International entrepreneurship bridges the fields of international business and entrepreneurship, focusing on new ventures that internationalise from inception (Oviatt & McDougall, 1994). The development of these ventures is closely tied to their home country’s socio-economic environment, which shapes entrepreneurial knowledge and provides essential networks and learning resources (Zahra et al., 2018).

Existing literature has examined the impact of the home country on internationalisation from three main perspectives: industry, network, and institution. From an industry perspective, scholars argue that a firm’s international competitiveness is rooted in the characteristics of its domestic industry (Porter, 1998). Patel et al. (2018) found that a dynamic domestic industry environment increases the risk of failure for born globals, while a munificent environment enhances their competitive position. Other studies show that high concentrations of industrial clusters (Baum et al., 2022) and home-peer entry density (Sui et al., 2019) significantly shape new ventures’ international strategies and survival.

The network perspective emphasises the role of relational resources in gaining an international advantage. Baier-Fuentes et al. (2021) highlight how interactions among triple helix actors within the domestic innovation ecosystem support internationalisation. Similarly, Zhang et al. (2016) demonstrate how domestic and international network ties affect the scope and speed of international market entry.

From the institutional perspective, firms’ behaviours are shaped by the regulatory and normative frameworks of their home countries (Szyliowicz & Galvin, 2010). Empirical evidence shows that institutional uncertainty can encourage early internationalisation (Rialp-Criado et al., 2019), while poor digital infrastructure and institutional voids may prompt firms to seek opportunities abroad (Brieger et al., 2022). Governance quality (Li, 2022), regulatory environments, and corruption (Pindado et al., 2023) have also been identified as key institutional factors, although findings on regulation remain mixed. Yang et al. (2023) further argue that institutional support for entrepreneurship increases the likelihood of internationalisation, especially in emerging markets.

While these three perspectives have established a robust foundation for understanding how domestic environments influence internationalisation, most existing studies focus on linear relationships and isolated mechanisms. In fact, the home-country environment, especially in digital contexts, is characterised by interdependent, dynamic, and nonlinear interactions. Within a DEE, elements such as digital infrastructure, user participation, platforms, and entrepreneurial capacity interact in complex and sometimes competing ways. These configurations vary across contexts, creating distinct conditions that shape internationalisation outcomes for new ventures.

The concept of the entrepreneurial ecosystem has gained increasing scholarly attention. It offers a systemic lens to study how new ventures emerge, grow, and interact with their environments (Autio et al., 2018; Stam, 2015). Entrepreneurial ecosystems are generally defined as interconnected sets of actors, institutions, and resources that collectively foster entrepreneurship within a defined spatial or functional boundary (Stam & Van de Ven, 2021). While entrepreneurial ecosystems emphasise geographic proximity, institutions, and industrial clusters, the DEE shifts attention towards digitally mediated structures, global linkages, and platform-based dynamics reflecting the changing nature of entrepreneurship in the digital era (Hajli et al., 2025; Li et al., 2024; Zahra et al., 2023). According to Du et al. (2018), the DEE comprises regional elements that support the growth and expansion of innovative start-ups leveraging digital technology. We adopt a four-part conceptualisation of DEE consisting of (Song, 2019) digital user citizenship (DUC), digital multisided platforms (DMP), digital infrastructure governance (DIG), and digital technology entrepreneurship (DTE). These components reflect the multiactor, multiscalar, and multifunctional nature of DEEs and their potential influence on IE.

Digital user citizenship refers to the digital literacy, participation, and co-creation behaviours of users within digital ecosystems (Sharma et al., 2022). Users are no longer passive consumers but active contributors to innovation, reputation, and content production, shaping entrepreneurial opportunity spaces through feedback loops and social amplification (Rachmad, 2024). High levels of digital citizenship, especially within transnational user bases, provide legitimacy, early-market access, and demand-side innovation for internationally oriented start-ups (Roshan et al., 2024; Sussan & Acs, 2017). From an ecosystem perspective, DUC represents a demand-side foundation that enables digital ventures to scale across borders through user engagement and network effects.

Digital multisided platforms are intermediary infrastructures that enable transactions and value exchange among multiple stakeholder groups (Etemad, 2023). Platforms such as Amazon, Google Play, Shopify, and Upwork provide standardised interfaces, governance rules, and scalable architectures that facilitate market access, reduce transaction costs, and enhance global visibility for ventures (Evans & Schmalensee, 2016). Importantly, DMPs function as market-enabling and coordination mechanisms. They provide access to global users, partners, and resources, thereby lowering liabilities of foreignness and smallness (Hu et al., 2024). Thus, DMPs primarily capture the platform-based transactional and market access dimension of the DEE.

Digital infrastructure governance includes the rules, policies, standards, and regulatory institutions that manage digital system access, security, and interoperability (Janssen et al., 2009). It encompasses both hard infrastructure (e.g. broadband coverage, cloud computing availability) and soft infrastructure (e.g. data protection regimes, platform regulation, cross-border digital trade policies). Effective DIG ensures digital trust, interoperability, and institutional predictability, enabling firms to scale internationally with lower compliance risks (Lafuente et al., 2024). Conversely, fragmented or restrictive governance increases uncertainty and transaction costs, constraining international expansion. Digital infrastructure governance, therefore, represents the institutional and regulatory backbone of the DEE.

Digital technology entrepreneurship refers to the creation and commercialisation of new ventures based on digital technological innovation, including artificial intelligence (AI), blockchain, Internet of Things (IoT), and other emerging technologies (Zahra et al., 2023). It reflects the supply-side innovative capacity of the ecosystem, encompassing entrepreneurs, start-ups, and supporting institutions such as universities, incubators, and accelerators (Sahut et al., 2021).

Unlike DMPs, which provide market access and transactional infrastructure, DTE captures the generation of novel digital products, services, and business models. It emphasises innovation capability, opportunity creation, and technological advancement, rather than platform-mediated exchange. High levels of DTE indicate strong capabilities in producing globally scalable, technology-driven ventures and adapting to rapidly evolving digital environments.

The interaction between the DEE components and the internationalisation of new ventures is inherently systemic and dynamic. Rather than operating independently, DEE elements co-evolve through continuous interactions among institutional structures, technological infrastructures, market mechanisms, and user communities.

From an institutional co-evolution perspective, entrepreneurial outcomes emerge from the mutual adaptation between formal institutions (e.g. digital infrastructure governance), market mechanisms (e.g. platforms) and entrepreneurial actors (Amitrano & Bifulco, 2024). In this process, digital infrastructure shapes the rules of interaction, platforms enable cross-border exchange, users generate demand-side feedback, and digital entrepreneurs drive innovation. These elements evolve interdependently, reinforcing or constraining one another over time.

This co-evolutionary logic implies that the internationalisation of new ventures cannot be explained by isolated factors. Instead, it results from specific configurations of interdependent conditions. For example, strong digital infrastructure may amplify the effectiveness of platforms, while an active digital user base may compensate for weaker technological entrepreneurship. Conversely, misalignment among these elements may limit international expansion.

Building on this perspective, a configurational approach is particularly appropriate, as it captures the combined and context-dependent effects of multiple DEE components. Rather than assuming linear and symmetric relationships, this approach recognises that different combinations of conditions can lead to similar internationalisation outcomes (equifinality), and that the absence of certain conditions may produce different effects (causal asymmetry).

Accordingly, this study proposes a configurational framework (Figure 1) in which the internationalisation of new ventures is shaped by alternative combinations of DEE components. This framework emphasises that international entrepreneurial success is contingent upon the alignment and interaction of ecosystem elements, rather than the presence of any single factor.

This study adopts the national level as the unit of analysis for two key reasons. Firstly, DEE components are predominantly shaped by national-level institutions, regulations, and policy environments, making the country a coherent and meaningful context for assessing systemic configurations (Roshan et al., 2024; Venâncio et al., 2023). Secondly, in the context of digitalisation and globalisation, entrepreneurial activities increasingly transcend regional boundaries, rendering the national level more appropriate for examining how DEEs support internationalisation.

The sample includes 47 countries with available data on both DEE components and international entrepreneurial outcomes. The outcome variable, IE, is measured using the TEAIMPACT2 indicator from the 2019 Global Entrepreneurship Monitor (GEM) Adult Population Survey. This metric captures the share of early-stage entrepreneurs who introduce new products or processes with an international market orientation. Data for the four DEE components are sourced from the Digital Platform Economy Index developed by Acs et al. (2022), accessible through the Global Entrepreneurship Development Institute (GEDI) database. As GEM and GEDI datasets are compiled using different update cycles and data collection frameworks, perfect temporal alignment is not feasible across all countries. To address this issue, the study adopts the closest available years, using DEE indicators from 2018 and IE outcomes from 2019. This 1-year lag is consistent with prior entrepreneurship research and helps to reflect the temporal sequence from ecosystem conditions to entrepreneurial outcomes, while mitigating concerns of reverse causality.

While the 2018–2019 pairing represents the best feasible compromise to balance sample size and data consistency, this misalignment may introduce minor measurement bias if cross-country variation in the timing of digital ecosystem development and entrepreneurial activity differs. We address this concern by focusing on country-level, relatively stable institutional and structural conditions, which are less sensitive to short-term fluctuations.

Table 1 ranks IE rates across the 47 countries, and Table 2 presents the corresponding DEE component measures.

Before necessity and sufficiency analyses, all variables were calibrated into fuzzy sets with values ranging from 0 to 1. Given the lack of established thresholds for high or low levels of DEE and IE, sample-based quartiles were used for direct calibration: the 75th percentile for full membership, the median for the crossover point and the 25th percentile for full non-membership.

To avoid ambiguity in cases where values equal the crossover point (0.5), a constant (0.001) was subtracted from those values, following prior fsQCA literature (Anton et al., 2022; Mattke et al., 2021). Table 3 provides descriptive statistics and calibration thresholds for all variables.

This article followed all ethical standards for research without direct contact with human or animal subjects.

Necessity analysis was conducted to examine whether any single component of DEE consistently leads to high-level IE. Following the threshold proposed by Schneider and Wagemann (2010), a condition is considered necessary only if its consistency exceeds 0.90 (Ragin, 2008). As reported in Table 4, all consistency scores fall below this threshold, indicating that no individual DEE element is necessary to achieve high-level IE.

To identify sufficient configurations of DEE elements that lead to the outcome, fsQCA was employed using the truth table algorithm (Ragin, 2008). The analysis applied standard thresholds: a minimum consistency of 0.80, a PRI (Proportional Reduction in Inconsistency) threshold of 0.70, and a minimum case frequency of 1 (Du et al., 2024). Given the limited theoretical consensus on how individual DEE conditions influence IE, both the presence and absence of each condition were included.

Core and peripheral conditions within each configuration were determined by comparing intermediate and parsimonious solutions. Conditions appearing in both are identified as core, while those only found in the intermediate solution are regarded as peripheral (Du et al., 2024; Ma et al., 2024).

As presented in Table 5, three configurations (S1, S2 and S3) are sufficient for high-level IE, with an overall solution consistency of 0.792 and a coverage of 0.734. In contrast, one configuration (NS) is sufficient for the absence of IE, with a consistency of 0.84 and a coverage of 0.71. These findings suggest that different configurations of DEE elements can lead to high-level IE. A detailed interpretation of each configuration is provided in the subsequent section.

(1) DIG-driven competition enhances IE. Configuration S1 reveals that the presence of DIG as a core condition, supported by the presence of DMP and DTE as peripheral conditions, contributes to elevated levels of IE. In this configuration, robust DIG and DMP foster innovation beyond traditional boundaries, while DTE intensifies market competition, encouraging firms to expand internationally. Countries exemplifying this configuration include the United States, Netherlands, United Kingdom, Sweden, Canada, Switzerland, Norway, Ireland, Luxembourg, Australia, Germany, South Korea, Spain, Portugal, Israel, Italy, Japan, Cyprus, Slovenia, Latvia and Chile. For instance, the U.S. government’s 2023 National Cybersecurity Strategy reflects its high-level DIG, and major platform companies like Google, Apple, and Amazon illustrate its leadership in DMPs. In addition, the U.S. hosts 669 digital unicorns, and countries such as the U.K., Canada, Germany, Israel, and South Korea each support over 15.

(2) DIG and DUC reinforce each other in supporting entrepreneurship. Configuration S2 shows that the presence of DIG and DUC serves as core conditions, while the absence of DMPs and DTE serves as peripheral conditions. In countries where DMPs and DTE are underdeveloped, governments rely on strong digital infrastructure and an engaged digital citizenry to foster IE. A robust digital user base enhances trust and market recognition, especially when it is aligned with effective digital governance. Slovakia and Poland illustrate this configuration. Although their DMP and DTE levels are relatively low, both countries demonstrate well-developed digital legislation and moderate-to-high levels of DIG and DUC, enabling supportive conditions for international entrepreneurial activities. Despite Slovakia’s limited digital financial and technology absorption capabilities, its legislation on information technology usage has been well-developed with effective protection of user rights.

(3) DUC-driven opportunity fuels international expansion. Configuration S3 identifies the presence of DUC as the core condition, the presence of DMPs as a peripheral condition, and the absence of DIG and DTE as peripheral conditions. In such settings, governments primarily leverage an engaged user base and digital platforms to promote IE. A large and digitally literate population provides local validation and market feedback, forming a foundation for global expansion. Meanwhile, efficient digital platforms reduce transaction costs and facilitate cross-border collaboration. Qatar exemplifies this configuration, where, despite limitations in digital freedom, strong digital literacy and regulatory frameworks support entrepreneurial efforts.

(4) Configuration NS captures conditions associated with low IE. This configuration is characterised by the simultaneous absence of DUC and DTE (core conditions), along with the absence of DIG and DMP (peripheral conditions). A lack of digital users diminishes market demand and engagement, while weak DTE suppresses innovation and startup activity. Furthermore, underdeveloped DIG increases uncertainty, and inadequate DMPs hinder access to international customers and partners. Together, these conditions indicate an unsupportive digital environment for the internationalisation of new ventures.

(5) Substitution between DIG and DMPs. The three configurations that constitute high-level IE reveal a certain degree of interrelation among DEE elements. In S2 and S3, under the presence of DUC, DIG and DMPs are substitutable and functionally equivalent for new ventures entering overseas markets. This relationship can be more clearly illustrated in Figure 2. When DIG is well-established, new ventures can directly leverage an open and mutually recognised institutional environment to complete cross-border activities without relying on specific multisided platforms; conversely, if infrastructure governance is weak, new ventures will use integrated multisided platforms to avoid institutional obstacles. In the context of high-level DUC, users’ adaptability is robust (Hajli et al., 2025), further reducing the need for simultaneous high-intensity investment in both, thus forming a substitution effect.

To assess the robustness of the fsQCA results, two sensitivity analyses were conducted. Firstly, the consistency threshold was raised from 0.80 to 0.85. The resulting configurations remained consistent with the original findings. Secondly, calibration thresholds for full membership and full non-membership were adjusted to the 90th and 10th percentiles, respectively. Although this led to minor changes in three configurations, the overall configurational structure remained stable. The robustness results are summarised in Table 6.

This study makes several contributions to the literature on IE and DEE. Firstly, by integrating the DEE perspective into a configurational analysis, it extends existing theories on international new venture development. Prior studies have typically emphasised linear relationships between home-country industrial, institutional, or network conditions and internationalisation. This study reveals that multiple distinct configurations of DEE elements can lead to similarly high levels of internationalisation, thereby confirming the principle of equifinality and answering calls for more holistic, system-level explanations (e.g. Roshan et al., 2024; Venâncio et al., 2023).

Secondly, the findings enrich DEE research by illustrating the interdependence and complementarity of ecosystem components. For example, configuration S1 demonstrates that when DIG is combined with strong DMPs and active DTE, it fosters a competition-driven environment conducive to international growth. This supports a systemic logic rather than treating ecosystem elements in isolation, aligning with theories emphasising institutional co-evolution in entrepreneurship (Amitrano & Bifulco, 2024). Meanwhile, the study challenges the assumption that all ecosystem components are equally important (Stam, 2015; Stam & Van de Ven, 2021). While all components matter, their roles vary in importance. The evidence aligns with Leendertse et al. (2022), showing that some DEE elements are core (e.g. DIG, DUC), some supportive (e.g. DMP), and others contextually useful (e.g. DTE).

Thirdly, this study broadens the performance evaluation framework for DEEs by incorporating internationalisation as a meaningful outcome. Moving beyond traditional metrics such as venture count or investment levels, it responds to growing calls for more comprehensive indicators of entrepreneurial ecosystem performance (Torres & Godinho, 2022; Venâncio et al., 2023). Internationalisation serves as a proxy for ecosystem maturity and global competitiveness, offering a novel lens for both academic inquiry and policy assessment.

The findings offer several actionable insights for policymakers and ecosystem builders.

For governments aiming to foster internationally oriented start-ups, context-sensitive strategies are essential. A one-size-fits-all model is unlikely to be effective given the varying strengths and weaknesses of national DEEs. Recognising the substitution effects among DEE components enables flexible strategies tailored to national circumstances. For instance, countries with limited DTE or DMP capacity (e.g. Slovakia, Qatar) can still facilitate IE by strengthening DIG and DUC through regulatory reforms, digital skills education, and citizen engagement programmes. Importantly, even when focusing on a specific alternative strategy, it is essential to ensure that all configurations surpass a minimum threshold of effectiveness to maintain overall balance and effectiveness within the ecosystem.

Given the central role of DIG and DUC in fostering IE, policy efforts should prioritise these two dimensions. In particular, the role of DUC in configurations S2 and S3 also highlights the demand-side importance of digitally engaged users. In environments with weak technological or entrepreneurial infrastructure, a digitally literate population can enhance legitimacy, traction, and user-driven innovation, partially compensating for other ecosystem deficiencies.

While this study contributes novel insights, several limitations should be acknowledged. Firstly, like many QCA-based studies, it relies on cross-sectional data, limiting the ability to capture dynamic ecosystem changes or causal evolution over time. Future research should incorporate longitudinal datasets and dynamic QCA techniques to explore how DEE configurations evolve and adapt to technological and geopolitical shifts.

Secondly, the national-level focus, although methodologically justified, may obscure important within-country variations. Future studies could conduct multilevel analyses to compare national ecosystems with regional or city-level ecosystems, offering a more granular view of DEE functionality.

Thirdly, future research may explore how specific combinations of DEE elements interact with firm-level capabilities such as entrepreneurial orientation, resource availability, or international experience to shape internationalisation pathways. This could deepen understanding of the micro–macro-linkages in global entrepreneurial ecosystems.