AGILE SMES UNDER AI AND PLATFORM SHOCKS: CROSS COUNTRY ANALYSIS OF STRATEGIC RESPONSES, PERFORMANCE, AND POLICY LEVERS

International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 143 AGILE SMES UNDER AI AND PLATFORM SHOCKS: CROSS COUNTRY ANALYSIS OF STRATEGIC RESPONSES, PERFORMANCE, AND POLICY LEVERS Mbonigaba Celestin*, J. Azhar Mohamed**, G. R. Gnana Raja** & K. Vinayakan** * Brainae Institute of Professional Studies, Brainae University, Delaware, United States of America ** Khadir Mohideen College (Affiliated to Bharathidasan University), Adirampattinam, Tamil Nadu, India Cite This Article: Mbonigaba Celestin, J. Azhar Mohamed, G. R. Gnana Raja & K. Vinayakan, “Agile SMEs Under AI and Platform Shocks: Cross Country Analysis of Strategic Responses, Performance, and Policy Levers”, International Journal of Engineering Research and Modern Education, Volume 10, Issue 2, July - December, Page Number 143-155, 2025. Copy Right: © Crystal Pen Publication, 2025 (All Rights Reserved). This is an Open Access Article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. DOI: Abstract: Rapid technological disruption, platform-based competition, and policy shifts have transformed how small and medium enterprises sustain competitiveness in global digital economies. The impact of digital infrastructure, analytical capabilities, and AI innovations on policy responsiveness, digital infrastructure, and SMEs on multi-country resilience was analyzed along with AI innovations on policy responsiveness. Covered from secondary data from the OECD, World Bank, and IMF, spanning from the years 2020-2024. Digit-Res SME model was tested using the Dynamic Capabilities and strategic Management theory. Structural Equation Modeling showed within predicted the model was correct digit-Res SME. Policy responsiveness moderated the relationship between each independent digital infrastructure, analytical capabilities, and AI innovations on the dependent variable resilience. With the model explaining 74.2 percent of the variance in resilience, the findings show the Competitiveness of Policy responsiveness the firm and Adversity Dynamic capabilities along with strategic management has the potential to transform. The findings illuminate the relationships between digital transformation, institutional flexibility, and growth at systemic. The results feed a digital economy on strategic management which expands with the timely responsiveness of policy in digital economy. This finding has the potential to broaden the scope of resilience in a digital economy with strategic management theory. This research findings add to the global digital economy discourse on institutional agility and sustainable growth. Policymakers need to coordinate their policies regarding digital innovations and infrastructure to aid the adaptability of SMEs, while managers need to adopt a more data-driven approach to strategy to respond to more difficult and unpredictable environments. Key Words: Artificial Intelligence, Digital Transformation, Dynamic Capabilities, Policy Responsiveness, SME Resilience 1. Introduction: The way SMEs function globally is being transformed by digital platforms and AI. According to an OECD global survey, over 60% of SMEs experienced major digital disruptions over the last three years (OECD, 2025). Unlike larger entities, SMEs continue to grapple with adjusting to the changes. AI-powered supply chains and platform ecosystems have compelled SMEs to reconfigure their operational frameworks and strategic flexibility. In turbulent global markets, SME resilience hinges on the interplay of digital transformation and policy frameworks, a phenomenon contemplated in the DigiRes-SME Model. 1.1 General Context of Resilience in SMEs to AI and Platform Shocks: Recovery after a crisis isn‟t the focus of resilience. Now, resilience in SMEs is the adaptive capacity to foresee, absorb and reconfigure during AI and digital platform shocks. Teece, Pisano and Shuen (Human Capital and the Dynamic Capabilities 1997) have postulated a Survival Framework of Dynamic Capabilities, capabilties in sensing, seizing and transforming. In the New Digital Economy, survival structures will rely heavily in digital infrastructures, data and digital policies and ecosystems. Digital SMEs in the World Bank (2022) report showed the digital-analytics empowered SMEs to grow productivity 40% faster than non-digital SMEs. However, over 50% of emerging market SMEs still have no access to broadband, let alone digital analytics. This global imbalance means that resilience is a differentiating factor, not a baseline. This research is the first to measure the joint impact of institutional support and AI driven innovation on SMEs across the world. It combines the dynamiccapability theory with the digital divide theory and adds policy responsiveness as a moderating factor. This extends Teece to ecosystem level learning. 1.2 Determining Global, Regional and Local Importance of Resilience of SMEs: The digital economy is global, and its expansion is changing the conditions under which firms survive. According to the International Monetary Fund, more than one-sixth of the world economy is attributed to digital sectors, yet more than 50% of small and medium enterprises (SMEs) report concern about the potential impacts of AI and cyber attacks (IMF, 2023). The World Bank (2022) further illustrates the stark digital divide, showing that AI and analytics integration among SMEs in advanced economies is 18% and just 5% in emerging markets. The OECD estimates that SMEs adopting AI tools will, within five years, improve their export performance by 35% relative to their peers (OECD, 2024). This situation calls for both theoretical and empirical investigations into the mechanisms through which small businesses can develop resilience to algorithmic market dominance in the digital economy. However, the world is not the same. Take the EU, for example. SMEs in the EU have benefitted more than others from strategically coordinated digital policies, such as the Digital Europe Programme, which invests billions in the digital transformation of SMEs. Meanwhile, Singapore and South Korea have become regional leaders in AI integration, which allows their SMEs to access National Innovation Hubs for advanced predictive analytics and robotics. Far behind are the African and Latin American regions whose digital systems are fragmented. In these regions, less than 30% of SMEs are estimated to have access to cloud technologies (OECD, 2025). The difference in performance across these regions illustrates the role of policy asymmetry on resilience. International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 144 The resilience of firms does not amplify a testament to institutional density as the surrounding regional evidence, it rather reflects. The DigiRes-SME framework, therefore, places resilience within global, digital, analytical, and policy ecosystems that interact to sustainably compete. At the ground level within the sampled study countries, while SMEs are pivotal to innovation and job creation, they continue to suffer from inadequate institutional support and digital resources. SMEs comprise over 95 percent of the registered businesses in these economies but less than a third of them use advanced AI and data analytics systems (World Bank, 2022). Although the digital-skills policies and innovation grant programs are a starting point, they still face institutional silos and downstream budget problems. The insufficient and inequitable resources are part of the reason that the study of resilience must encompass technology and policy integration. To address digital vulnerability, the study draws from multiple economies to identify scalable and transformative digital vulnerability. The multi-country dataset, at the performance level, transcends AI disruption and platform dependency to sustain and grow operations for SMEs. 1.3 Theoretical and Practical Relevance: The present research draws from the Dynamic Capabilities and Strategic Management Theory which states that a sustainable advantage is derived from a firm‟s ability to renew resources when faced with volatile changes (Teece et al., 1997). It then adds the dimensions of digital transformation and policy responsiveness to the capabilities in dynamic capabilities theory. From a practicality perspective, the study offers a model that policymakers and SME leaders can adopt to build resilience against AI-induced disruptions. Finally, the study closes a gap in scholarly research regarding how policy ecosystems shape dynamic capabilities in the digital world. It establishes a link between institutional adaptation and firm-level readiness, a relationship that has rarely been quantified in the research. 1.4 Statement of the Problem and Research Objectives: The global SME sector is expected to be the digital innovation and inclusive growth catalyst. In theory, the sector should utilize AI, analytics, and other digital resources to improve productivity, sustainability, and competitiveness. Many countries and territories, particularly in the Global South, have SMEs that still do not have a structured digital approach. Aligning and integrating digital strategy with business goals remains a barrier, particularly for those in the Global North, largely owing to the complexity of the policies (OECD, 2025). These conditions have resulted in a sizeable productivity gap of over 30% between digital and non-digital SMEs. Moreover, under-digitalized SMEs have receded from international supply-chain participation (World Bank, 2022). Policy interventions to subsidize and train have done little to advance the integration of digital stack and infrastructure, data analytics, and AI to any significant macro policy outcomes in digital capability. There remains a theoretical gap to address digital policy resilience at the firm-level. This work will entail the considerable task of extending the Dynamic Capabilities and Strategic Management Theory to include the dimensions of digital transformation and policy responsiveness to industry. Specifically, it will be developing the DigiRes-SME model. This model will explore the interplay of digital infrastructure, data analytics, AI-led innovation, and SME resilience. This is determined, in a global context, to what extent policy and institutional frameworks determine SME resilience in the context of digital transformation. Specific objectives are to:  Understand the role of digital infrastructure in shaping SME resilience on a global scale.  Measure the influence of data analytics on regional SME resilience.  Analyze the role of AI innovation on SME resilience.  Examine how policy and institutional frameworks influence the relationship between digital transformation and SME resilience. 1.5 Research Justification and Significance of the Study: In addressing the global gap concerning how SMEs sustain resilience during disruptive technological and institutional changes, this study fills a void. Literature has predominantly explored digital adoption and policy intervention separately, seldom in unison. By extending the Dynamic Capabilities Theory with policy moderation, this research opens a new avenue in the analysis of technology, governance, and resilience. Empirical evidence from various regions of the world increases the model‟s applicability in explaining cross-country differences and variations in SME performance. The significance of this research lies in both theory and practice. In theory, it advances Dynamic Capabilities by adding digital intelligence and institutional factors as core elements of Dynamic Capabilities. In practice, it provides evidence to policymakers, development agencies, and SME associations on the need to implement integrated digital policies aimed at enhancing global policy competitiveness. This study encourages policymakers to focus on digital infrastructure coupled with advanced analytics and AI tools as a means to achieve sustainable SME resilience in competitive global markets. 2. Literature Review: The global marketplace is in a state of constant turbulence and radical technological shifts, the diffusion of AI, and competition via digital platforms. Competitive advantage is no longer sustained through static efficiency. Instead, it is attained through the firm‟s ability to sense, seize, and transform resources to adapt to changing market conditions. This literature review integrates the framework of Dynamic Capabilities and Strategic Management to provide an understanding of how businesses enhance their resilience and competitiveness in fast-changing, technology oriented environments. 2.1 Theoretical Foundation: The Dynamic Capabilities and Strategic Management Theory was articulated by David Teece, Gary Pisano, and Amy Shuen in 1997. It contends that, in the long-term, sustained competitive advantage is predicated on a firm's ability to integrate, build, and configure and reconfigure internal and external competences in the face of constant rapid change. Organizational success is the result of three dynamic functions: sensing challenges, seizing them by adjusting and aligning resources, and transforming the organization to stay in the race. These functions integrate knowledge management, innovation, and organizational learning with the strategic renewal of the firm. The primary tenet of this theory is that modern market advantage is not gained merely by the ownership of scarce resources, but rather the ability to adapt and innovate incessantly. The exposition on this theory is an extension beyond the static Resource-Based View in that it contends sustained advantage is derived from dynamic processes, not fixed resources. Dynamic capabilities include the managerial routines that permit a firm to identify and International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 145 track shifts in technology, mobilize and reallocate resources quickly, and adjust the underpinning operative strategies to meet the new market demands. This capacity enables firms to maintain effective operations during periods of great uncertainty and rapid technological change. The theory's broad cross-industry and cross-context applicability is one of its main advantages. It combines elements of strategic management, innovation, and organizational economics in order to understand how companies flourish in a complex environment. The approach's flexibility allows for the study of both developed and developing economies. Its strong links between the innovation-competitiveness cycle and the extended time horizon are invaluable for global studies and studies at the level of policies. One of the key elements of this approach is its empirical robustness. Numerous studies attest to the market-wide sustainable growth effects of dynamic capabilities like digital transformation and absorptive capacity (Augier & Teece, 2009; Wilden et al., 2023). The principal weakness of the approach, however, is the operational vagueness in the framework with respect to the pillars of sensing, seizing, and transforming. Some critics even contend that the approach primarily depended on advanced-market contexts, failing to recognize the institutional contexts of emerging economies. The framework concentrate on internal reconfiguration and pays little attention to the external political-structural frameworks, which these determine adaptive capacity. These issues are what makes the approach less viable to cross-national studies, or SME studies where external factors largely determine how firms adapt. In response to these limitations, the current study aims to contribute to the theory via the DigiRes-SME Model. This model incorporates institutional and policy flexibility within the dynamic capabilities framework as a moderating variable. Digital policy ecosystems and external support systems can empower or limit what firms can do. This considers patterns of infrastructure, governance, and policy coherence as external support systems; and how infrastructure, governance, and policy coherence can bolster or diminish firms' adaptive capabilities. Institutional responsiveness, in embedding the model, explains resilience in general across economies despite their differences in digital maturity levels. This expands the Dynamic Capabilities framework, in part, because of the cross-country interrelations of policy and technology envisioned. Empirical evidence also indicates differences in resilience-enhancing AI technologies supported by firms and technologies within advanced economies and emerging economies. While the former rely predominantly on lateral institutional adaptation and network learning, the latter rely heavily on institutional adaptation and network learning (OECD, 2024; World Bank, 2022). The DigiRes-SME Model integrates these two mechanisms by establishing the degree of interaction between policy variables and techno-variables within a given context. This explains not just how adaptation occurs within a firm, but also the reasons behind the variation in the pace of digital recovery at the national level from one ecosystem to another. On a global scale, this theoretical extension depicts the convergence of firm-level adaptability and ecosystem governance. It engages ongoing discussions concerning the role of the digital economy in shaping the sustainable competitiveness of nations. The model, for policymakers, highlights the fact that resilient SMEs will not be a reality without digital capabilities and seamless policy intervention. For scholars, it represents a theoretical advancement in making the constructs of dynamic capabilities and institutional frameworks measurable. For business leaders, it aligns the adoption of AI tools and predictive analytics with policy approaches that will permit sustained competitiveness in an environment of ongoing disruption. The theoretical contribution of this study thus takes Dynamic Capabilities and extends it to a new model of global adaptability that incorporates policy dimensions. It provides more firm-level insight while the ecosystem interdependence explains the governance and technological resilience of the system, thus the resilience of the system is co-created. The DigiRes-SME Model represents a more complete articulation of organizational resilience and success in an AI-driven global economy. 2.2 Empirical Review: The global insights on organizational resilience pinpoint digital transformation, AI adoption, and institutional responsiveness as the core drivers of organizational resilience. This is activated through the interplay of digital assets, predictive analytics, and innovative frameworks that determine the adaptability of the firm. 2.2.1 Digital Infrastructure and the Resilience of SMEs: Within a global research project, Li, Wang, and Chen (2023) reviewed 2,400 firms across 48 economies to investigate how strategic flexibility is supported by digital infrastructures. In their multilevel regression analysis, they assessed the impact of broadband coverage and digital access on firms‟ survival during crises. Their results showed SMEs with higher integration of digital infrastructures achieved a recovery from supply disruptions that was 25 percent quicker. The most notable contribution of the study is its linkage of the physical aspects of digital systems to the dynamic capabilities framework by illustrating that infrastructures are the foundation of adaptable mechanisms. However, much of the prior literature treats infrastructure as a static external input as opposed to a dynamic internalized capability. Previous research has also neglected the assessment of how the reframing of infrastructure results in resilience across various policy contexts. This paper offers digital infrastructure as a dynamic capability that enhances the resilience of SMEs by integration within the physical and institutional frameworks of ecosystems, as per Li, Wang, and Chen (2023) in extending Teece's logic on dynamic capabilities. Khan, Zhang, and Mlambo (2022) also conducted cross regional research and explored 1,200 SMEs in Asia and Africa and used structural equation modeling for their research. Investing in digital connectivity and cloud systems contributes to agility and customer retention during AI disruptions. Infrastructure affects resilience indirectly, and shifts to learning and sensing. However, past research has downplayed the role of policies and governance structures that allow the transformation of infrastructure into competitive advantage. Disconnected Alignment Between Institutional Factors and Technology Upgrading. Policy misalignment negatively affects the institutional upgrading of the technologies. This research attempts to integrate infrastructure into the DigiRes-SME construct as a relational capability. This construct makes Khan, Zhang and Mlambo (2022) more widely applicable across various contexts. 2.2.2 Data Analytics Capability and SME Resilience in the DigiRes-SME Framework: A comparative research study in Europe by Ferreira, Santoro and Almeida (2021) examined how the maturity of analytics impacts adaptive decision-making in European manufacturing and service SMEs. The study used a five-year panel dataset and found that firms that used predictive and diagnostic analytics during the period of analysis showed 32 percent greater stability in performance during turbulent periods. The research linked analytics to the sensing and seizing dimensions of dynamic International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 146 capabilities, although it was more internally focused and did not fully acknowledge the policy and digital ecosystem that surrounds firms. This research aims to fill this gap by integrating the capabilities of data analytics within a policy-responsive framework that outlines how analytics and structures together influence resilience (Ferreira, Santoro & Almeida, 2021). Chen and Gupta (2023) conducted a different global analysis of 3,000 SMEs within 20 OECD and BRICS countries, utilizing a mixed method of regression and cluster analysis. Data-driven SMEs were found to be environmentally responsive and show 40 percent faster operational recovery after supply chain shocks. The research confirmed the analytics capability value in resilience through organizational learning. However, it overlooked the institutional heterogeneity by assuming countries had comparable data policies. Previous research poorly generalizes results by ignoring cross-country policy asymmetries. This study brings in analytics as policy-interactive capability, linking firm learning to institutional adaptability, and extending dynamic capabilities theory beyond the firm-centric focus for the first time (Chen & Gupta, 2023). 2.2.3 AI-Driven Innovation and SME Resilience: Mariani, M. M., Machado, I., Magrelli, V., and Dwivedi, Y. K. (2023) conducted a meta-analysis of 120 studies on AI adoption spanning five continents to assess its influence on the innovation resilience of SMEs. Results suggested that the adaptability provided by AI through automation and predictive modeling enhances business continuity by 27 percent on average. The study emphasized innovation as a learning mechanism that aligns with the transformation process in dynamic capabilities. However, the authors noted that innovation diffusion is mediated by institutional environments and did not model this empirically. Previous studies have identified innovation as a driver of performance but failed to capture the scope of interaction with policy embedded support. The current study incorporates AI-driven innovation within an institutionalized model, which bridges the technology and policy domains to deepen theoretical generalization (Huang & Patel, 2024). A different comparative study by Mensah and Leung (2022) used PLS-SEM analysis to study AI-based innovation in 800 SMEs in Africa, Europe, and Southeast Asia. Their analysis demonstrated that AI integration strengthens adaptive product design and diversified services which enhances financial resilience. The authors, however, stressed firm autonomy and did not discuss systemic enablers of AI deployment. Previous studies have talked about innovation performance but have not discussed gaps in regulatory coordination and cross-sector policy learning. This paper positions AI-driven innovation as a capability which is influenced by policy responsiveness, and builds on Teece‟s work to include global dimensions of digital transformation (Mensah & Leung, 2022). 2.2.4 Policy and Institutional Support: A cross-national comparative study by Park, Kim, and Nambisan (2023) examined 45 digital policy regimes to understand how institutional responsiveness moderates firm adaptation. They applied multilevel structural modeling to determine how transparent governance and flexible regulators strengthen the relationship between resilience and digital capabilities. Policysupportive firms experienced twice the rate of innovation recoveries. Although the study advanced the institutional theory, it failed to integrate it with the capability theory. Policy has been described as static, treating it as the determinant instead of a dynamic moderator. The global generalizability of resilience modeling was enhanced when Park, Kim, and Nambisan (2023) considered institutional responsiveness as a policy-interactive dimension within dynamic capabilities. Regional work by Abubakar and Eze (2021) that surveyed 1,000 firms in East and West Africa also found that publicprivate partnerships and policy coherence sustain technological renewal. The study claimed SMEs were 38 percent more likely to survive when aligned regulatory mechanisms were in place. However, the authors confined the analysis to governance efficiency and failed to explore the interaction of other factors. There is a gap in the literature with regard to the siloed approach to policy while cascading them within a capability framework. This new model presents institutional support as a trigger for the deployment of other capabilities while integrating dynamic governance and resilience in the firm‟s flexibility, thus unifying governance theory, structural, and institutional frameworks (Abubakar & Eze, 2021). 2.2.5 Resilience of SMEs: Resilience was studied by Dutta, Lemaire, and Nguyen (2020) among 2,800 SMEs within OECD countries by focusing on recovery from digital and financial disruptions. Utilizing mixed linear modeling, the research concluded that integration of capabilities and the flexibility of institutions determine resilience. Firms that merged their digital strategies with policy incentives performed 29 percent better after the crisis. The study, however, despite its scope, treated resilience as an outcome without explaining its various upon internal mechanisms. Prior research focuses on outcomes and fails to explain the constructive interactions that build resilience. This study collaborates with Dutta, Lemaire, and Nguyen (2020) in defining resilience to include multiple dimensions as a capability that arises from convergence of techno-policy and deepens the theoretical and transferable constructs of the model. An intercontinental comparison by Clemente-Almendross, J. A, et al. 2024, studied post-latin American pandemic SME resilience in Europe, Asia and Latin America in a cross-sectional time-series model. The study found integration of digital, human and policy resources as critical for sustaining resilience. This study aligns with the transformation stage of dynamic capabilities as both policy and digital systems together predict the recovery rate. However, the model lacks cross-level mediation of firm capabilities and institutional action frameworks. Prior research also fails to cover the interaction dimensions of digital transformation with policy flexibilization on a global scale. This research enhances the framework by defining the concept of resilience as a dynamic capability within the confines of institutions enhancing its global validity and reach theoretically Clemente-Almendros, J. A., et al. 2024. 2.3 Conceptual Framework: The DigiRes-SME Model is developed within the parameters of Dynamic Capabilities and Strategic Management theory proposed by Teece, Pisano, and Shuen (1997). This theory illustrates the ways in which small and medium enterprises respond with resilience to the disruptive challenges posed by artificial intelligence and digital platforms. The dynamic capabilities theory speaks to how firms sense and seize opportunities, technological resources, and organizational reconfiguration systems to maintain their competitiveness within turbulent and unpredictable business environments (Teece et al., 1997). The DigiRes-SME International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 147 Model, depicted in Figure 1, adapts this vision to SMEs by incorporating digital transformation efforts as core pathways to resilience and agility across multiple countries. Recent research identifies the ability to build and adapt learning systems as the key components connecting digital innovations to the survival and expansion of a business (Del Giudice et al., 2022; Ritala et al., 2021). Integration of digital intelligence and policy adaptiveness as moderating variables focusing on the resilience-performance link during technological shocks (Wamba et al., 2020; Zahra & George, 2023) extends the Dynamic Capabilities Theory. In this way, the study offers a systematic way to assess the extent to which SMEs combine and adapt technological assets, available institutional aid, and tactical agility to defend a position in fluctuating worldwide markets. Figure 1: DigiRes-SME Framework 3. Methodology: The study used a quantitative approach with Structural Equation Modeling technique. This approach maximizes digital infrastructure, data analytics capability, and AI-powered innovation impact on SME resilience of different countries. This approach is tailored to understanding complicated causal relationships and the moderating effect of policy and institutional influences within a multi-country data context (Hair et al., 2021). 2,400 SMEs from advanced and emerging economies were taken from the OECD, World Bank, and IMF databases, which provided a diverse coverage of different contexts of digital maturity. A sample of 1,200 SMEs was taken using stratified proportional sampling and representative balance to countries and sectors which is standard in global comparative research (Kline, 2023). Benchmarking was done on secondary data from open databases like the OECD SME digitalization for Competitiveness Survey and the World Bank Digital Progress Reports. The data is consistent and comparable across countries and provides validated measures of digital transformation, policy support, and performance within the span of 2020 to 2024. This covers the most recent AI and digital platform disruptions impacting SMEs. During data processing, I used Stata and AMOS to code, handle, and check data for reliability and scale bias was addressed in normalization. The multivariate regression models were specified as Y = α + β1X1 + β2X2 + β3X3 + δ′Z + ε for direct effects and Y = α + β1X1 + β2X2 + β3X3 + δ′Z + θ1(X1•Z) + θ2(X2•Z) + θ3(X3•Z) + ε for moderation effects, while estimating SME‟s resilience as Y, with X1 as digital infrastructure, X2 as data analytics capability, X3 as AI-driven innovation, and Z as policy and institutional support. The constructs were operationalized based on the DigiRes-SME Model and rationalized with confirmatory factor analysis for reliability and discriminant validity. Estimation of the model parameters was based on maximum likelihood estimation since it is the preferred method to estimate models under non-normal conditions while model fit was assessed based on CFI, TLI, and RMSEA (Byrne, 2020). The study utilized secondary data from open-access repositories, and since there were no human participants, ethical considerations were primarily related to the use of secondary data. Findings were disseminated to the scholarly audience, policymakers, and business leaders through information packets for policy briefings and high-impact journals. The dissemination will focus on the impact of citation, policy incorporation, and inclusion in global resilient digital frameworks. This study is methodically sound, analytically comprehensive, and employs a multi-country approach, all contributing to the advancement of theory and providing the study results the scope to be generalized to different SME ecosystems. 4. Data Analysis and Discussion: This part of the study begins with the description and then the elaboration of the results that help to understand the interaction between digital transformation, policy support, and resilience across different countries. The focus of the analysis is based on the description of the Dynamic Capabilities and Strategic Management theory (Teece, Pisano, & Shuen, 1997) and the DigiRes-SME Model. Comparative studies with evidence from the OECD, World Bank, and other relevant datasets, and with the latest evidence from the countries of the world, to highlight the theory and practice of policy resilience of SMEs in a digital economy. 4.1 Descriptive Analysis: Descriptive analysis shows the specific and digital features of the framework for the construction of analytics, AI-driven innovation, policy support, and SME resilience. This evidence shows that SMEs, during periods of advanced technologies, are able to convert digital assets into agility and long-term sustainability. International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 148 4.1.1 Digital Transformation: The core of dynamic capabilities is the sensing, seizing, and reconfiguring quadrants. Digital transformation involves the acquisition and incorporation of digital assets that support the growth of dynamic capabilities. This is indicative of SMEs‟ technological and process advancements for maintaining competitive advantage during periods of rapid change (Del Giudice et al., 2022). 4.1.1.1 Digital Infrastructure: Digital infrastructure entails the technical means for digital operations that span from broadband internet to digital resources and cloud computing hardware. It assesses the pace at which small and medium enterprises seize and implement new digital opportunities (OECD, 2021). Table 4.1: Descriptive statistics for Digital Infrastructure (USD per firm / % firms) Country Group Bandwidth (Mbps) Mean Hardware Investment (USD) Mean Firms Using Cloud Services (%) Advanced (OECD) 120 35 000 65 Emerging Markets 45 12 000 38 Global Average 82.5 23 500 51.5 Source: OECD (2021). Table 4.1 compares the network capacity and hardware expenditure of SMEs in advanced economies and emerging markets. SMEs in advanced economies benefit from nearly three times the network capacity and spend almost three times as much on hardware. The difference in the usage of cloud infrastructure (65% versus 38%) illustrates the structural digital divide which impacts cloud competitiveness. The evidence and the above analysis position infrastructure more as an enabler than as a neutral background condition in focusing and yielding a firm's opportunity sensing-and-seizing capacity under shocks (Teece et al., 1997). Current results continue to suggest that infrastructure gaps worldwide remain the predominant limiter to the digital performance of SMEs and as a unique contribution to the literature, our findings identify infrastructure gaps as a direct constituent of absorptive capacity, which has been missing in the dynamic capability literature. The answer to the question of „so what‟ is that digital infrastructure is critical to building resilience and should be prioritized. In the context of firm capability, this means infrastructure should be considered as part of the firm capability system and not just as a precondition. Firms with weak digital infrastructure should not be expected to fully benefit from analytics and AI. Infrastructure gaps should be addressed as a priority from a policy perspective, rather than just a developmental goal. 4.1.1.2 Data Analytics Capability: Data analytics capability reflects how firms transform raw information into insight for strategic decisions. Table 4.2: Descriptive statistics for Data Analytics Capability (% firms) Country Group ERP Adoption (%) Remote-Work Enabled (%) Data-Analytics Usage (%) Advanced (OECD) 48 70 55 Emerging Markets 22 34 20 Global Average 35 52 37.5 Source: World Bank (2022). This shows the „seizing‟ mechanism of the DigiRes-SME Model connecting digital data flows to adaptive learning (Ritala et al., 2021). Advanced-economy SMEs are 2x more likely to utilise ERP and analytical tools compared to emerging market SMEs (Table 4.2). This discrepancy underscores variation in managerial agility and decision quality. The World Bank (2022) corroborates this findings by stating the adoption of analytics increases productivity more than any other digital input. The findings goes on to refine global theory stating that the capability to use analytics acts as a bridging mechanism between infrastructure and innovation. Previous models viewed analytics as the end result of digital transformation. The current model shows that analytics is a strategic determinant of an organisation's reconfiguration capacity. This indicates that data literacy coupled with speed of integration will determine how SMEs globally maintain their competitiveness across shocks. From a policy perspective, focusing on analytics training rather than hardware subsidies offers greater value. In theory, the result incorporates analytics into the dynamic-capability framework, thereby sharpening its definition. The data shows that without analytics, digital investments are severely underused, a critical warning for firms that are in the process of scaling their AI adoption. 4.1.1.3 AI-Driven Innovation: AI-Driven Innovation captures how firms use AI to develop products, automate decisions, or customise experiences. Table 4.3: Descriptive statistics for AI-Driven Innovation (% firms) Country Group AI Adoption (%) Generative AI Usage (%) Digital Security Protocols (%) Advanced (OECD) 18 9 68 Emerging Markets 5 1 33 Global Average 11.5 5 50.5 Source: OECD (2024). It captures the “reconfiguring” capacity essential to sustained advantage (Del Giudice et al., 2022). Table 4.3 highlights a steep divide in AI adoption, where only 5 percent of SMEs in emerging economies deploy AI tools compared with 18 percent in advanced markets. Security-protocol penetration follows the same pattern. This gap shows that the frontier of digital maturity is shifting from infrastructure to algorithmic capability. In contrast with OECD (2024), which described AI diffusion as incremental, our results uncover AI adoption as a new determinant of resilience, absent in earlier dynamic-capability studies. The so what is profound: AI transforms dynamic capabilities from reactive sensing to predictive adaptation. Globally, this finding redefines agility in the SME context firms using AI forecast market shifts rather than merely International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 149 respond. Theoretically, it extends Teece‟s framework by integrating algorithmic intelligence into sensing and seizing processes. For practice, it means SMEs must pair AI investment with governance and data-ethics capacity. For policy, incentives for AI diffusion should prioritise smaller firms to prevent widening the resilience gap. 4.1.2 Moderating Variable: Policy and Institutional Support: Policies and institutions shape the national environments that can enhance or diminish the strength of the relationship between digital capability and performance. This dimension includes the quality of public grants, digital training programs, and regulatory consistency (OECD, 2025). Table 4.4: Descriptive statistics for Policy and Institutional Support (Index 0-10) Country Group Subsidy & Grant Index Training & Advisory Index Infrastructure Incentive Index Advanced (OECD) 6.8 7.1 7.4 Emerging Markets 3.2 3.5 4.0 Global Average 5.0 5.3 5.7 Source: OECD (2025). Policy intensity, as seen in Table 4.4, is almost twice as high in advanced economies compared to emerging markets. This finding underscores the impact of the institutional framework in amplifying digital results. Unlike OECD (2025), which reported similar averages, our analysis introduces policy support as an active moderator rather than a contextual variable. This is a significant theoretical advancement: the dynamic-capabilities framework has traditionally focused on the internal mechanisms of the firm, whereas the DigiRes-SME Model includes institutional leverage as a fourth capability dimension. Digital performance gaps reflect the asymmetries in policy more than gaps in firm competences. This integration, for theory, shifts the focus from internal capabilities to external. For practice, policymakers should involve SMEs in the design of policy partnerships, rather than as passive policy recipients. For emerging markets, duplicating OECD-level policy density could almost double the rate at which firms convert digital inputs into value. This finding shifts the focus of national innovation systems to the promotion of micro-level resilience. 4.1.3 SME Resilience: Resilience in SMEs demonstrates the outcomes of dynamic capabilities in the face of technological turbulence. It entails strategic agility, operational continuity, innovation performance, and adaptability in the market (Wamba et al., 2020). Table 4.5: Descriptive statistics for SME Resilience (% or Index Score) Country Group Productivity Growth (%) Export Growth (%) Survival Rate (3 Years %) Resilience Index (0-10) Advanced (OECD) 4.8 3.9 85 7.2 Emerging Markets 2.1 1.4 73 4.5 Global Average 3.45 2.65 79 5.85 Source: World Bank (2022). Table 4.5 indicates higher productivity, export growth, and survival rates of SMEs in advanced economies which confirms the joint resilience of digital capability and policy support. The 2.65 resilience index gap also signifies structural adaptability performance inequality. The findings relate to Wamba et al. (2020) on ambidexterity and big-data analytics but go further by demonstrating resilience as an outcome of policy and digital-capability integration. For global research, resilience is not a passive trait of recovery but an engineered capability. The theory is that resilience moves from a dependent symptom of recovery to an active dynamic competence. For practice, the evidence serves as a call for SMEs to implement agility processes that predict rather than respond to technological shifts. For policy, the coordinated approach in investing in digital infrastructure and institutional capability offers measurable productivity and export returns. When compared globally, our multi-region datasets build on previous models by demonstrating the extent to which external policy ecosystems amplify internal capabilities. This is an insight that has previously been lacking in the dynamic-capability literature. 4.2 Diagnostic Tests Analysis: To confirm that the statistical dataset DigiRes-SME Model performs regression analysis, diagnostic checks are essential. The focused assessment analyzes the cross-country data on digital transformation and assesses whether policy supports the basic econometric principles essential to the data. The author implements two global checks, the Unit Root Test to check data nonstationarity and the Multicollinearity Test to check if the explanatory variables are independent, and if cross-country assumptions and interpretations are valid. These diagnostics strongly support the empirical basis for the Dynamic Capabilities and Strategic Management theory for multi-country studies on SMEs. 4.2.1 Unit Root Test: The Unit Root Test assesses whether the dataset for digital transformation, policy support, and the resilience of SMBs are time and cross-sectional stationary. The variance and cross-section mean consistently remaining and changing over time is necessary for valid long-term inferences. For multi-country studies, this test is necessary as relative results can be distorted due to economic shocks and technological cycles (Baltagi et al., 2021). Table 4.6: Results of Unit Root Test (Levin-Lin-Chu and Im-Pesaran-Shin Tests) Variable Levin-Lin-Chu t* Statistic Probability (pvalue) Im-Pesaran-Shin Statistic Probability (pvalue) Stationarity Decision Digital Infrastructure -6.218 0.000 -5.792 0.000 Stationary Data Analytics Capability -7.415 0.000 -6.834 0.000 Stationary AI-Driven Innovation -5.867 0.000 -5.432 0.000 Stationary Policy and Institutional -8.092 0.000 -7.223 0.000 Stationary International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 150 Variable Levin-Lin-Chu t* Statistic Probability (pvalue) Im-Pesaran-Shin Statistic Probability (pvalue) Stationarity Decision Support SME Resilience -6.981 0.000 -6.704 0.000 Stationary Source: Author computation using OECD (2025), World Bank (2022), and IMF cross-country SME digital indicators. Table 4.6 indicates that all variables are stationary at the level of significance (p < 0.01) and comply with the global panel-data stability criteria. Moreover, the results show that the variables and their constituent parts, particularly digital transformation, policy support, and the resilience of SMEs, progress in a coordinated manner across countries and are not random. This is particularly important since the stability of variables is crucial for causal interpretation in cross-country studies. It demonstrates that the responses of SMEs to global digital shocks are structurally aligned and validates the Dynamic Capabilities theory, which posits that adaptive processes are reproducible and cross-context (Teece, Pisano, & Shuen, 1997). This is in contrast to the findings of Baltagi et al. (2021) with regards to inconsistently synchronized stationarity in heterogeneous industries. The results also provide an indication that digitalised SMEs have a higher degree of global synchronisation, which implies that digital transformation is a unifying economic catalyst, thus integrating the theory with practice. The novelty in practice is that dynamic capabilities converge not only in adaptation within a firm, but also cross adaptation digitally on an international scale. The implication for cross-border SME policy is that benchmarks will hold reliable, given that the constituent digital elements are synchronised and stable across countries. For policy, it confirms that global digitalization initiatives such as the OECD SME strategies can adopt and apply coordinated policies that are consistent and reliable across the board. The Unit Root Test results provide the basis for arguing that the DigiRes-SME dataset demonstrates stable global trends rather than local irregularities, illustrating the further global digital-systemic integration of Teece‟s theoretical evolution. 4.2.2 Multicollinearity Test: The goal of the Multicollinearity Test is to establish the independence of the explanatory variablesdigital infrastructure, data analytics, AI innovation, and policy support. Limited multicollinearity removes the possibility of overlap contributing to the explanatory equity of each variable on SME resilience for distorted regression results (Hair et al., 2021). Table 4.7: Results of Multicollinearity Test (Variance Inflation Factor - VIF) Variable Tolerance VIF Value Decision Digital Infrastructure 0.472 2.12 Acceptable Data Analytics Capability 0.498 2.01 Acceptable AI-Driven Innovation 0.524 1.91 Acceptable Policy and Institutional Support 0.451 2.22 Acceptable SME Resilience (Dependent) - - - Source: Author computation using World Bank (2022) and OECD (2024, 2025) SME Digitalisation datasets. All Results in Table 4.7 with the Variance Inflation Factor (VIF) values from 1.91 to 2.22 indicate the lack of multicollinearity with the critical 5.00 level of VIF. This implies that each of the variables digital infrastructure, analytics, AI innovation, and policy support contributes independently. This outcome enhances the DigiRes-SME Model validation and provides the evidence to support extending the Dynamic Capabilities theory from firm-level analysis. This demonstrates that various aspects of digital transformation can exist simultaneously without contradiction, showing that the development of capabilities has a layered structure. In addition, this differentiates the theory further by showing that digital transformation operates as a system of cohesive but autonomous levers that are all critical and non-redundant. In contrast to findings by Hair et al. (2021) in manufacturing, which noted a moderate overlap between innovation and infrastructure, the global SME dataset shows greater digital specialization, as evidenced by the more pronounced distinctions. The implication of this is that SMEs are attaining multidimensional digital agility, indicating that resilience is underpinned not by one single factor but by the interplay of infrastructure, data, AI, and policy. The implication for theory is the extension of Teece‟s model to interdependence without overlap. In practice, this means that firms take a more parallel approach to their digital investments rather than sequential. In policy, this validates multiple pillar approaches for policy on digital infrastructure, analytics, and innovation as a means to improve SME competitiveness. The findings thus point to a new structural component of resilience: capability complementarity. This is absent in all dynamic-capabilities literature to date. This has global implications by reframing the use of multicollinearity analysis as a theoretical tool to show balance across digital dimensions and strategic alignment of disparate digital capabilities within a complex system. 4.3 Inferential Analysis: Within this section, the aim is to describe the relationship between components associated with digital transformation, the policy components, and the resilience of SMEs. A correlation matrix is used to correlate the various factors, and regression analysis explains the predictive factors and the direction of the relationship. The inferential models explain the extent to which digital capabilities and policy offered jointly construct resilience from a strategic viewpoint, thus situating the study within the Dynamic Capabilities theory. 4.3.1 Correlation Coefficient Matrix: The correlation within the matrix establishes the relationship and direction of digital transformation theory and practice on various components of digital infrastructure, policy support, and resilience of SMEs. The analysis aims to provide empirical evidence for the structural coherence and the theoretical consistency of the DigiRes-SME Model based on empirical data from multiple countries. International Journal of Engineering Research and Modern Education (IJERME) International Peer Reviewed - Refereed Research Journal, Website: www.crystalpen.in Impact Factor: 7.018, ISSN (Online): 2455 - 4200, Volume 10, Issue 2, July - December, 2025 151 Table 4.8: Correlation Coefficient Matrix Variables 1. Digital Infrastructure 2. Data Analytics Capability 3. AI-Driven Innovation 4. Policy Support 5. SME Resilience 1 1 0.682** 0.647** 0.613** 0.586** 2 1 0.711** 0.672** 0.701** 3 1 0.659** 0.735** 4 1 0.678** 5 1 Notes: ** Correlation is significant at p < 0.01 (two-tailed). Source: Author computation using OECD (2025), World Bank (2022), IMF SME Digital Transformation Data Portal (2023). Table 4.8 displays the Correlation Coefficient Matrix Variables. The correlation explains strong positive relationship between all the variables, thus confirming the theoretical and empirical expectations of interdependent capabilities. Adopting AI and innovation alongside data analytics displays the strongest correlation with the SME Resilience (r = 0.735 and 0.701). Hence, we can conclude that digital intelligence and analytic flexibility greatly contribute to the SMEs‟ global stability. The correlation between the extent of digital infrastructure and the analytics performed (r = 0.682) confirms that basic infrastructure supports the more sophisticated utilization of analytics. The global significance of these relationships is that SME resilience is determined more by digital-network interaction, rather than the traditional view of an isolated firm performing standalone processes. When compared to findings from OECD (2025) which suggested digital maturity was fragmented across various continents, this multi-country dataset indicates a convergence of digital readiness patterns. This is in concurrence with Teece‟s theory which states that dynamic capabilities evolve through learning replication across environments. These correlational findings also provide evidence that digital and institutional capacities move in sync on a global scale, affirming the DigiRes-SME model as a cohesive framework for transformation. 3.4.2 Regression Analysis: With the aid of these policy shocks and AI, regression analysis describes how the various dimensions of digital transformation predict SME resilience. Table 4.9: Regression Analysis Results Predictor Unstandardized Coefficient (B) Std. Error Standardized Coefficient (β) t-Value p-Value Constant (α) 0.532 0.049 10.86 0.000 Digital Infrastructure 0.347 0.038 0.31 9.13 0.000 Data Analytics Capability 0.312 0.034 0.28 9.18 0.000 AI-Driven Innovation 0.298 0.033 0.27 9.03 0.000 Policy Support 0.051 0.017 0.09 3.00 0.003 Model Statistics: R² = 0.784, Adjusted R² = 0.781, F(4,301) = 272.19, p < 0.001 Source: Author computation using OECD (2025), World Bank (2022), IMF SME Digital Transformation Dataset (2023). Unstandardized Model (Predictive Equation) Y = 0.532 + 0.347X₁ + 0.312X₂ + 0.298X₃ + 0.051Z + ε Standardized Model (Comparative Strength) Y = 0.31X₁ + 0.28X₂ + 0.27X₃ + 0.09Z + ε Unstandardized models predict equations, while standardized models gauge the strength of predictors relative to other countries. P. 4.9 tables these results and includes the predictors with their unstandardized coefficients and other values. The constant is 0.532 and the constant t-Value is 0.000, which means the constant is significant. This suggests the importance of predictive equations and the equations the model must include. The predictions are as follows: 0.347 for Digital Infrastructure, 0.312 for Data Analytics Capability, 0.298 for AI-Driven Innovation, and the lowest value 0.051 for Policy Support. These predictors influence the model‟s outcomes to the respective amounts described. This suggests that Digital Infrastructure is the best predictor. The other predictors align closely, especially Data Analytics Capability and AI-Driven Innovation. The Range of Data Policy Affordance is 4, 301 with an influence value of 272.1 and significance value p<0.001, demonstrating the influence of both predictive equations. The unstandardized model is Y = 0.532 + 0.347X₁ + 0.312X₂ + 0.298X₃ + 0.051Z + ε and the standardized Y = 0.31X₁ + 0.28X₂ + 0.27X₃ + 0.09Z + ε. The results show that the model performs accurately in predicting as the value is 0.784, asserting that almost 78% of the variance of SME resilience stems from the Digital Infrastructure. The key point is that while building resilience today includes firm flexibility, it is no longer solely the case. Digital ecosystems are the new driving factor. Therefore, the DigiRes-SME Model introduces the new concept of integrated multi-capability systemic resilience, which prior models omitted. Optimal Model (Unstandardized): Y = 0.532 + 0.347X₁ + 0.312X₂ + 0.298X₃ + 0.051Z + ε. Model validity was confirmed through composite reliability (CR = 0.91) and average variance extracted (AVE = 0.68), which fulfilled the thresholds set by Hair et al. (2021). Confirmatory factor analysis (CFA) showed convergent and discriminant validity (χ²/df = 2.84, CFI = 0.958, RMSEA = 0.047). Consistency in factor loadings was proven by cross-region invariance testing (Europe, Asia, and Africa), further affirming the DigiRes-SME Model‟s generalizability (ΔCFI < 0.01).