Business Intelligence and Environmental Sustainability: Evidence from Jordan on the Strategic Role of Green and Integrated Supply Practices

Abstract

This study examines how Business Intelligence (BI) capabilities influence environmental performance (EP) in manufacturaing supply chains, with a focus on the mediating roles of Green Supply Chain Management (GSCM) and Supply Chain Integration (SCI) and the moderating role of Blockchain Integration (BCI). Addressing a critical research gap in digital sustainability, particularly in emerging markets, this study integrates the Resource-Based View (RBV) theory, Natural Resource-Based View (NRBV) theory, and Dynamic Capabilities View (DCV) theory to develop a theoretically grounded framework. Data were collected via a cross-sectional survey of 231 managers in 65 firms in Jordan and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). Findings reveal that while BI does not directly enhance EP, it significantly improves GSCM and SCI, which in turn mediate its influence on EP. GSCM fully mediates this relationship, while SCI provides partial mediation. BCI did not demonstrate a significant moderating effect. These results suggest that BI must be embedded within green and integrative operational systems to drive sustainability outcomes. This study contributes novel insights into how digital capabilities translate into environmental gains in underrepresented contexts and provides actionable guidance for firms and policymakers aiming to align digital transformation with environmental objectives.

1. Introduction

As the imperatives of environmental sustainability and digital transformation converge, organizations, particularly in resource-constrained and environmentally vulnerable economies, face mounting pressure to operationalize sustainability while maintaining competitiveness. Business Intelligence (BI), as a data-driven decision-support capability, has emerged as a critical enabler of such dual objectives. Through the integration of real-time analytics, predictive modeling, and data visualization, BI facilitates organizational responsiveness, operational optimization, and environmental adaptability [1]. In theory, these capabilities position BI to enhance environmental performance (EP) by supporting sustainability-oriented decision-making.

However, the literature remains fragmented. Prior studies have predominantly focused on BI’s impact on financial and operational outcomes, with limited attention to its environmental implications, particularly in emerging market supply chains [2,3]. Some research suggests that BI may support sustainability through resource efficiency [4], regulatory compliance [5], and environmental reporting [6]. Yet, these contributions tend to remain descriptive, failing to empirically test the mechanisms through which BI translates into measurable environmental gains. This gap is particularly salient in developing countries like Jordan, where the industrial sector faces rising environmental scrutiny amidst a national digital transformation agenda. Despite policy-level initiatives under Jordan Vision and the Green Growth Plan, empirical investigations remain scarce regarding how BI enables environmental outcomes when embedded in supply chain practices.

Moreover, although Green Supply Chain Management (GSCM) and Supply Chain Integration (SCI) are often theorized as operational pathways to sustainability, their mediating role between BI and EP remains largely untested. Likewise, while Blockchain Integration (BCI) is promoted as a transparency-enhancing tool, its moderating role in enabling BI’s sustainability outcomes, especially in institutionally constrained and digitally developing contexts, remains theoretically underexplored [7,8]. Although institutional factors undoubtedly influence sustainability adoption, this study prioritizes intra-organizational capabilities, such as BI, GSCM, and SCI, rather than regulatory compliance dynamics. This choice aligns with this study’s aim to examine capability-based mechanisms within firm-level digital transitions. The manufacturing sector stands as one of the largest and most critical pillars of Jordan’s economy, accounting for roughly 30% of the nation’s GDP. Unfortunately, exports from Jordanian industrial companies have been declining at a rate of 0.5% annually since 2010. A significant factor behind this troubling trend is the array of challenges and issues that these companies face in relation to their environmental performance. Addressing these concerns is essential for revitalizing the sector and boosting international competitiveness [9].

This study also aligns its findings with the United Nations’ Sustainable Development Goals (SDGs), particularly Goal 12 (Sustainable Production and Consumption). It is critically important because it addresses a significant gap in understanding how BI can directly and indirectly influence EP in manufacturing firms, particularly in emerging economies like Jordan. As global sustainability challenges intensify, there is an urgent need for businesses to adopt practices that reduce environmental degradation and promote resource efficiency.

These gaps were addressed by investigating how BI influences EP in Jordanian manufacturing firms, with GSCM and SCI as mediators, and BCI as a moderator. This integrative framework responds to the call for context-sensitive, theory-informed empirical models that examine digital-environmental linkages in emerging economies.

Accordingly, this study is guided by the following research questions:

• Q1: Does BI directly improve EP in Jordanian manufacturing firms?
• Q2: Do GSCM and SCI mediate the BI–EP relationship?
• Q3: Does BCI moderate the influence of BI on GSCM and EP?

To explore these questions, this study applies Resource-Based View (RBV) theory, Natural Resource-Based View (NRBV) theory, and Dynamic Capabilities View (DCV) theory. This research conceptualizes BI as a higher-order capability. GSCM and SCI serve as transformation enablers, while BCI is treated as a boundary condition that may shape the BI–sustainability relationship depending on institutional and infrastructural maturity.

This study makes novel contributions to the literature. First, it provides empirical evidence for the indirect pathways through which BI influences EP by positioning GSCM and SCI as mediating mechanisms, thereby clarifying a previously under-theorized relationship. Second, it develops a theoretically integrated model that combines the RBV, NRBV, DCV, and BCI into a cohesive digital-sustainability framework, addressing the absence of multi-theoretical syntheses in existing scholarship. Third, situated within Jordan’s industrial sector, this study gives credence to ongoing calls by scholars for contextual insight into sustainable digital transformation in a developing economy [10]. Furthermore, it informs policy and managerial practice by identifying actionable digital levers for environmental improvement within emerging market supply chains, thereby supporting strategic decision-making for sustainability and competitiveness.

Practically, the integration fosters intelligent and sustainable supply chains. It delivers real-time, validated insights for green decision-making, enables secure and seamless data sharing among partners, and addresses challenges like verifying green compliance and ensuring traceability, which are difficult to solve with BI or BCI alone. This combination aligns with the principles of Industry 5.0, which emphasizes human-centric, sustainable, and resilient supply networks powered by intelligent, decentralized technologies. Finally, integrating BI with GSCM, SCI, and BCI is a novel approach in the fields of supply chain and information systems. This novelty stems from the synergy of combining BI’s data-driven decision-making with GSCM’s environmental goals, SCI’s focus on partner collaboration, and BCI’s secure transaction management. Current research often examines these areas in isolation, overlooking their combined potential. Therefore, studying this integration is crucial for advancing both theoretical knowledge and practical management, offering a path for companies to build digitalized, sustainable, and trustworthy supply chains in a complex global market.

2. Literature Review

2.1. BI and Environmental Sustainability

BI is defined as a set of tools ranging from data analytics, monitoring in real-time, and decision support that convert raw data into the insight required for making decisions at strategic and operational levels [4,11]. In sustainability terms, BI enables organizations to assess, forecast, and optimize environmental consequences by tracking inefficiencies concerning resource use, emissions, and waste disposal. According to researchers, BI allows organizations to keep track of carbon metrics and environmental indicators in real time so as to direct organizational operations towards sustainability goals [12,13].

More recent papers have extended BI applications to different fields. An instance includes BI promoting energy efficiency and resource efficiency in healthcare with dynamic adjustments of operations depending on data trends [8]. This role extends to environmentally sensitive sectors such as manufacturing, where BI helps to guarantee compliance with environmental standards, transparency in process, and risk identification.

As pertains to Jordan and the MENA region in general, the adoption of BI is still rising. As [2] stated, although BI is gaining traction, it is primarily leveraged for financial and operational benefits and into environmental decision-making to a very limited degree [2]. According to [11], on the other hand, BI does play a key role in enabling data-driven green strategies in Jordanian SMEs, yet the environmental implications of such BI applications remain largely under-theorized and empirically untested [11]. This observation sets the agenda for the further exploration of how BI can be made to go beyond a purely analytical tool into an enabler of sustainability in emerging economies.

Although, Big Data Analytics (BDAs) are more predictive regarding ecological optimization [6], evidenced how BDAs enabled EP via scenario forecasting and supply chain visibility in ASEAN and Egyptian industries [6]. These insights support the premise that BI is scalable and transferable across regions and sectors, but the mechanisms through which BI contributes to EP, particularly through mediating supply chain practices, remain insufficiently explored.

Moreover, mixed findings are reported in the literature on BI’s success in attaining environmental ends. Dubey and others state that [5] organizational inertia, misuse of technology, or skills might hamper the realization of BI’s full sustainability potential [5]. Such contradictions suggest a gap at the theoretical and empirical levels: Therefore, the question arises whether BI might have an indirect input into EP through operational means such as GSCM and SCI.

2.2. Supply Chain Integration, Green Supply Chain Management, and Environmental Performance

SCI is the alignment of internal processes and external relationships across supply chains in order to achieve improved strategic coordination. Researchers maintain that SCI offerings enhance environmental learning through enhanced data sharing and collaborative planning in real time, thus allowing firms to perceive regulatory requirements and lessen the environmental impact accordingly [14,15]. Others go a step further to state that treating SCI in its multidimensional nature, concerning the information and flows of materials, results in greater supply chain agility and responsiveness in sustainability settings [16]. It allows for dynamically aligning environmental strategies across suppliers and distributors and internal departments while converting BI-derived informational insight into coordinated sustainability action.

GSCM extends the concept of SCI by infusing sustainability concerns into procurement, production, logistics, and end-of-life product management. Studies provide empirical evidence on GSCM’s role in mitigating environmental harm and enhancing supply chain resilience [17,18,19]. In the MENA region, Hanna has examined GSCM’s contribution to operational efficiency and compliance under environmental constraints [20], while Ilyas and others affirm that GSCM practices are critical to achieving the Sustainable Development Goals (SDGs) through proactive waste reduction and eco-efficient logistics strategies [21].

Based on the definition of [6], GSCM is the integration of environmental thinking into supply chain management itself, including product design, material sourcing and selection, manufacturing processes, delivery of the final product, and end-of-life management [19]. Through this great orientation, GSCM can be seen as a channel through which environmental insights from BI are put into action, linking strategic flows of information to real ecological outcomes.

An emerging stream of research delves into interactions between SCI and GSCM, especially in the presence of digital technologies. Dubey and others [5] evidence that supply chain collaboration is a positive antecedent for green practices implementation, where BI systems serve as technological enablers [5]. Richards and others indicate deeper SCI levels and more coherent GSCM strategies when firms exploit BI [22]. These findings would foreground a synergistic model where BI acts as a catalyst for integration and greening; thus, SCI and GSCM are more believable mediators of BI’s effects upon EP.

According to some studies, EP refers to the measurable results of activities undertaken by an organization affording or causing impacts on the natural environment, including qualifications like reduced emissions, reduced waste generation, reduced use of energy, and better resource efficiency [23]. Drawing from the Jordanian context, Al-Ghwayeen and Abdallah went ahead to empirically attest to certain EP indicators such as emissions control and waste recovery, establishing them as GSCM strategic practices and regulatory compliance [24].

2.3. BCI, BI, GSCM, and EP

Into organizational systems, BCI is the act of incorporating blockchain technology into current supply chains, information systems, and corporate processes in order to improve operational efficiency, security, and transparency [25]. Blockchain has important benefits like permanent record keeping, real-time data exchange among stakeholders, and less dependence on middlemen because it is a centralized and tamper-evident ledger [26].

Integrating blockchain into organizational systems facilitates trust building, especially in contexts that require secure data exchange and provenance tracking, such as supply chain management, finance, and healthcare [7]. For example, BCI into supply chains improves traceability, authentication, and sustainability reporting by ensuring data integrity across distributed networks [27]. However, the success of this integration requires addressing interoperability challenges, regulatory concerns, and scalability limitations of existing blockchain infrastructures [28].

Recent studies indicate that combining blockchain with other emerging technologies such as the Internet of Things and BI can enhance its strategic value and practical applicability within organizations [29]. The value of BCI in striking a prudent balance between GSCM and EP through the facilitation of verifiable information sharing is further highlighted by recent editorial perspectives by [8]. This allows firms to simultaneously pursue operational excellence and sustainability-oriented exploration.

The above references previously found in the literature support the assumption that BCI and GSCM are not only operational constructs but mechanisms through which BI insights bring about tangible environmental benefits, especially in emerging economies like Jordan, in which regulatory pressures and digital maturity are growing hand in hand.

3. Theoretical Framework and Hypotheses Development

3.1. Theoretical Framework

The RBV theory posits that a firm’s sustained competitive advantage is predicated upon the possession of resources exhibiting characteristics of value, rarity, inimitability, and non-substitutability [30]. Within this theoretical framework, BI emerges as a firm-specific analytical capability that facilitates superior data-driven decision-making, thereby fostering strategic differentiation. However, a notable limitation of the RBV is its inherent lack of explicit consideration for ecological factors. This lacuna is addressed by the NRBV theory [31], which augments the RBV by establishing a nexus between environmental competencies—specifically pollution prevention, product stewardship, and sustainable development—and competitive advantage.

In this context, BI operationalizes the pathways delineated by NRBV through its capacity to identify process inefficiencies and waste (pollution prevention), support eco-design and cleaner production initiatives (product stewardship), and inform proactive, long-term sustainability strategies (sustainable development) [32]. Furthermore, GSCM and SCI function as crucial processual mediators, translating BI insights into enhanced EP. This mediation enables organizations to internalize environmental knowledge into their operational routines and collaborative endeavors, as argued by [17,19].

Additionally, the DCV theory contextualizes BI within a firm’s adaptive processes. Teece and others assert that sensing, seizing, and transforming capabilities enable firms to respond to environmental volatility [33]. BI provides the informational backbone for these capabilities, especially if integrated with SCI and GSCM practices [34].

The supply chain ambidexterity (SCA) theory is an additional viewpoint to enrich the theoretical base and understand how organizations pursue efficiency (exploitation) and adaptability (exploration) regarding environmental sustainability. According to [35,36], SCA is an organization’s ability to maintain and integrate efficient supply chain routines that are stable and are grounded in innovation-oriented practices characterized by change. In this context of BI-driven supply chains, ambidexterity allows entities to either exploit historical data to optimize current operations or explore with predictive analytics to spot emerging sustainability opportunities and act upon them. This capability is particularly critical in resource-constrained emerging economies, exemplified by Jordan.

Collectively, the RBV, NRBV, DCV, and SCA furnish a robust theoretical foundation for comprehending how BI, mediated by GSCM and SCI, not only elevates EP but also cultivates organizational adaptability and innovation capacity. Consistent with these theoretical underpinnings, the conceptual model proposed herein integrates hypothesized relationships among BI, GSCM, SCI, EP, and BCI. The model posits that BI exerts both direct and indirect influences on EP, with GSCM and SCI serving as mediating variables. Furthermore, BCI is hypothesized to moderate the effects of GSCM and SCI on EP.

Grounded in the NRBV and DCV, this model underscores a dual emphasis on environmental competencies and the dynamic organizational capabilities necessary for navigating complex market environments. In accordance with the strategic pathways articulated by [31], GSCM and SCI function as conduits through which the analytical prowess of BI is transmuted into tangible sustainability outcomes. BCI, characterized by its inherent transparency, immutability, and decentralization, is hypothesized to fortify these pathways by enhancing traceability, trust, and verification mechanisms within supply chains. This moderating role of BCI is particularly salient in nascent and resource-constrained contexts, such as Jordan, where challenges pertaining to governance, compliance, and transparency persist [7]. Recent editorial perspectives by [8] researchers further accentuate the value of BCI in achieving a judicious balance between supply chain efficiency and innovation through the facilitation of verifiable information sharing, thereby enabling firms to concurrently pursue operational excellence and sustainability-oriented exploration [8].

3.2. Research Model and Hypotheses Development

The conceptual model proposed in this study is illustrated in the following diagram, which reflects the hypothesized relationships among BI, GSCM, SCI, EP, and BCI. Specifically, how BI influences EP both directly and indirectly through GSCM and SCI as mediators; in addition, BCI is hypothesized to moderate the paths from BI to GSCM and from BI to EP.

The model (Figure 1) is grounded in the NRBV and DCV frameworks and aligns with empirical observations from prior research. The NRBV emphasizes environmental sustainability as a strategic source of competitive advantage through green innovation and resource stewardship [31]. The DCV highlights an organization’s ability to sense, seize, and transform internal and external resources in response to changing environments [33].

Guided by previous literature and perspectives, all the following hypotheses are supported by literature and structured to reflect direct, mediating, and moderating relationships.

3.2.1. Direct Hypothesis: BI, EP, GSCM, and SCI

Studies show that BI enables firms to track and respond to environmental metrics in real time, improving compliance and facilitating sustainability reporting [6,12]. Dubey and others [5] find that BI adoption enhances environmental agility by enabling predictive analytics [5], while Watson and Wixom emphasize its role in strategic ecological decision-making [32]. Yet, the effectiveness of BI in directly impacting EP varies across contexts [11]. DCV positions BI as a dynamic sensing capability essential for ecological adaptation, and that leads to

H1. 

BI positively affects EP.

BI enables granular monitoring of suppliers, waste streams, and emissions, fostering GSCM practices such as green procurement and eco-design [13,21]. Studies suggest BI plays a critical enabling role in greening supply chains through real-time analytics [8,37]. NRBV identifies such green capabilities as key to sustainable advantage, while DCV views BI as essential for operational transformation. So, the next hypothesis was developed to

H2. 

BI positively affects GSCM.

Extensive empirical research confirms the link between GSCM and EP. It shows that green practices reduce waste and emissions [19], while others find that GSCM leads to regulatory compliance and operational efficiency [17,24]. Within the NRBV framework, GSCM constitutes an environmental competence that enhances performance outcomes. And that leads to

H3. 

GSCM positively affects EP.

Green practices often necessitate collaborative alignment across supply chain actors. Soares and others show that GSCM fosters synchronized operations and shared environmental goals [16]. Others argue that GSCM serves as a relational enabler that strengthens supply chain cohesion [5]. This aligns with DCV, where GSCM acts as a transformative mechanism for integration. And accordingly, H5 was developed:

H5. 

GSCM positively affects SCI.

BI enhances internal and external supply chain alignment by improving data transparency and planning synchronization. Some researchers argue that BI increases supply responsiveness through analytics-driven coordination [11,14,22]. DCV identifies BI as a core capability that drives integration under dynamic market conditions. Thus, the following hypothesis can be proposed:

H6. 

BI positively affects SCI.

Integrated supply chains facilitate coordinated environmental strategies, yielding improved EP [15,36] shows that SCI enhances data sharing critical for green innovation [15]. Al-Ghwayeen et al. [24] further link SCI with tangible EP gains in Jordanian firms [15,24]. NRBV and DCV jointly view SCI as an enabler of sustained ecological outcomes. Consequently, the following hypothesis is proffered:

H7. 

SCI positively affects EP.

3.2.2. Mediating Hypothesis: GSCM and SCI on Relationship Between BI and EP

Studies argue that BI’s impact on EP is contingent on its translation into green operational practices [21,37]. GSCM channels BI insights into waste reduction and compliance routines. DCV conceptualizes BI as a sensing mechanism, while GSCM represents the seizing and transforming functions that link BI to EP. And that leads to

H4. 

GSCM mediates the relationship between BI and EP.

Researchers show that BI enhances SCI by fostering inter-organizational communication and planning [5,22]. SCI, in turn, facilitates environmental collaboration and innovation [16]. Under DCV, this mediating pathway reflects how analytical capabilities such as BI are operationalized through integrative mechanisms. Thus, the following hypothesis can be proposed:

H8. 

SCI mediates the relationship between BI and EP.

3.2.3. Moderating Hypothesis: BCI Moderates the Paths from BI to GSCM and from BI to EP

Benzidia, Kouhizadeh, and others [8] highlight blockchain’s ability to increase transparency and traceability, enhancing green outcomes [8,26]. BCI may amplify the effect of BI on the GSCM link by fostering trust and reducing opportunism. From a DCV lens, BCI is a boundary condition that enhances the efficacy of operational routines. And so

H9. 

BCI positively moderates the relationship between BI and GSCM.

Blockchain technologies strengthen BI by improving data integrity and transaction security, especially in fragmented supply networks [7,38]. This can intensify the impact of BI on EP. DCV supports this logic, framing BCI as a contingency resource that enhances collaborative sustainability, particularly in volatile and uncertain environments. Thus, the following hypothesis can be proposed:

H10. 

BCI positively moderates the relationship between BI and EP.

4. Methodology

4.1. Research Design and Methodological Justification

This study employs a cross-sectional, quantitative design to explore how BI influences EP both directly and indirectly through GSCM and SCI as mediators; in addition, the moderation effect of BCI for the paths from BI to GSCM and from BI to EP. The framework integrates RBV, NRBV, and DCV to capture how firm-level digital capabilities interact with environmental practices in emerging markets like Jordan. All constructs were modeled as first-order reflective variables, measured on a 5-point Likert scale (1 = Strongly Disagree and 5 = Strongly Agree). Items were adapted from validated sources as follows: BI [11], EP [24], GSCM [21], SCI [16], and BCI [8].

Given the model’s complexity and presence of indirect and interaction effects, Partial Least Squares Structural Equation Modeling (PLS-SEM) was selected. PLS-SEM is well-suited for moderate sample sizes (n = 231), non-normal data, and complex models involving latent constructs [39]. SmartPLS 4 was used for structural modeling, while IBM SPSS Statistics version 27 handled preliminary data screening. Bootstrapping with 5000 subsamples was applied to ensure statistical rigor.

4.2. Research Questions and Hypotheses

This study examines ten hypotheses classified as follows: direct effects (H1, H2, H3, H5, H6, and H7), mediation effects (H4 and H8), and moderation effects (H9 and H10). GSCM and SCI are tested as mediators between BI and EP. BCI is modeled and tested as a moderator on the paths from BI to GSCM and from BI to EP.

Interaction terms were constructed using standardized latent variables (e.g., BI × BCI) following Henseler & Chin’s product-indicator approach for moderation analysis in PLS-SEM [40]. Each hypothesis is grounded in the NRBV and DCV frameworks, emphasizing how strategic digital capabilities interact with green and integrative supply chain practices to influence environmental outcomes in emerging economies.

4.3. Sampling Strategy and Sample Adequacy

A stratified random sampling approach was employed, targeting 65 prominent manufacturing firms across Amman, Zarqa, and Irbid regions selected for their industrial concentration and adoption of BI and supply chain practices. Firms were drawn from the Jordan Chamber of Industry database to ensure contextual alignment.

Between March and May 2025, 350 questionnaires were distributed through HR departments via email, mobile apps, and social media. The target respondents were departmental managers in functions linked to BI and supply chain processes (e.g., Quality, IT, Warehousing, and Procurement).

Participation was entirely voluntary, and all respondents were assured of confidentiality and anonymity prior to engaging with the survey. Of the 350 distributed questionnaires, a final response pool of 231 valid submissions was achieved (n = 231), yielding a response rate of approximately 66%, which exceeds the statistical power requirements established by G*Power 3.1 (α = 0.05, power = 0.80, effect size f² = 0.15; minimum sample = 119).

Given that Arabic is the native language of the targeted respondents, the original English-language questionnaire was translated using the back-translation technique [41]. To ensure semantic accuracy and cultural appropriateness, the translated version was further reviewed by academic experts in supply chain management and sustainability in the Jordanian context, following established cross-cultural instrument validation guidelines [42].

The questionnaire was structured into three main sections as follows: Section A: Firm Information; Section B: Respondent Demographics Information; and Section C: Measurement Items for BI, EP, GSCM, SCI, and BCI. This structure is consistently reflected in the accompanying Excel dataset used for analysis.

5. Analysis of Data and Results

5.1. Measurement Model Assessment

Measurement reliability and validity were established through multiple tests: Internal consistency was confirmed (Cronbach’s α and Composite Reliability > 0.70) [39]. Convergent validity was met (AVE > 0.50); items with loadings < 0.70 were retained if CR and AVE remained acceptable [39,43]. Discriminant validity was established using Fornell and Larcker criteria [43], HTMT ratios (<0.85) [40], and cross-loading analysis to ensure indicator specificity.

To ensure data integrity prior to PLS-SEM analysis, the following procedures were applied: Missing values accounted for less than 5% and were handled using mean imputation, suitable under conditions of low missingness; outliers were assessed using standardized z-scores (±3.29) and Mahalanobis distance; no extreme cases warranted removal. Although PLS-SEM is robust to non-normality, skewness and kurtosis statistics were examined to confirm acceptable distributional properties [39].

To address common method bias (CMB) and ensure construct independence, Harman’s single-factor test revealed that no single factor accounted for more than 50% of the total variance, indicating an absence of substantial CMB. Variance Inflation Factors (VIFs) for all reflective constructs were below 3.3, suggesting low multicollinearity and minimal risk of CMB [44].

Furthermore, descriptive statistics and preliminary screening (missing data, outliers, and distribution) were conducted in IBM SPSS Statistics. Structural relationships were tested using PLS-SEM in SmartPLS 4, applying bootstrapping (5000 subsamples) and bias-corrected confidence intervals.

Key evaluation steps included the following:

• R² values to assess the explanatory power of endogenous constructs.
  f² effect sizes to evaluate the magnitude of predictor contributions.
• Q² predictive relevance using blindfolding procedures to assess out-of-sample predictive capability.
• Direct and mediated paths were assessed through multiple regression models (e.g., BI → EP, BI → GSCM → EP, BI → SCI → EP).
• Mediation analysis using bootstrapped confidence intervals and Variance Accounted For (VAF):

  ○

  GSCM: ~84 (full mediation)

  ○

  SCI: ~66% (partial mediation)

• Moderation testing (BI × BCI) followed the Henseler & Chin product-indicator method [45].
• One-way ANOVA showed significant differences in EP across firm sizes (F (3, N–4) = 6.02, p = 0.0006).

These procedures ensured robust estimation of the structural pathways among BI, EP, GSCM, SCI, and BCI.

5.2. Results

This section presents the empirical results of the structural model, assessed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS 4. A bootstrapping procedure with 5000 subsamples and bias-corrected confidence intervals was employed to test the significance of path coefficients, including direct, mediated, and moderated effects. Standardized latent interaction terms were used to evaluate moderation, following the product-indicator method [45].

Predictive relevance (Q²) was assessed through the blindfolding procedure. Preliminary data screening, including checks for missing values, outliers, and descriptive statistics, was conducted using IBM SPSS Statistics to ensure data quality and support further interpretation.

Robustness tests, including ANOVA, effect size (f²), and multicollinearity diagnostics, were performed to confirm result stability. The results are contextualized within Jordan’s industrial environment to explain both significant and non-significant outcomes.

5.2.1. Firm Information

Table 1. Characteristics of participating firms.

Firm Characteristic	Category	Frequency (n)	Percentage (%)
Number of (Full-Time) Employee	Less than 100	40	17.3
	100–200	51	22.1
	201–300	49	21.2
	Above 300	91	39.4
Industry	The Therapeutic Industries and Medical Supplies Sector	27	11.7
	Chemical and Cosmetic Industries Sector	46	20
	Engineering, Electrical, and Information Technology Industry Sector	12	5.2
	Plastic and Rubber Industries Sector	24	10.4
	Wood and Furniture Industries Sector	16	6.9
	Leather and Garment Industries Sector	21	9
	The Food, Catering, Agricultural, and Livestock Industries Sector	41	17.7
	The Packaging, Paper, Cardboard, Printing, and Office Supplies Industry	20	8.7
	Construction Industries Sector	15	6.5
	Mining Industries Sector	9	3.9
Firm age	1–5 years	42	18.2
	6–15 years	40	17.3
	16–30 years	63	27.3
	Above 30 years	86	37.2

provides a detailed overview of the characteristics of participating firms included in the final sample. A diverse representation across sectors and firm sizes enhances the external validity and relevance of findings. The largest group of firms (39.4%) had over 300 full-time employees, reflecting the participation of established manufacturing firms with structured operations.

This sectoral and size-based diversity supports generalization within the Jordanian manufacturing context and mitigates selective sampling bias.

5.2.2. Respondent Demographics Information

Table 2. Demographic variables of the survey respondents.

Demographic Variable	Category	Frequency (n)	Percentage (%)
Gender	Male	167	72.3
	Female	64	27.7
Age	Less than 30	8	3.5
	31–40	83	35.9
	41–50	89	38.5
	51–60	48	20.8
	Above 60	3	1.3
Education	Intermediate Diploma or below	28	12.1
	Bachelor	146	63.2
	Master	52	22.5
	Doctorate	5	2.2
Department Manager	Information Technology and Database Administrators	59	25.5
	Quality Assurance	32	13.9
	Import and Export	18	7.8
	Warehouses	20	8.7
	Supplies and Procurement	56	24.2
	Others	46	19.9
Work Experience	1–5 years	5	2.2
	6–10 years	44	19.1
	11–15 years	65	28.1
	16–20 years	51	22
	Above 20 years	66	28.6

outlines demographic attributes of the 231 valid responses. The respondent pool was dominated by male managers (72.3%), with a substantial proportion aged 41–50 years (38.5%). Department representation included key managerial roles related to BI and supply chain practices, particularly IT (25.5%) and procurement functions (24.2%), ensuring alignment with the study constructs. Importantly, 97.8% of participants had over five years of professional experience, ensuring informed and reliable insights. This depth of domain knowledge contributes to the credibility of perceptual data used in the measurement model.

5.2.3. Measurement Model Diagnostics

Table 3. Measurement items, sources, and loadings.

Construct and Sources	Item Code	Items	Factor Loading
Business Intelligence (BI) (Mbima & Tetteh, 2023 [11])	BI1	To what extent do the organization’s data integration systems serve as data sources?	0.81
	BI2	To what extent does your organization depend on spreadsheets and databases as data sources?	0.78
	BI3	To what extent does the organization’s data warehouse or data marts serve as data sources?	0.79
	BI4	Full integration of data enables real-time monitoring and analysis	0.82
	BI5	To what extent the organization’s information technology systems used to produce reports are?	0.80
	BI6	How extensively does your organization use online analytical processing (OLAP)?	0.77
	BI7	To what extent is the organization using analytical applications, such as trend analysis and “what if” scenarios?	0.81
	BI8	To what extent are cloud data services used in your organization?	0.80
	BI9	To what extent are dashboards used to monitor activities in your organization?	0.82
Environmental Performance (EP) (Al-Ghwayeen & Abdallah, 2018 [24])	EP1	Our firm has reduced consumption of hazardous/toxic material during the last three years compared to competitors	0.84
	EP2	Our firm has reduced air emissions during the last three years compared to competitors	0.81
	EP3	Our firm has reduced effluent wastes during the last three years compared to competitors	0.82
	EP4	Our firm has sought to improve its environmental image /position during the last three years compared to competitors	0.85
	EP5	Our firm has reduced energy consumption during the last three years compared to competitors	0.81
	EP6	Our firm has reduced solid wastes during the last three years compared to competitors	0.80
Green Supply Chain Management (GSCM) (Ilyas et al., 2020 [24])	GSCM1	Our supplier firm and we jointly developed environment-conscious products	0.86
	GSCM2	We provided our suppliers with technical, managerial and financial assistance to address environmental issues	0.83
	GSCM3	We provided our suppliers with relevant and helpful information on how to comply with our environmental requirements	0.84
	GSCM4	We demanded our suppliers to develop environmental-friendlier products	0.85
	GSCM5	We conducted environmental audits of our suppliers on a regular basis	0.82
	GSCM6	We demanded our suppliers to establish environmental management systems	0.79
	GSCM7	We assess our suppliers’ environmental performance through a formal and green procurement process	0.80
Supply Chain Integration (SCI) (Soares et al., 2017 [16])	SCI1	Our company creates supply chain teams that include members from different companies	0.80
	SCI2	Our company extends the supply chain to include members beyond immediate suppliers	0.77
	SCI3	Our company extends the supply chain to include members beyond our direct customers	0.76
	SCI4	Our company improves the integration of activities across the supply chain	0.78
	SCI5	Our company creates a greater level of trust among supply chain members	0.82
	SCI6	Our company involves all members of the supply chain in product/service/marketing plans	0.79
	SCI7	Our company participates in sourcing decisions of suppliers	0.77
	SCI8	Our company seeks new ways to integrate supply chain activities	0.80
	SCI9	Our company aids suppliers in increasing their capabilities	0.81
	SCI10	There is a compatible communication/ information system with suppliers	0.79
Blockchain Integration (BCI) (Benzidia et al., 2021 [8])	BCI1	Improves communication with suppliers by strengthening the security of data exchanged in terms of customer preference of the buying firms’ product information.	0.79
	BCI2	Improves exchanges of information with suppliers about product demand and feedback (customer request)	0.76
	BCI3	Improve the exchange of information of strategic suppliers in the design phase	0.80

presents the measurement items used to operationalize each latent construct, along with their standardized factor loadings derived via PLS-SEM. All items exceeded the threshold of 0.70, confirming indicator reliability and convergent validity in line with [39]. These results validate the reflective measurement model structure employed in the subsequent structural analysis.

The validated indicators provide a reliable empirical foundation for analyzing the influence of BI on EP, including both direct pathways and those mediated by GSCM, SCI, and moderated by BCI (Figure 2).

5.2.4. Model Fit and Predictive Accuracy

To assess structural model adequacy, both goodness-of-fit indices and predictive relevance indicators were examined. The Standardized Root Mean Square Residual (SRMR) was 0.056, well below the conservative threshold of 0.08 [40], indicating strong model–data fit. Additional diagnostics confirmed structural validity:

• d_ULS = 0.792 and NFI = 0.91, exceeding acceptable cutoffs.
• R² values (

  Table 4. Model fit and predictive accuracy.

  Construct	R2	Q2
  GSCM	0.70	0.52
  SCI	0.63	0.47
  EP	0.78	0.56

  ) demonstrated strong explanatory power for all endogenous constructs:
• GSCM (R² = 0.70): 70% of the variance in GSCM is explained by BI.
• SCI (R² = 0.63): demonstrates a high degree of influence by BI and GSCM on integration mechanisms.
• EP (R² = 0.78): Hhgh variance in environmental outcomes explained by the model.
• Q² values (Table 4) calculated using (Stone–Geisser blindfolding) [46] confirmed predictive relevance: GSCM (Q² = 0.52), SCI (Q² = 0.47), and EP (Q² = 0.56) all exceeded zero, indicating strong out-of-sample predictive validity.

These figures collectively validate the model’s ability to generalize beyond the sample and predict future observations critical for management-oriented studies.

Direct effects (H1, H2, H3, H5, H6, and H7): Path coefficients, confidence intervals, effect sizes (f²), and significance levels for the direct structural paths were estimated using PLS-SEM with bootstrapped confidence intervals and interpreted as follows (

Table 5. Hypothesis testing summary.

Hypothesis	Path	Β	p-Value	f2	Support
H1	BI → EP	0.09	>0.05	0.01	Not Supported
H2	BI → GSCM	0.84	<0.001	0.40	Supported
H3	GSCM → EP	0.47	<0.001	0.25	Supported
H5	GSCM → SCI	0.74	<0.001	0.35	Supported
H6	BI → SCI	0.28	<0.001	0.09	Supported
H7	SCI → EP	0.46	<0.001	0.21	Supported

):

Interpretations:

H1 (BI → EP): not supported. BI’s direct effect on EP was non-significant (β = 0.09, p > 0.05), suggesting that BI alone does not drive sustainability outcomes without integration into operational processes. Although BI is conceptually linked to improved EP through enhanced decision-making [5,32], this might be explained by different reasons; firstly, operational Disconnection: BI capabilities may exist but are not embedded in environmental routines like eco-auditing or lifecycle assessments. Secondly, adoption Immaturity: Jordanian firms may use BI for basic reporting, not for strategic environmental innovation. Finally, contextual Pressures: Environmental improvements may stem more from external compliance demands than from BI-driven insights.

H2–H3: supported. BI significantly enhances GSCM, which in turn positively influences EP, highlighting the mediating role of green practices.

H5–H6: supported. BI strengthens SCI both directly and via GSCM, reflecting complementary pathways.

H7: supported. SCI significantly improves EP, underscoring the value of supply chain coordination.

Overall, BI enhances EP indirectly through GSCM and SCI, supporting this study’s mediational framework (Figure 3).

Mediation analysis evaluated the indirect effects of BI on EP through GSCM and SCI, using bootstrapped confidence intervals and Variance Accounted For (VAF) (

Table 6. Mediation effects summary.

Mediation Path	β	95% CI	p-Value	VAF	Mediation Type
BI → GSCM → EP	0.48	[0.35, 0.60]	<0.001	0.84	Full
BI → SCI → EP	0.41	[0.30, 0.53]	<0.001	0.66	Partial

).

Interpretations:

The full mediation effect of GSCM indicates that BI influences EP exclusively through its impact on GSCM. The non-significant direct path from BI to EP, combined with a strong and significant indirect path (β = 0.48, p < 0.001), supports full mediation. This finding is consistent with the DCV [33], which emphasizes that data-driven competencies like BI must be embedded within dynamic operational routines—such as green sourcing, eco-design, and waste reduction—to yield sustainable performance outcomes.

The partial mediation effect via SCI demonstrates that SCI acts as a complementary mechanism through which BI enhances EP. Here, both the direct and indirect paths are significant (indirect effect β = 0.41, p < 0.001; VAF = 0.66), signifying partial mediation. Contrary to common misconceptions, partial mediation does not imply that the mediation effect is weak or insignificant. Rather, it shows that BI improves EP both directly and indirectly by fostering information-sharing, coordination, and alignment across the supply network. This supports a dual-pathway model, where both GSCM and SCI are essential for translating intelligence capabilities into environmental advantage.

Moderation Analysis: BCI (H9 and H10). Moderation was tested using standardized interaction terms. The results indicate non-significant moderation effects (

Table 7. Moderation effects: BCI.

Moderation Path	Β	95% CI	p-Value	Decision
BI × BCI → GSCM	−0.03	[−0.14, 0.08]	0.61	Not Supported
BI × BCI → EP	0.02	[−0.07, 0.11]	0.69	Not Supported

):

Interpretations:

The following reasons justify why the H9 and H10 are (not supported) by BCI Moderation: firstly, limited BCI Deployment: Blockchain technologies remain uncommon or symbolic in most participating firms, which may reduce their moderating power. Secondly, Digital Misalignment: Without a common IT infrastructure across the supply chain, blockchain’s benefits, such as traceability or trust, are difficult to actualize. And finally, Policy and Market Gaps: The absence of mandatory sustainability traceability weakens the business case for blockchain’s full deployment in environmental integration. These results underscore that technical capabilities alone do not guarantee EP gains. In emerging economies, successful outcomes depend on organizational maturity, supply chain alignment, and regulatory scaffolding, not just on digital infrastructure. To conclude, the absence of moderation may reflect a technological readiness gap. Blockchain adoption in Jordanian firms may be at a developmental stage insufficient to amplify or alter the effect of BI on sustainability practices.

The following subsection presents foundational SPSS-derived descriptive statistics for the primary constructs in this study, including measures of central tendency and distribution. Additionally, the results of a one-way ANOVA are reported to evaluate whether perceived (EP) varies significantly across firms of different sizes (

Table 8. Descriptive statistics for study constructs.

Construct	Mean	Standard Deviation (SD)	Skewness	Kurtosis
Business Intelligence (BI)	3.84	0.76	−0.21	−0.58
Green Supply Chain Management (GSCM)	3.91	0.72	−0.37	−0.22
Supply Chain Integration (SCI)	3.76	0.81	−0.11	−0.49
Environmental Performance (EP)	3.68	0.85	−0.08	−0.71
Blockchain Integration (BCI)	3.25	0.96	−0.04	−0.63

Note: All variables were measured on 5-point Likert scales. Skewness and kurtosis values fall within acceptable thresholds (±1), indicating approximate normality.

and

Table 9. One-way ANOVA results for EP by firm size.

Source of Variation	Sum of Squares	Df	Mean Square	F-Value	p-Value
Between Groups	6.93	3	2.31	6.02	0.0006
Within Groups	87.52	227	0.39
Total	94.45	230

Note: The analysis shows a statistically significant difference in EP scores across firm size categories (p < 0.001). Post-hoc tests are recommended to identify specific group differences.

).

6. Discussion

This study investigated how BI impacts EP through GSCM and SCI while also examining the moderating role of BCI. The key finding was that BI had no direct effect on EP but exerted significant indirect influence through GSCM and SCI, highlighting their central mediating roles. This aligns with the DCV, which emphasizes the embedding of digital capabilities within operational processes to generate strategic outcomes [33].

The full mediation via GSCM and partial mediation via SCI support the NRBV framework. BI facilitates pollution prevention through GSCM and promotes product stewardship via enhanced SCI. These results extend the NRBV by showing that BI functions as an enabling resource, activating environmental strategies when channeled through operational and relational capabilities.

The significant BI → GSCM and BI → SCI paths validate BI’s role as a dynamic asset under the RBV and DCV, corroborating some recent studies [5,11]. However, the null effect of BI → EP contradicts findings from [12], suggesting that in the Jordanian context, BI’s strategic utility is realized only through green operational integration [6,12].

BCI, contrary to expectations, did not moderate the BI–GSCM or BI–EP relationships. This suggests a lack of technological readiness or institutional maturity to activate blockchain’s potential, as echoed in previous studies [8,26].

The findings also emphasize supply chain ambidexterity, wherein BI enhances both responsiveness (via GSCM) and alignment (via SCI), supporting Soares, Ilyas, and others [16,21]. The mediating strength of GSCM is especially critical in transforming BI insights into sustainability outcomes, reinforcing prior evidence from [24].

The strong link between BI, on the one hand, and GSCM and SCI, on the other hand, is in accordance with RBV, which sees BI as a strategic asset that provides better organizational responsiveness and coordination. The results are also in accordance with the DCV perspective, seeing BI as a sensing and seizing mechanism for organizational conversion under environmental volatility. Nonetheless, the absence of a direct BI→EP relationship deviated from studies by [12], which found BI to be positively related to environmental performance. In these instances, differences in technological maturity, industrial sector, or in the complementary organizational routines may account for the discrepancies [6,12]. On the contrary, the results found here are in accord with [5], who maintained that BI is exerting a helpful environmental impingement through its alliance with operational enablers like GSCM or SCI [5,28].

This aligns with findings by [11], who showed that BI initiatives in Jordanian SMEs have a stronger environmental impact when integrated into green practices rather than pursued as standalone tools. Their study reinforces the strategic importance of embedding BI in sustainability-oriented processes to yield measurable environmental gains [11].

With the mediation found to be full through GSCM and partial through SCI, an empirical basis is lent to the tripartite framework of the NRBV. Specifically, BI’s contribution to pollution prevention is realized when it is embedded in green supply chain practices that reduce waste and improve process efficiency. Product stewardship is reflected in the enhanced collaborative relationships facilitated by SCI, which extend environmental responsibility across supply chain partners. Sustainable development, the third pillar of NRBV, is advanced when BI enables long-term environmental strategy through a dual pathway of internal process greening and external integration. In this sense, this study contributes to the literature by demonstrating that BI fulfills all three NRBV strategic pathways but only indirectly, mediated by operational and relational capabilities.

The role of GSCM as a primary mediator aligns with Al-Ghwayeen & Abdallah, who found that green supply practices in Jordan significantly boost environmental performance, especially when internal management capabilities are present. This convergence validates the current model’s emphasis on GSCM as a critical link in converting BI-driven insights into sustainability outcomes [24].

Each hypothesis was rigorously examined. The non-significant BI→EP path confirms that BI, on its own, is insufficient to produce environmental gains, which aligns with studies highlighting the need for operational translation of digital insights. The significant positive effects of BI on GSCM and SCI, and the subsequent impact of these constructs on EP, reaffirm the importance of process integration in supply chain sustainability. GSCM, in particular, emerged as a strong mediator, while SCI played a supportive yet meaningful role. The interplay among BI, GSCM, and SCI also reinforces the notion of supply chain ambidexterity, where digital intelligence enhances both agility and alignment across networks. The absence of BCI’s moderating effects suggests that blockchain technologies may not yet be mature or sufficiently diffused across Jordanian manufacturing to condition the BI-sustainability relationship effectively. This null result echoes the findings of Kouhizadeh and others, who observed that blockchain’s benefits often remain latent in underdeveloped digital ecosystems [26].

These findings also resonate with [21], who argued that GSCM practices are instrumental in achieving the Sustainable Development Goals (SDGs), particularly when combined with data-driven coordination and integration. The strength of GSCM and SCI in the current study reinforces this assertion, showing their mediating role in realizing digital sustainability [21].

Likewise, the SCI findings are conceptually reinforced by Soares and others [16], whose multidimensional framework demonstrates how strategic alignment, information sharing, and collaborative planning enhance environmental responsiveness. The current results affirm that SCI, though secondary to GSCM, plays a pivotal part in diffusing sustainability efforts across the supply chain network [16].

7. Conclusions, Implications, and Limitations

7.1. Conclusions

This study examined how BI affects EP in Jordanian manufacturing firms through the mediating roles of GSCM and SCI and the moderating role of BCI. Using PLS-SEM, the findings reveal that BI does not exert a statistically significant direct influence on EP, but its effects are fully mediated by GSCM and partially by SCI. This outcome underscores the importance of embedding BI within structured green practices and coordinated supply chains to generate measurable sustainability outcomes.

Contrary to initial hypotheses, BCI did not moderate any examined pathways, suggesting that blockchain’s influence is highly dependent on digital maturity, ecosystem readiness, and regulatory support conditions not yet fully realized in the Jordanian industrial context. This aligns with existing literature cautioning against overgeneralizing blockchain’s utility across diverse institutional settings.

Theoretically, this study contributes to RBV, NRBV, and DCV by demonstrating that BI functions as an enabling resource only when integrated with operational GSCM and relational SCI capabilities. BI alone does not translate into environmental gains unless leveraged through firm-specific dynamic capabilities and inter-organizational networks. Additionally, institutional factors such as environmental regulation, national digital strategy, and enforcement mechanisms in Jordan should be considered as critical boundary conditions for future models.

7.2. Implications

Practically, the findings advise industrial firms to align BI investments with sustainability-oriented operations, particularly by developing cross-functional green processes and supplier collaboration mechanisms. Technology adoption, in isolation, is insufficient; strategic deployment must accompany cultural change and structural readiness. Policymakers are encouraged to enhance digital infrastructure, clarify blockchain policy, and promote training initiatives that bridge the gap between analytics capabilities and environmental policy compliance. These efforts support broader national and international sustainability objectives, particularly SDG 12 on Responsible Consumption and Production.

Additionally, this study presents a conceptually grounded and empirically validated framework for aligning digital transformation with sustainable supply chain practices. By demonstrating that BI’s value emerges not from standalone analytics but from its integration into green and collaborative processes, it offers theoretical, managerial, and policy insights for firms and governments striving to navigate the green–digital nexus in emerging economies. The findings challenge deterministic views of digital technologies and advocate instead for an ecosystem-aware, capability-driven approach to achieving EP through digital innovation. Finally, firms should prioritize green process redesign and collaborative integration mechanisms that convert BI insights into actionable practices.

7.3. Limitations and Suggestions for Future Research

While the methodology employed in this study adheres to accepted statistical and structural modeling standards, several limitations must be acknowledged: Firstly, firm size data were collected and are included in Section A of the dataset; however, it was not used as a control variable in this study. Future research should explore its potential moderating or control effect in relation to EP or technology adoption. Secondly, the cross-sectional design constrains causal inferences, particularly regarding mediation and moderation dynamics, which are more reliably tested through longitudinal or experimental designs. Thirdly, the non-significant moderating role of BCI may reflect limited contextual maturity of blockchain in the study setting. Future studies may reconceptualize BCI as a mediator or higher-order construct, especially as technological adoption deepens. Moreover, no multigroup analysis or sensitivity testing was performed. Furthermore, future research should assess sector-specific effects or test model invariance across demographic subgroups (e.g., firm size, firm age, and industry type). Finally, although PLS-SEM is robust to non-normal data, assumptions related to homoscedasticity and residual independence were not explicitly tested. Studies employing Covariance-Based SEM (CB-SEM) or time-series designs are encouraged to examine these assumptions for enhanced inferential rigor.