This study employed a cross-sectional survey design targeting professionals within the Vietnamese digital sector. This context is particularly salient as it represents a rapidly digitising emerging market characterised by institutional voids and talent shortages – challenges that directly mirror the current digital landscape in South Africa (Mhlungu et al., 2019; Urban & Wood, 2017).

Data were collected using a purposive sampling technique to ensure that respondents had sufficient experience with digital transformation processes. To guarantee data validity, participants were selected based on three strict inclusion criteria: (1) currently working full time in technical specialist or managerial roles within enterprises operating in the digital or technology sector, thereby ensuring direct exposure to organisational changes; (2) having a minimum of 1 year of continuous involvement in active digital transformation initiatives (e.g. system migration, digital product development or agile workflow adaptation) and (3) possessing sufficient professional experience to reflect on workplace dynamics, consistent with the sample distribution where over 76% of respondents held more than 5 years of experience. Preliminary screening questions based on these criteria were administered at the beginning of the survey, and only eligible respondents were permitted to complete the questionnaire. A total of 595 valid responses were obtained. As shown in Table 1, the sample includes a wide range of roles (including technical specialists and managers) and professional experience, providing a robust and heterogeneous basis for configurational analysis.

All constructs in this study were operationalised using 5-point Likert scales, ranging from 1 (‘Strongly Disagree’) to 5 (‘Strongly Agree’), with instruments adapted from established literature to ensure content validity. Specifically, PsyCap was measured using eight items adapted from Luthans et al. (2007) and Luthans and Youssef-Morgan (2017), covering self-efficacy, hope, optimism and resilience. Digital readiness (DR) was assessed with five items from Parasuraman (2000) to evaluate employees’ adaptability to technology. Both job autonomy (AU) and skill variety (SV) utilised three-item scales from Morgeson and Humphrey (2006) to measure work-method independence and the diversity of required skills, respectively. Work engagement (WE) consisted of nine items adapted from Schaufeli et al. (2006), focusing on vigour, dedication and absorption. Perceived organisational support was measured via six items from Eisenberger et al. (1986) to capture employees’ beliefs regarding organisational care. Finally, IWB was operationalised with six items from Scott and Bruce (1994) and De Jong and Den Hartog (2010), encompassing idea generation, promotion and implementation.

To ensure linguistic and conceptual equivalence, a rigorous back-translation process was conducted by two independent bilingual experts. Furthermore, the instrument underwent a content validity review with academic experts and senior managers in the Vietnamese technology sector to ensure the items were contextually appropriate for digital enterprises. During this review, minor adjustments were made to refine professional business terminology and align with the standard corporate vocabulary used in Vietnamese digital firms. For instance, in the POS scale, terms such as ‘strongly cares about my well-being’ were carefully nuanced in Vietnamese (e.g., using ‘quan tâm đến phúc lợi và đời sống’ rather than a literal translation of ‘well-being’) to accurately capture the perceived paternalistic support typical in Vietnamese corporate culture. Similarly, certain abstract items in the PsyCap scale were rephrased to ensure technical specialists and line employees could easily comprehend them without semantic ambiguity, thereby maximising linguistic validity and reducing cognitive fatigue during response selection.

Given the self-reported nature of the data, both procedural and statistical controls were implemented to mitigate potential common method bias (CMB). Procedural measures include ensuring strict anonymity for respondents and informing them that there were no right or wrong answers.

Statistically, two tests were performed. Firstly, Harman’s single-factor test indicated that the largest factor accounted for only 32.4% of the total variance, well below the 50% threshold. Secondly, a full collinearity assessment was conducted. All inner variance inflation factor (VIF) values were below the 3.3 threshold, ranging from 1.42 to 2.15 (Kock, 2015). These results collectively confirm that CMB does not threaten the integrity of the findings.

This study foregrounds fsQCA to capture the non-linear drivers of innovation. The analysis followed a two-step protocol.

To establish a foundation for the study, a partial least squares structural equation modelling (PLS-SEM) analysis was initially conducted to examine the baseline relationships between the studied constructs. As shown in Table 2, all constructs demonstrate satisfactory reliability and convergent validity, with Cronbach’s alpha and composite reliability (CR) exceeding 0.80, and average variance extracted (AVE) above the 0.50 threshold.

As shown in Table 2, the measurement model meets all standard requirements for reliability and validity, with Cronbach’s alpha and CR values exceeding 0.80 and AVE values above 0.50. In the structural model, all hypothesised direct associations are statistically significant (p < 0.001). The reported effect sizes (f²) range from 0.044 to 0.204. Specifically, while several predictors exhibit small effect sizes, WE (f² = 0.204) and POS (f² = 0.122) demonstrate moderate substantive explanatory power. These varying effect sizes, coupled with the fact that the linear model explains 42% of the total variance in IWB (R² = 0.42), indicate that no single resource acts as a dominant ‘silver bullet’ for fostering innovation. This suggests that traditional symmetrical approaches may oversimplify the complex, interdependent nature of innovation drivers. Consequently, these empirical results provide a strong mandate for the subsequent configurational analysis. By moving beyond net-effect estimations, the following fsQCA explores how these resources combine in asymmetrical and equifinal patterns to achieve high IWB, particularly in contexts where certain individual resources may be scarce.

Prior to identifying the causal configurations, a necessity analysis was performed to determine whether any single resource is a mandatory prerequisite for high IWB. Following the established standards in configurational research, a consistency threshold of 0.90 was utilised to identify necessary conditions (Ragin, 2008). The selection of this conservative, higher threshold, rather than the more lenient 0.80 often used for exploratory sufficiency testing, was deliberately intended to reinforce the empirical rigour of our findings. In qualitative comparative analysis, a condition must demonstrate an exceptionally high and non-trivial consistency score to be meaningfully interpreted as a strict prerequisite. Adopting the 0.90 baseline protects the model from Type I error (false positives), ensuring that any claims regarding resource necessity have robust explanatory certainty within the volatile environment of digital transformation.

The results presented in Table 3 indicate that none of the examined conditions, whether present or absent, meets the 0.90 threshold. This confirms that no single factor is a necessary condition for the occurrence of high IWB in digital enterprises.

Notably, job autonomy (0.892), POS (0.868) and skill variety (0.852) demonstrate high consistency scores that approach the threshold, suggesting that while these organisational resources are highly relevant, they are not indispensable. This finding is theoretically significant as it reinforces the configurational logic of this study: innovation does not depend on universal prerequisites but is instead generated through specific, interdependent combinations of resources.

The empirical evidence from this necessity analysis provides strong support for Proposition 5 (non-necessity). By demonstrating that even critical resources such as digital readiness and organisational support are not individually sufficient in isolation for innovation, the study shifts the focus towards how these conditions are strategically combined to drive outcomes in resource-constrained environments.

The sufficiency analysis identifies multiple configurations associated with high levels of IWB, as detailed in Table 4. The overall solution consistency (0.919) and coverage (0.871) exceed the recommended thresholds, indicating that these configurations are both reliable and empirically relevant.

The results provide robust empirical evidence for the principle of equifinality, confirming that within the complex ecosystem of Vietnamese digital enterprises, there is no single path to fostering innovation. Instead, innovative behaviour emerges from two overarching patterns: synergistic resource alignment and compensatory resource substitution.

The synergistic pattern is exemplified by Solution 1 (Digital champion) and Solution 3 (Supported innovator), in which internal personal capabilities are reinforced by external organisational resources. Specifically, Solution 1 characterises the profile of ‘digital elites’, in which high levels of PsyCap, digital readiness, WE and job autonomy converge. In this pathway, individuals possess a surplus of psychological and technical resources that lowers their perceived risk threshold, allowing them to initiate innovation even without explicit organisational backing. Similarly, Solution 3 identifies the ‘supported innovator’, in which high PsyCap and job autonomy are activated by strong POS. Notably, in this configuration, high digital readiness is not a prerequisite, as POS functions as a critical cultural catalyst that legitimises risk-taking and creates a safe environment for experimentation.

In contrast to these resource-rich scenarios, Solution 2 (Institutional scaffolding) represents a compensatory pathway that is particularly significant for resource-constrained emerging economies. This configuration reveals that high levels of IWB can still be achieved even in the absence of high PsyCap, digital readiness and WE. In this ‘scaffolding’ mechanism, the robust presence of job autonomy, skill variety and POS functionally substitutes for individual-level deficits. By establishing a robust institutional support mechanism, the organisation enables the ‘average’ workforce to contribute to innovation without requiring elite internal psychological reservoirs or extraordinary drive. This finding confirms that strategic organisational design can effectively ‘crowd in’ innovation even when human capital is temporarily depleted or lacks specific technical skills.

Beyond equifinality, the configurational findings provide compelling evidence for the principle of causal asymmetry, indicating that the drivers of innovation are not merely the mirror image of the factors that inhibit it. While traditional linear models often assume a symmetric relationship in which the absence of a driver implies the absence of the outcome, this study demonstrates that high IWB is achieved through specific, complex recipes of resource alignment or substitution.

The presence of both synergistic and compensatory pathways suggests that the decline of innovation is likely the result of distinct patterns of resource depletion or misalignment, rather than a simple lack of individual factors. This reinforces Proposition 4, confirming that the pathways leading to high IWB are fundamentally asymmetric to those leading to its absence. For management, this implies that preventing the decline of innovation requires a different set of interventions than those used to spark it, necessitating a nuanced understanding of how resources are combined within the specific organisational context.

This study makes several important contributions to the literature on IWB by advancing a configurational perspective that moves beyond the dominant net-effect paradigm. Firstly, it contributes to the ongoing debate on the drivers of innovation by demonstrating that IWB is not the result of isolated, additive effects of individual or organisational factors but rather emerges from multiple, interdependent configurations of conditions. By providing empirical support for equifinality, the study challenges the prevailing assumption that innovation is primarily driven by the linear accumulation of resources (Anderson et al., 2014; Parker et al., 2017), and instead aligns with complexity theory in emphasising that different combinations of antecedents can produce equivalent outcomes (Fiss, 2011; Misangyi et al., 2017).

Secondly, the study extends the configurational literature by offering a theoretically grounded set of propositions – equifinality, synergistic resource configuration, resource substitution, causal asymmetry and non-necessity – and empirically validating them within the context of digital enterprises. In doing so, it contributes to a more nuanced understanding of how personal and organisational resources interact, rather than operate independently, in shaping innovative behaviour. This addresses recent calls for greater theoretical alignment between configurational methods and the development of configurational hypotheses or propositions (Pappas & Woodside, 2021).

Thirdly, identifying the institutional scaffolding configuration provides a novel theoretical insight by demonstrating a clear mechanism for resource substitution. While prior research has largely focused on high-performing individuals as the primary drivers of innovation, this study shows that organisational resources can compensate for the absence of strong individual capabilities. This finding extends COR theory by illustrating that resource investment can be externally structured and institutionally supported, rather than solely internally accumulated (Hobfoll et al., 2018). In parallel, the evidence of synergistic configurations refines the JD–R framework by showing that the value of resources is contingent upon their co-presence within specific configurations (Bakker & Demerouti, 2017).

Fourthly, by demonstrating the absence of universally necessary conditions and the presence of causal asymmetry, the study reinforces the core tenets of configurational theory. The findings show that the pathways leading to high and low IWB are fundamentally distinct, thereby challenging symmetrical assumptions embedded in conventional variance-based models (Woodside, 2014). This contributes to a growing body of research advocating for asymmetric and complexity-informed approaches to organisational phenomena (Fiss, 2011; Misangyi et al., 2017).

Fifthly, the study contributes to the contextualisation of innovation theory by situating configurational mechanisms within resource-constrained and institutionally complex environments. By incorporating evidence from Vietnam and articulating implications for similar emerging economies such as South Africa, the study responds to calls for more context-sensitive theorising in management research. It demonstrates that configurational approaches are particularly well suited to capturing the realities of environments characterised by resource scarcity, institutional voids and high interdependence, thereby extending the external validity of configurational theory in non-Western and emerging-market contexts (Khanna & Palepu, 2010; Urban & Wood, 2017).

The shift from a ‘war for talent’ towards an architectural approach to work design represents a fundamental pivot for executives navigating digital transformation in emerging markets. Rather than prioritising the acquisition of scarce ‘superstar’ employees with exceptional psychological resilience and digital readiness, firms should focus on the strategic configuration of organisational systems to enable innovation across the broader workforce. This perspective necessitates a departure from managing through isolated key performance indicators (KPIs) or fragmented HR interventions, which often yield limited impact because they ignore the interdependence of innovation drivers. Instead, managers must adopt a systems-oriented diagnostic lens, recognising that innovation is not a byproduct of any single ‘silver bullet’ factor but emerges from specific, mutually reinforcing ‘causal recipes’ of job autonomy, task variety and organisational support.

Central to this strategic blueprint is the operationalisation of ‘Institutional Scaffolding’ as a viable substitute for individual-level resource deficits. For organisations operating amid persistent talent shortages, the deliberate combination of high job autonomy and skill variety – reinforced by a visible climate of organisational support – creates an external reservoir that compensates for modest PsyCap or digital readiness. By establishing ‘safe-to-fail’ zones and institutionalising support through consistent recognition and managerial backing, firms can lower the perceived psychological cost of experimentation. This approach ensures that the innovation engine remains resilient even when human capital is temporarily depleted, shifting the managerial focus from merely accumulating personal resources to actively providing the structural scaffolds that activate them.

To translate these insights into practice, managers should adopt a contingency-based decision-making logic tailored to their teams’ specific resource profiles. In scenarios where employees already possess high psychological and digital capabilities – the ‘Digital Champions’ – management should prioritise the removal of hierarchical bottlenecks by maximising decision latitude. Conversely, for employees with high motivation but emerging competencies, the ‘Supported Innovator’ pathway suggests that intensifying organisational support can successfully trigger innovative behaviour. In contexts where both individual agency and engagement are constrained, activating the ‘Institutional Scaffolding’ pathway becomes critical, allowing structural job design and paternalistic support to carry the weight of the innovation process.

Finally, these findings provide a transformative blueprint for South African enterprises and public-sector organisations grappling with systemic skill gaps and labour market segmentation. By decentralising decision-making and redesigning roles to encompass diverse problem-solving tasks, South African managers can democratise the innovation process and broaden participation beyond a narrow pool of elite talent. At the policy level, fostering an environment that rewards experimentation and institutionalises support mechanisms can expand the national base of innovative contributors. Such scalable and adaptive interventions offer a sustainable pathway towards inclusive innovation, ensuring that digital transformation efforts are not paralysed by resource constraints but are instead supported by robust organisational architectures.