	0.73	0.98	0.726	Acceptable
TE4	Your company uses technology in managing and planning organizational business resources.	0.48	0.73	0.98	0.726	Acceptable
TE5	Your company has systematic data storage and data management.	0.48	0.73	0.98	0.726	Acceptable
TE6	Your company uses data analytics to aid in executive decision-making.	0.48	0.73	0.98	0.726	Acceptable
Organization Context (OR)
OR1	Your company has departments responsible for business technology operations.	0.50	0.75	1.00	0.750	Acceptable
OR2	Your company has a vision for continuously implementing technology in business operations.	0.50	0.75	1.00	0.750	Acceptable
OR3	Your company encourages employees to continuously develop knowledge in modern technology.	0.50	0.75	1.00	0.750	Acceptable
OR4	Your company has human resource planning to achieve maximum operational efficiency.	0.48	0.73	0.98	0.726	Acceptable
OR5	Your company has an organizational culture of continuous technology use.	0.48	0.73	0.98	0.726	Acceptable
OR6	Your company has regular communication channels between employees and management.	0.48	0.73	0.98	0.726	Acceptable
External Enviroment Context (EX)
EX1	Your company has increasing market competition.	0.43	0.68	0.93	0.679	Acceptable
EX2	Your company has competitors with effective customer attraction strategies.	0.43	0.68	0.93	0.679	Acceptable
EX3	Your company has competitors with diverse products or services.	0.43	0.68	0.93	0.679	Acceptable
EX4	Your customers have more options in purchasing products or services.	0.43	0.68	0.93	0.679	Acceptable
EX5	Your customers’ needs for products or services change over time.	0.43	0.68	0.93	0.679	Acceptable
EX6	Your customers have relatively high bargaining power.	0.43	0.68	0.93	0.679	Acceptable
Business Strategy (BU)
BU1	Your company has a pricing policy that is lower than competitors.	0.43	0.68	0.93	0.679	Acceptable
BU2	Your company plans business resources for maximum benefit and value.	0.48	0.73	0.98	0.726	Acceptable
BU3	Your company can reduce certain business costs to be lower than competitors.	0.45	0.70	0.95	0.702	Acceptable
BU4	Your company creates a memorable image that differs from competitors.	0.50	0.75	1.00	0.750	Acceptable
BU5	Your company creates product or service differentiation to be more interesting.	0.48	0.73	0.98	0.726	Acceptable
BU6	Your company increases distribution of different products or services.	0.48	0.73	0.98	0.726	Acceptable
BU7	Your company presents products or services faster than competitors.	0.45	0.70	0.95	0.702	Acceptable
Entrepreneur Orientation (EN)
EN1	You initiate new things in your business before competitors in the same industry.	0.48	0.73	0.98	0.726	Acceptable
EN2	You have vision and regularly seek new business opportunities.	0.50	0.75	1.00	0.750	Acceptable
EN3	You make great efforts to make your business a leader above competitors.	0.48	0.73	0.98	0.726	Acceptable
EN4	You have plans or risk management when presenting new concepts.	0.48	0.73	0.98	0.726	Acceptable
EN5	You have good governance in business focusing on quality products or services.	0.48	0.73	0.98	0.726	Acceptable
EN6	You often introduce new services in your business first to market despite uncertainties.	0.45	0.70	0.95	0.702	Acceptable
EN7	You have excellent knowledge and expertise in your business.	0.50	0.75	1.00	0.750	Acceptable
Product Innovation (IN)
IN1	Your company continuously develops new products or services to meet customer needs.	0.48	0.73	0.98	0.726	Acceptable
IN2	Your company consistently improves products or services to be superior to competitors.	0.48	0.73	0.98	0.726	Acceptable
IN3	Your company regularly conducts research and designs products or services for the market.	0.48	0.73	0.98	0.726	Acceptable
IN4	Your company continuously develops new service processes.	0.48	0.73	0.98	0.726	Acceptable
IN5	Your company regularly provides modern equipment for production and service processes.	0.45	0.70	0.95	0.702	Acceptable
Social media adoption (SO)
SO1	Your company uses social media for rapid product or service promotion.	0.50	0.75	1.00	0.750	Acceptable
SO2	Your company uses social media to increase marketing channels.	0.50	0.75	1.00	0.750	Acceptable
SO3	Your company uses social media to increase access to new customers.	0.50	0.75	1.00	0.750	Acceptable
SO4	Your company uses social media to effectively increase product or service repurchases.	0.50	0.75	1.00	0.750	Acceptable
SO5	Your company uses social media to effectively develop customer relationships.	0.50	0.75	1.00	0.750	Acceptable

. All seven key dimensions received “Acceptable” ratings, demonstrating robust validation results. The technology context showed strong performance (0.726–0.750), particularly in modern technology adoption. The organization context exhibited similar strength (0.726–0.750), excelling in technology operations, vision implementation, and employee development. The external environment context maintained consistent levels at 0.679, while business strategy varied positively (0.679–0.750), notably in brand differentiation. Entrepreneur orientation demonstrated solid results (0.702–0.750), especially in vision and expertise. Product innovation showed steady scores (0.702–0.726), and social media adoption emerged as particularly strong with uniform scores of 0.750. All dimensions surpassed the minimum threshold of 0.65, thereby confirming the reliability of the research findings.

4.2. Quantitative Results

In addition to the quantitative analysis, the researchers conducted both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to evaluate the model’s fit. The analysis incorporated both first-order and second-order factor structures. The detailed findings are presented below.

4.2.1. Exploratory Factor Analysis (EFA) Results

The Exploratory Factor Analysis (EFA) was applied to examine internal reliability and discover the factor structure, ultimately helping the researchers create a more robust measurement instrument and establish stronger evidence for construct validity before conducting the CFA. The analysis utilized the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy, where the recommended threshold exceeds 0.60. The results yielded a KMO value of 0.985, meeting the measurement criteria. Additionally, Bartlett’s Test of Sphericity was significant (χ² = 25,937.663, df = 1081, p < 0.001), confirming the appropriateness of factor analysis, as shown in

Table 5. KMO and Bartlett’s Test (Source: Author).

KMO and Bartlett’s Test			Result
Kaiser–Meyer–Olkin Measure of Sampling Adequacy		0.985	Accepted
Bartlett’s Test of Sphericity	Approx. Chi-Square	25,937.663
	df	1081
	Sig.	0.000	Accepted

KMO and Bartlett’s Test with result accepted and Table 9 also provides the EFA outcomes, indicating that individual item loadings were greater than 0.5, with all values falling within acceptable ranges and all items loaded appropriately on their respective factors.

4.2.2. CFA Test Result for Each Contract

Confirmatory Factor Analysis (CFA) was conducted on each construct as a preliminary step before testing the higher-order model, aiming to validate the consistency of observed variables with their intended constructs and ensure alignment with theoretical frameworks. The analysis revealed that all factor loadings surpassed the recommended 0.7 threshold, with model fit indices meeting established criteria. The CFA results shown in

Table 6. CFA index result for each construct.

Index	CMIN/DF	AGFI	GFI	CFI	NFI	TLI	RMSEA	RMR
Measurement	≤3	≥0.9	≥0.9	≥0.9	≥0.9	≥0.9	≤0.08	≤0.08
BU	2.64	0.96	0.99	0.99	0.99	0.99	0.04	0.01
TE	2.08	0.98	0.99	0.97	0.99	0.99	0.04	0.03
OR	2.16	0.98	0.99	0.97	0.99	0.99	0.04	0.01
EX	2.94	0.98	0.99	0.99	0.99	0.99	0.05	0.02
EN	2.85	0.97	0.99	0.99	0.99	0.99	0.05	0.01
IN	2.11	0.99	0.99	0.99	0.99	0.99	0.04	0.01
SO	2.26	0.98	0.99	0.99	0.99	0.99	0.04	0.01
Decision	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted

demonstrated exceptional model fit across all constructs, with indices consistently exceeding recommended thresholds—CMIN/DF (2.08–2.94, threshold ≤ 3), AGFI (0.96–0.99, threshold ≥ 0.9), and optimal scores of 0.99 for GFI, CFI, NFI, and TLI (threshold ≥0.9). Additionally, RMSEA (0.04–0.05) and RMR values (0.01–0.03) remained well below the acceptable threshold of 0.08. These comprehensive results confirmed strong construct validity and excellent model fit for all seven constructs in the research framework: business strategy (BU), technology context (TE), organization context (OR), external Environment context (EX), entrepreneur orientation (EN), product innovation (IN), and social media adoption (SO), leading to the acceptance of all constructs.

4.2.3. First-Order Confirmatory Factor Analysis

The First-order Confirmatory Factor Analysis (CFA) was conducted as a crucial prerequisite step before proceeding with second-order testing, validating the measurement model’s fundamental structure and ensuring that the measured variables accurately represented their intended constructs. The analysis shown in

Table 7. Fit index of First-order Confirmatory Factor Analysis.

Index	CMIN/DF	AGFI	GFI	CFI	NFI	TLI	RMSEA	RMR
Criteria	≤3	≥0.9	≥0.9	≥0.9	≥0.9	≥0.9	≤0.08	≤0.08
Result	1.55	0.93	0.94	0.98	0.95	0.98	0.03	0.16
Decision	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted

revealed significant factor loadings exceeding the 0.7 threshold, demonstrating good construct validity and reliability. The results showed excellent model fit across all indices, with strong absolute fit measures including CMIN/DF = 1.60 (threshold ≤ 3), GFI and AGFI both at 0.93 (threshold ≥ 0.9), RMSEA = 0.03, and RMR = 0.02 (both threshold ≤ 0.08), while incremental fit indices demonstrated superior fit with CFI = 0.98, NFI = 0.94, and TLI = 0.98 (all exceeding threshold ≥ 0.9) as shown in Figure 7. These comprehensive results not only met but exceeded their respective criteria, confirming that the lower-level factors were properly specified and providing a solid foundation for subsequent second-order testing while maintaining theoretical integrity and avoiding potential violations of regression assumptions.

4.2.4. Second-Order Confirmatory Factor Analysis

Second-order Confirmatory Factor Analysis plays a vital role in research by simplifying complex data structures, particularly in this research, which comprises seven constructs. The analytical results shown in

Table 8. Fit index of Second-order Confirmatory Factor Analysis.

Index	CMIN/DF	AGFI	GFI	CFI	NFI	TLI	RMSEA	RMR
Criteria	≤3	≥0.9	≥0.9	≥0.9	≥0.9	≥0.9	≤0.08	≤0.08
Result	1.60	0.93	0.93	0.98	0.94	0.98	0.03	0.02
Decision	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted	Accepted

from the Second-order Confirmatory Factor Analysis revealed strong model fit indices, demonstrating the validity of the hierarchical structure. The analysis yielded satisfactory results with CMIN/DF = 1.60, well below the acceptable threshold of 3. The model demonstrated excellent fit with GFI = 0.93, CFI = 0.98, and NFI = 0.94, all meeting or exceeding the recommended cutoff value of 0.90. The TLI value of 0.98 further supports the model’s robustness. Additionally, RMSEA = 0.03 and RMR = 0.02 indicate excellent model fit, as they are well below the 0.08 threshold as shown in Figure 8. The model fit results not only improve theoretical models by creating more nuanced frameworks but also enable researchers to account for both specific and general patterns in the data, ultimately leading to more accurate and meaningful conclusions.

The structural equation modeling analysis revealed that the Second-order Confirmatory Factor Analysis demonstrated excellent model fit indices, as shown in Table 9, with all results meeting acceptable criteria. The measurement model reported strong construct reliability with CR values (0.86 to 0.91) exceeding 0.7 and AVE values (0.55 to 0.59) above 0.5. Factor loadings across all seven factors exceeded 0.7, with business strategy (0.94), entrepreneur orientation (0.92), and social media adoption (0.92) demonstrating the strongest loadings for this model. The model exhibited strong explanatory power through R-squared values ranged from 0.73 to 0.88, with business strategy showing the highest value at 0.88 and entrepreneur factor at 0.84, indicating high explained variance. The Standard Error (S.E.) values were consistently low, ranging from 0.03 to 0.05, indicating high precision in the parameter estimates. The statistical significance of all relationships was confirmed (p < 0.001), with Critical Ratio (C.R.) values well above 1.96, collectively establishing strong construct validity and reliability of the measurement model. Additionally, all constructs exceeded the threshold for Cronbach’s alpha (>0.7). The residual measures (RMSEA = 0.03, RMR = 0.02) indicated minimal discrepancy between observed and model-implied matrices, further confirming the model’s excellent fit. These comprehensive results validate the measurement model’s reliability and validity for examining social media adoption factors among Thai digital SMEs.

Table 9. Statistical value of Second-order and EFA outcomes.

NO.	Factor Landing	EFA	S.E.	C.R.	p	CR	AVE	CA	R2
Business strategy (BU)	0.94					0.90	0.55	0.90	0.88
BU1	0.73	0.61
BU2	0.73	0.56	0.05	21.07	***
BU3	0.75	0.58	0.05	21.75	***
BU4	0.74	0.61	0.05	21.54	***
BU5	0.74	0.56	0.05	21.58	***
BU6	0.73	0.57	0.05	21.24	***
BU7	0.76	0.55	0.05	22.14	***
Technology Context (TE)	0.90					0.88	0.55	0.88	0.81
TE1	0.78	0.67
TE2	0.71	0.69	0.03	23.81	***
TE3	0.74	0.59	0.04	25.30	***
TE4	0.73	0.57	0.04	25.03	***
TE5	0.75	0.58	0.04	25.89	***
TE6	0.76	0.59
Organization Context (OR)	0.89					0.89	0.57	0.89	0.80
OR1	0.74	0.66
OR2	0.77	0.64	0.04	22.62	***
OR3	0.75	0.61	0.05	21.88	***
OR4	0.78	0.59	0.04	22.75	***
OR5	0.71	0.63	0.05	20.63	***
OR6	0.77	0.66	0.05	22.42	***
External Environment Context (EX)	0.86					0.88	0.56	0.88	0.73
EX1	0.78	0.71
EX2	0.72	0.69	0.04	24.45	***
EX3	0.73	0.59	0.04	24.69	***
EX4	0.77	0.63	0.04	26.67	***
EX5	0.73	0.63	0.04	24.74	***
EX6	0.75	0.69
Entrepreneur Orientation (EN)	0.92					0.91	0.59	0.91	0.84
EN1	0.83	0.58
EN2	0.75	0.63	0.03	24.91	***
EN3	0.74	0.67	0.04	24.79	***
EN4	0.74	0.66	0.03	24.72	***
EN5	0.79	0.60	0.03	27.06	***
EN6	0.77	0.58	0.03	25.90	***
EN7	0.76	0.60	0.04	25.76	***
Product Innovation (IN)	0.89					0.86	0.55	0.86	0.78
IN1	0.81	0.61	0.04	23.79	***
IN2	0.70	0.69	0.04	20.33	***
IN3	0.72	0.68	0.04	20.92	***
IN4	0.73	0.60	0.04	21.21	***
IN5	0.76	0.59
Social Media Adoption (SO)	0.92					0.87	0.58	870.00	0.84
SO1	0.79	0.60
SO2	0.74	0.60	0.04	23.01	***
SO3	0.74	0.61	0.04	23.01	***
SO4	0.79	0.58	0.04	24.87	***
SO5	0.72	0.54	0.04	22.27	***

statistically significant *** = p < 0.001.

Table 9. Statistical value of Second-order and EFA outcomes.

NO.	Factor Landing	EFA	S.E.	C.R.	p	CR	AVE	CA	R2
Business strategy (BU)	0.94					0.90	0.55	0.90	0.88
BU1	0.73	0.61
BU2	0.73	0.56	0.05	21.07	***
BU3	0.75	0.58	0.05	21.75	***
BU4	0.74	0.61	0.05	21.54	***
BU5	0.74	0.56	0.05	21.58	***
BU6	0.73	0.57	0.05	21.24	***
BU7	0.76	0.55	0.05	22.14	***
Technology Context (TE)	0.90					0.88	0.55	0.88	0.81
TE1	0.78	0.67
TE2	0.71	0.69	0.03	23.81	***
TE3	0.74	0.59	0.04	25.30	***
TE4	0.73	0.57	0.04	25.03	***
TE5	0.75	0.58	0.04	25.89	***
TE6	0.76	0.59
Organization Context (OR)	0.89					0.89	0.57	0.89	0.80
OR1	0.74	0.66
OR2	0.77	0.64	0.04	22.62	***
OR3	0.75	0.61	0.05	21.88	***
OR4	0.78	0.59	0.04	22.75	***
OR5	0.71	0.63	0.05	20.63	***
OR6	0.77	0.66	0.05	22.42	***
External Environment Context (EX)	0.86					0.88	0.56	0.88	0.73
EX1	0.78	0.71
EX2	0.72	0.69	0.04	24.45	***
EX3	0.73	0.59	0.04	24.69	***
EX4	0.77	0.63	0.04	26.67	***
EX5	0.73	0.63	0.04	24.74	***
EX6	0.75	0.69
Entrepreneur Orientation (EN)	0.92					0.91	0.59	0.91	0.84
EN1	0.83	0.58
EN2	0.75	0.63	0.03	24.91	***
EN3	0.74	0.67	0.04	24.79	***
EN4	0.74	0.66	0.03	24.72	***
EN5	0.79	0.60	0.03	27.06	***
EN6	0.77	0.58	0.03	25.90	***
EN7	0.76	0.60	0.04	25.76	***
Product Innovation (IN)	0.89					0.86	0.55	0.86	0.78
IN1	0.81	0.61	0.04	23.79	***
IN2	0.70	0.69	0.04	20.33	***
IN3	0.72	0.68	0.04	20.92	***
IN4	0.73	0.60	0.04	21.21	***
IN5	0.76	0.59
Social Media Adoption (SO)	0.92					0.87	0.58	870.00	0.84
SO1	0.79	0.60
SO2	0.74	0.60	0.04	23.01	***
SO3	0.74	0.61	0.04	23.01	***
SO4	0.79	0.58	0.04	24.87	***
SO5	0.72	0.54	0.04	22.27	***

statistically significant *** = p < 0.001.

5. Discussion

This research successfully developed a comprehensive framework for digital entrepreneurship in Thailand through an innovative mixed-methods approach that combined qualitative expert insights with quantitative validation. The methodological integration proved particularly effective in establishing a robust understanding of the factors driving superior long-term business performance.

The measurement model demonstrated excellent psychometric properties, with factor loadings consistently above 0.70 across all constructs. This robust factor structure aligns with similar findings by Chatterjee and Kar [68] in their study of social media adoption in emerging economies and Abdullah and Ampauleng [69] in social media usage on MSME’s performance. The model’s exceptional fit indices (CFI = 0.98, TLI = 0.98, RMSEA = 0.03) surpass those reported in comparable studies, such as the study by Ahmad et al. [70], who found similar but slightly lower fit values in their investigation of SME digital transformation.

The business strategy construct showed the highest explained variance (R² = 0.88) and strong reliability (CR = 0.90), suggesting its crucial role in social media adoption. This finding reinforces conclusions from Palladan et al. [71] and Qalati et al. [72], who identified strategic alignment as a key determinant of successful social media implementation in SMEs. The high factor loadings (0.73–0.76) for business strategy items indicate that SMEs are increasingly viewing social media as a strategic tool rather than merely a communication channel.

The technology factor (CR = 0.86) and organization factor (CR = 0.89) demonstrated strong reliability, supporting the Technology–Organization–Environment (TOE) framework’s applicability. These results parallel findings by Cvjetković [73], who reported similar construct validities in their study of digital technology adoption among Asian SMEs. The high convergent validity suggests that both technological readiness and organizational capacity are fundamental to successful social media adoption.

The external environment factor (AVE = 0.56) and entrepreneur factor (AVE = 0.59) showed significant influence on social media adoption, consistent with research by Scholtz et al. [74,75]. The entrepreneur’s role, in particular, emerged as a critical factor with the highest reliability (CR = 0.91), supporting previous findings about the importance of leadership in technology adoption decisions.

The innovation factor’s strong relationship with social media adoption (factor loadings 0.70–0.76) suggests that innovative capacity significantly influences social media implementation. This finding aligns with recent work by Kamboj [76], and Gun [77], who found similar correlations between innovation capability and digital platform adoption in SMEs.

These findings contribute to the growing body of literature on social media adoption in SMEs by validating previously identified factors while providing new insights into their relative importance and interrelationships in the current digital business environment. For practical implementation, Thai SMEs can leverage these research findings through several actionable approaches. First, businesses should integrate social media strategies into their overall business planning by setting clear objectives and KPIs aligned with their business goals (business strategy, R² = 0.88). Second, organizations need to assess and enhance their technological readiness (technology factor, CR = 0.86) through systematic infrastructure development and staff training programs. Leadership teams should actively promote digital transformation (entrepreneur factor, CR = 0.91) by cultivating a digital-first mindset and allocating appropriate resources. Additionally, establishing innovation teams and regular workshops can strengthen innovative capabilities (innovation factor, 0.70–0.76), while building strategic partnerships with technology providers can support successful implementation. These practical steps can help Thai SMEs effectively translate the theoretical frameworks into tangible business outcomes in their digital transformation journey.

6. Conclusions

This research advances our understanding of social media adoption in Thai digital SMEs through a novel integrated framework that extends beyond the traditional TOE model. This study employed a robust dual-methodology approach combining Rough Set and Fuzzy Theory with Confirmatory Factor Analysis, which validates the integration of business strategy, entrepreneurial orientation, and innovation factors with the TOE framework. The structural model demonstrated strong relationships between all constructs, with business strategy and entrepreneurial factors showing particularly strong influence on social media adoption outcomes. The findings reveal that successful digital transformation requires a holistic approach that considers strategic alignment, entrepreneurial capabilities, and innovative capacity alongside technological factors. This comprehensive approach not only provides valuable insights for practitioners but also establishes a more robust theoretical foundation for understanding social media adoption in SMEs. By incorporating these additional dimensions into the traditional TOE framework, this study contributes to both theory and practice by offering a more complete understanding of the factors that drive successful social media adoption and implementation in the context of Thai digital SMEs.

Limitations and Suggestions

This study acknowledges several methodological limitations that warrant consideration. The research sample was predominantly composed of small enterprises (44% with revenue less than 50 M Baht) using quota sampling, which may limit the generalizability of findings to larger SMEs. Additionally, the research scope was confined to non-manufacturing sectors, excluding potential insights from manufacturing industries that might be significantly impacted by digital transformation. The strategic framework analysis was narrowly focused on cost leadership and differentiation strategies, while the external environment assessment was constrained to customer and competitor factors, omitting other critical environmental elements such as technological changes, regulatory environments, and economic conditions. The cross-sectional nature of this research provided only a snapshot view rather than capturing long-term strategic impacts. To address these limitations, future research should aim to achieve the following: (1) expand the investigation to larger SMEs and manufacturing industries, (2) incorporate additional strategic frameworks, (3) broaden the external environment analysis, (4) implement mixed-methods approaches combining quantitative and qualitative insights, and (5) conduct longitudinal studies to track strategy implementation and its long-term performance impacts. These research directions would enhance our understanding of how firm size and industry context influence digital transformation outcomes in the evolving business landscape.