Innovative work practices and disruptive innovation: the roles of dynamic capabilities, innovation culture and transformational leadership

The 21st century has been widely recognized as an era of rapid innovation and transformation, where technological progress and new business paradigms continuously reshape industries and economies (Caparrós Ruiz, 2021). Disruptive innovation (DI) has emerged as a strategic mechanism that enables firms to challenge established market structures and create new sources of value (Kuhlmann et al., 2023; Pang and Wang, 2023). Through DI, products and services become more accessible, affordable and scalable across diverse sectors (Agyei-Boapeah et al., 2022; Lai et al., 2023). Examples such as Uber, Airbnb, Zoom and FinTech platforms illustrate how DI reshapes consumer experiences and business operations, while enabling sustainable performance improvements (Distelmans and Scheerlinck, 2024; Low and Wong, 2021; Robertson et al., 2022). Hence, firms capable of integrating DI strategies can convert uncertainty into opportunity and achieve competitive advantage in turbulent markets (Forrest et al., 2022).

Post-pandemic market conditions have intensified the need for innovative managerial approaches. The nature of work and employee roles has evolved dramatically, requiring firms to cultivate new organizational practices that enhance creativity and resilience (Zhang and Chen, 2023). Innovative work practices (IWP) have therefore gained renewed importance as mechanisms that enable knowledge sharing, collaboration and experimentation. By promoting open communication, cross-functional interaction and recognition of creativity, IWP empower employees to engage in continuous improvement and innovation (Cardon et al., 2019; Amabile and Pratt, 2016). Such practices help organizations transform employee ideas into viable innovations that can evolve into disruptive outcomes.

Pakistan’s economy, despite structural challenges, has shown notable resilience and adaptability in innovation-related performance (World Intellectual Property Organization (WIPO), 2023). Within its financial sector, dynamic capabilities (DC) and innovation culture (IC) have emerged as pivotal forces that convert managerial initiatives into meaningful innovation outcomes. DC allow firms to sense, seize and reconfigure resources to respond to environmental shifts and technological turbulence (Li et al., 2022; Zahra et al., 2022). IC, on the other hand, nurtures a collective mindset that values experimentation and collaboration (Rubin and Abramson, 2018). Together, these two factors determine whether a firm’s IWP will lead to incremental adaptation or genuine DI.

Nevertheless, empirical studies rarely examine how IWP translate into DI through the joint mechanisms of DC and IC, particularly within service-sector contexts such as banking. Existing research often isolates these constructs rather than considering their interdependence. This study, therefore, addresses a critical gap by developing and testing an integrated moderated-mediation model, grounded in contingency theory, to explain how internal organizational factors shape DI-driven growth.

Contingency theory posits that there is no single best way to organize or innovate; instead, the effectiveness of strategies depends on their fit with contextual conditions such as technology, structure and environment (Talukder and Wang, 2023). By applying this theoretical lens, the current research explores how IWP contributes to DI under specific organizational contingencies in Pakistan’s banking sector. This perspective is especially relevant because banks face heightened technological disruption and regulatory constraints that demand context-specific innovation approaches (Aggarwal and Patel, 2023). Adopting contingency theory allows this study to explain how leadership, capability development and cultural factors jointly moderate the success of innovative work initiatives.

Transformational leadership (TL) is introduced in this model as a boundary condition that strengthens or weakens the influence of IWP on DI. TL, characterized by inspiration, vision and individualized consideration, creates a psychological climate that stimulates creativity and organizational learning (Shafi et al., 2020; Lim and Moon, 2022). Leaders who exhibit transformational qualities provide intellectual stimulation and emotional support, thereby amplifying the positive effects of IWP on both DC building and innovation performance. However, this moderating role of TL has remained underresearched within the DI literature, particularly in emerging-market financial institutions.

Although prior research has examined IWP, DC, IC and TL largely in isolation, considerably less is known about how these internal organizational elements jointly interact to convert employee-level innovation initiatives into disruptive outcomes, particularly in regulated service contexts. For example, recent research has examined digital DI as an external turbulence condition influencing firm performance through managerial competence and IWP (Al Farooque et al., 2026). However, such studies primarily focus on performance consequences rather than on how internal organizational mechanisms jointly shape DI itself. Recent studies increasingly call for process-based explanations that clarify the mechanisms and boundary conditions through which human-centric practices translate into DI rather than assuming uniform effects across organizations (Amankwah-Amoah et al., 2024; Volberda et al., 2021). Responding to this gap, the present study adopts a contingency-theoretic perspective to explain how IWP are transformed into DI through distinct structural, cultural and leadership contingencies within Pakistan’s banking sector.

Drawing on the above arguments, this study proposes and empirically tests five hypotheses linking IWP, DC, IC, TL and DI. Specifically, it investigates:

• the direct effect of IWP on DI;

• the mediating roles of DC and IC; and

• the moderating effects of TL on both IWP–DI and IWP–DC relationships.

By focusing on Pakistan’s banking sector, the research contributes novel empirical evidence to the ongoing dialogue about how human-centric management practices drive disruptive growth in service organizations. It extends contingency theory by demonstrating that managerial innovation outcomes depend on the contextual alignment between leadership style, capability development and cultural support.

The remainder of this paper is organized as follows. Section 2 reviews the theoretical background and develops hypotheses. Section 3 outlines the data collection and analytical procedures. Section 4 presents empirical findings, Section 5 discusses theoretical and practical implications and Section 6 concludes with limitations and directions for future research.

This study is anchored in contingency theory, which explains organizational outcomes as contingent on the alignment between managerial practices and contextual conditions rather than on universally effective strategies. Accordingly, innovation practices do not generate uniform outcomes; their effectiveness depends on how they interact with internal organizational configurations and environmental demands (Osborn and Marion, 2009; Talukder and Wang, 2023).

This perspective is particularly relevant for DI in service-based and highly regulated sectors such as banking. Recent research emphasizes that disruption in financial services rarely emerges from technology adoption alone, but from organizational processes that integrate employee practices, adaptive capabilities and leadership responses under uncertainty (Aggarwal and Patel, 2023; Amankwah-Amoah et al., 2024). As a result, firms may invest heavily in innovation initiatives yet fail to achieve disruptive outcomes due to misalignment among internal contingencies (Pemer and Werr, 2023).

Conventional DI research has predominantly focused on market displacement and technological trajectories, often treating organizational factors as secondary or supportive (Christensen et al., 2018; Kuhlmann et al., 2023). However, contemporary scholarship increasingly recognizes that internal organizational contingencies are central in shaping whether innovative activity results in incremental adaptation or genuine disruption (Volberda et al., 2021). This shift calls for process-oriented explanations that clarify how innovative practices are converted into disruptive outcomes rather than assuming direct causal effects.

Contingency theory provides such an explanation by highlighting internal conversion mechanisms that condition the effectiveness of innovation practices and initiatives. IWP and innovation initiatives stimulate experimentation and idea generation, but disruptive outcomes materialize only when organizations possess adaptive capabilities and cultural and leadership conditions that enable coordination, learning and scaling (Felin et al., 2015).

Recent theoretical advances reinforce this logic by emphasizing the dynamic interaction between organizational capabilities and environmental change. Rather than passively responding to disruption, firms actively shape innovation trajectories through capability reconfiguration and managerial sense-making, consistent with contingency theory (Cristofaro et al., 2025). Empirical evidence from financial institutions further confirms that DC are critical in translating internal innovation initiatives into organizational transformation under digital disruption (Li et al., 2022; de Paula Pereira et al., 2024).

Accordingly, this study conceptualizes IWP as an innovation stimulus, DC and IC as conversion mechanisms and TL as a contextual boundary condition shaping DI in emerging-market banking contexts.

DI describes processes through which new value propositions, organizational practices or business models alter established market logics and competitive structures over time. Contemporary research increasingly emphasizes that DI is not defined solely by technological novelty, but by how organizations reorganize activities, redefine customer value and scale alternative solutions (Christensen et al., 2018; Chen et al., 2023).

Recent studies highlight that disruption in service and financial sectors tends to emerge gradually through organizational adaptation rather than abrupt market displacement. In regulated environments such as banking, DI often results from internal experimentation, recombination of resources and strategic reorientation rather than from standalone digital technologies (Kuhlmann et al., 2023; Preißner et al., 2024). This view challenges technology-deterministic accounts of disruption by foregrounding organizational processes as central drivers of disruptive outcomes.

Empirical evidence further suggests that firms fail to realize DI when innovative initiatives are decoupled from internal coordination and learning mechanisms (Pemer and Werr, 2023). Accordingly, DI is best conceptualized as an organizational outcome shaped by how firms mobilize internal practices and capabilities under environmental constraints, rather than as an automatic consequence of innovation investment (Amankwah-Amoah et al., 2024).

Consistent with this reasoning, the present study treats DI as an outcome contingent on internal organizational alignment rather than as a purely technology-driven phenomenon.

IWP refer to organizational arrangements that encourage employee learning, experimentation, collaboration and problem-solving in daily work activities. Rather than functioning as isolated HR tools, IWP shape how employees generate, share and refine ideas that may challenge existing routines and market assumptions (Amabile and Pratt, 2016; Martin, 2017).

Recent research emphasizes that IWP contribute to DI by expanding variation and enabling bottom-up experimentation within organizations. Through practices such as cross-functional collaboration, autonomy and knowledge sharing, employees are more likely to explore alternative solutions that diverge from established practices (Mazzei et al., 2016; Lei and Le, 2021). However, emerging evidence suggests that IWP do not automatically produce disruptive outcomes; their effectiveness depends on whether organizations can integrate and scale employee-driven innovation (Brunswicker and Chesbrough, 2018; Khan and Liu, 2023).

Contemporary studies further argue that in service and financial sectors, IWP are particularly important because disruption often originates from process and service redesign rather than from technology alone (Izzo et al., 2022; Preißner et al., 2024). Accordingly, IWP increase the likelihood of DI by fostering exploratory behavior, while the realization of disruption remains contingent on supportive organizational conditions:

DC refer to a firm’s ability to sense environmental changes, seize emerging opportunities and reconfigure resources to maintain competitiveness under uncertainty (Teece, 2014; Farzaneh et al., 2022). Within innovation research, DC are increasingly viewed as organizational processes that convert dispersed innovative activity into coordinated strategic outcomes rather than as standalone resources (Zahra et al., 2022).

Recent studies emphasize that IWP generate ideas and experimentation, but DI materializes only when organizations possess the capability to integrate, prioritize and scale these initiatives (Karimi and Walter, 2015; Volberda et al., 2021). In service and financial sectors, DC are particularly critical because disruption often requires reconfiguring routines, compliance structures and customer interfaces rather than launching entirely new technologies (Li et al., 2022).

Emerging evidence further shows that DC operate as a conversion mechanism linking internal practices to innovation outcomes under digital disruption (de Paula Pereira et al., 2024; Helfat and Raubitschek, 2025). Accordingly, DC are theorized as a mediating mechanism through which IWP are translated into DI rather than as a direct antecedent of DI:

IC reflects shared organizational values, norms and beliefs that legitimize experimentation, learning from failure and the pursuit of novel solutions. Unlike DC, which emphasize structured processes of sensing and reconfiguring, IC operates as a social and normative mechanism that shapes how employees perceive risk, creativity and deviation from established routines (Hogan and Coote, 2014; Rubin and Abramson, 2018).

IWP contribute to the development of IC by embedding innovation-oriented behaviors into daily work activities. When employees observe that experimentation is encouraged and failure is tolerated, they are more likely to engage in exploratory behavior that supports DI (Arsawan et al., 2022; Barros and Ramos, 2023). Recent studies emphasize that IC is particularly important in regulated service sectors, where informal norms often determine whether innovative ideas are pursued or suppressed despite formal support mechanisms (Hidayat and Kassim, 2023; Gad David et al., 2023).

Emerging evidence further suggests that IC functions as a mediating mechanism by translating employee-level innovation initiatives into organization-wide acceptance and continuity (Ismail, 2023). Accordingly, IC is theorized as a distinct social pathway through which IWP influence DI outcomes:

TL is characterized by vision articulation, intellectual stimulation and individualized consideration, enabling leaders to motivate employees to transcend routine expectations and embrace change. Within innovation contexts, TL does not operate as a direct source of disruptive outcomes but as a contextual condition that shapes how innovation practices are enacted and interpreted (Herold et al., 2008; Shafi et al., 2020).

Recent research emphasizes that leadership becomes particularly influential under conditions of uncertainty, where employees require sense-making and legitimacy to pursue nonroutine innovation activities (Lim and Moon, 2022; Tian et al., 2023). From a contingency perspective, TL strengthens the effectiveness of IWP by creating psychological safety, aligning experimentation with strategic intent and reducing resistance to deviation from established routines (Osborn and Marion, 2009).

Emerging empirical evidence further suggests that TL amplifies innovation outcomes not by replacing organizational practices, but by conditioning their effectiveness in dynamic and regulated environments (van Dun and Kumar, 2023; Schiuma et al., 2024). Accordingly, TL is conceptualized as a boundary condition that moderates the relationship between IWP and DI:

DC do not emerge automatically from IWP; they require leadership conditions that coordinate learning, prioritize experimentation and integrate dispersed initiatives into organizational routines. From a contingency perspective, TL provides this enabling context by shaping how innovation-oriented behaviors are interpreted, supported and institutionalized (Osborn and Marion, 2009; Teece, 2014).

Recent research emphasizes that TL plays a critical role in capability formation by aligning employee-driven innovation with strategic direction and by legitimizing resource reconfiguration under uncertainty (Wamalwa, 2023; Helfat and Raubitschek, 2025). In the absence of TL, innovative practices may remain fragmented, limiting their translation into repeatable sensing, seizing and reconfiguring processes (Karimi and Walter, 2015).

Empirical evidence further suggests that leadership is especially salient in regulated service sectors, where capability development requires overcoming structural inertia and risk aversion (de Paula Pereira et al., 2024). Accordingly, TL is theorized as a boundary condition that strengthens the relationship between IWP and DC rather than as a direct antecedent of capability development:

Figure 1 presents the conceptual framework illustrating the relationships among the study variables based on the related literature.

The study participants’ population comprised managers from the banking sector in Punjab, Pakistan. Data were gleaned using both online (430) and self-administered (70) questionnaires. Banks were stratified based on their public (4) and private (21) sector relevance, using a stratified sampling approach (Hayat et al., 2018). The survey was accompanied by a cover letter explaining its purpose. The data were garnered from the five largest cities in Punjab, Pakistan: Lahore, Faisalabad, Rawalpindi, Gujranwala and Multan. The process of data collection spanned over three months, starting from mid-February 2023. We disseminated a total of 500 questionnaires, of which 427 (368 online and 59 hardcopy) responses were collected. Analysis was run on 396 responses after data screening. Post hoc power analysis showed 100% power yield (1.00, p < 0.05) with four independent variables and two interaction terms, this power was well above the 80% threshold, justifying that 396 was more than enough sample size to test the hypothesized relationships (Faul et al., 2009; see  Appendix). Table 1 bears the details on demographics of the study.

Six items, developed by Martin (2017), were used to assess IWP. A specimen item is “I am creative at my work.” DC, the first mediator, was assessed on an eight-item scale proposed by Lin and Wu (2014); and IC, the second mediator, was evaluated on a ten-item scale proposed by Dobni (2008). The model questions for DC and IC are given respectively: “We rapidly respond to market changes” and “Innovation is an underlying culture not just a word.” TL was evaluated against eight items adopted from MacKenzie et al. (2001). A sample question is: “The leader goes beyond self-interests for the good of the group.” DI was gauged using five items from Song et al. (2005). A typical question is: “The positive and negative impact of technological changes on our business is high.” All items were assessed on a seven-point Likert scale. The survey instrument, consisting of 37 questions, was pretested with a small group of respondents representative of the target population, resulting in minor modifications to improve item clarity, readability and overall comprehensibility (Hayat et al., 2019; Nawaz et al., 2021). Following the methodological guidance of Hayat et al. (2025), content validity was established through an extensive review of the literature related to the study constructs, coupled with careful definition of each construct to ensure conceptual clarity. In line with established measurement practices, well-validated scales from prior research were adopted to operationalize the constructs. The retained items collectively capture the core conceptual dimensions of each construct – IWP, TL, DC, IC and DI – thereby ensuring comprehensive coverage of their respective theoretical domains. These measures were selected because they have been extensively validated in previous studies across multiple contexts, providing strong theoretical and empirical support for their content validity. This approach is widely recognized as a rigorous method for establishing content validity in structural equation modeling (SEM) research (Hair et al., 2021).

Descriptive and demographic analysis were conducted using SPSS 27, and data analysis was carried out using SmartPLS 4. The analysis proceeded in two stages, following Hair et al.’s (2019) methodology: first, the measurement model was evaluated, and then the structural model was assessed.

Hair et al. (2019) recommend three types of validity to evaluate a study’s measurement model, namely, content; convergent; and discriminant.

According to Hair et al. (2021), an item must have a minimum loading of 0.50 (reference value for factor loading or FL), indicating that the constructs accounted for a minimum of 50% of the variance in the indicators. The lowest FL value was 0.53, whereas the highest was 0.94. Although a few items had factor loadings slightly below 0.70, they were retained because all constructs demonstrated satisfactory internal consistency (CR, α, rho_A ≥ 0.70) and convergent validity (average variance extracted [AVE] ≥ 0.50). Consistent with established guidelines for measurement models and partial least squares structural equation modeling (PLS-SEM) recommendations, items with loadings between 0.50 and 0.70 may be retained when overall construct validity is adequate and their removal does not meaningfully improve reliability or AVE (Hair et al., 2019, 2021). This approach is also supported by prior research (Hayat et al., 2025). The FL values for a total of four items (one for DC, two for IC and one for TL) were below the threshold value of 0.50 and were excluded from the final analysis. The removed indicators reflected weak or overlapping aspects of their respective constructs and failed to capture theoretically distinct facets, whereas the retained items continued to adequately represent each construct’s conceptual domain, thereby preserving content validity. This item refinement approach is consistent with prior empirical research using established scales and does not undermine content validity when the retained indicators sufficiently cover the construct domain (e.g. Hayat and Afshari, 2022; Sadiq et al., 2026). The outcomes of indicator reliability and convergent validity are given in Table 2.

Internal consistency and convergent validity of the constructs were assessed against composite reliability (CR), Cronbach’s alpha (α), rho_A and AVE. The recommended minimum criterion for CR, α and rho_A is 0.7, as proposed by Cohen et al. (2013) and Hair et al. (2019). The study analysis revealed CR, α and rho_A values ranging from 0.75 to 0.93, 0.74 to 0.91 and 0.75 to 0.92, respectively, which confirms that the constructs exhibited satisfactory levels of internal consistency. In a similar vein, the AVE scores exhibited values from 0.57 to 0.82, surpassing the established benchmark of 0.5, as indicated by Bagozzi and Yi (1988) and Hair et al. (2021). The findings serve to validate the constructs, affirming their acceptable convergent validity. Table 2 displays the comprehensive outcomes of CR, α, rho_A and AVE.

The outcomes of the descriptive analysis, HTMT, Fornell and Larcker’s (1981) criterion and correlation among variables are demonstrated in Table 3. Hair et al. (2019) recommend that the desirable score for HTMT should be < 0.85. Our scores for HTMT range from 0.33 to 0.72, validating DV through the HTMT ratio. Furthermore, the Fornell and Larcker’s (1981) technique was used to affirm the DV. Based on this particular criterion, constructs must have a square root of AVE greater than their respective correlation with any of the remaining constructs. This criterion also verified the DV, as the scores for square root of the AVE were greater than their association with other constructs. Thus, the study findings confirmed the presence of DV.

All the study variables shared positive correlations, indicating that an increase in one variable is associated with an increase in another. Intercorrelation among variables ranged from weak to strong. The weakest correlation was detected between IC and TL. Conversely, the strongest correlation was detected between DC and DI (see Table 3).

Given the cross-sectional and single-source nature of the data, the potential risk of common method bias (CMB) was carefully considered. Following recommendations by MacKenzie and Podsakoff (2012), both procedural and statistical remedies were used. Although marker-variable techniques represent an alternative approach, they are not mandatory when strong theoretical grounding and multiple procedural and statistical remedies are jointly applied. Procedurally, respondents were assured of anonymity and confidentiality, no sensitive or personally identifiable information was requested, and participants were informed that there were no right or wrong answers, thereby reducing evaluation apprehension and social desirability concerns (Podsakoff et al., 2003). Statistically, CMB was assessed using the full collinearity variance inflation factor (VIF) approach, a robust diagnostic in PLS-SEM that captures both vertical (predictor–predictor) and lateral (predictor–criterion) collinearity (Kock, 2015; Kock and Lynn, 2012). Full collinearity VIF is distinct from traditional multicollinearity diagnostics, as it captures both vertical and lateral collinearity and is specifically designed to detect common method variance in PLS-SEM models (Kock, 2015). Lateral collinearity reflects shared variance due to method effects and may spuriously inflate structural path estimates. VIF values for all latent constructs were well below the conservative threshold of 3 (Table 4), indicating that CMB is unlikely to distort the estimated relationships. This approach, widely adopted in prior PLS-SEM studies (e.g. Hayat and Afshari, 2022; Hayat et al., 2025), provides a rigorous safeguard, particularly when combined with the procedural remedies and theoretical justification for the hypothesized relationships. While no method can fully eliminate CMB in cross-sectional, single-source data, these combined safeguards provide reasonable assurance that the findings are not seriously compromised.

To address potential endogeneity in the structural model, we conducted a Gaussian copula test in SmartPLS (Hair et al., 2019). The main predictor, IWP and the mediators, DC and IC, were tested for correlation with the residuals of DI. TL, as a moderator, was not included, as moderators are typically not assessed for endogeneity. The results indicated that none of the predictors were significantly endogenous: IWP → DI (t = 0.412, p = 0.680), DC → DI (t = 1.113, p = 0.266) and IC → DI (t = 0.517, p = 0.838). All p-values exceed 0.05, confirming no endogeneity and suggesting that the estimated structural relationships are unlikely to be biased (Hair et al., 2019). However, given the cross-sectional nature of the data, causal interpretations should be made with caution.

The researchers tested the hypotheses and the suggested model’s significance using SmartPLS’s bootstrapping method on a subsample of 5,000. The guidelines provided by Hair et al. (2021) were followed. We evaluated the structural model in terms of its explanatory power (R²), effect size (f²), predictive relevance (Q²) and the goodness of fit. The findings are presented in Table 4, entitled “summary of the structural model.” The R² score for DC, IC and DI was 0.23, 0.31 and 0.54, respectively. The study R² scores are weak for the mediators (DC and IC) and moderate for the dependent variable (DI), indicating sufficient explanatory power of the exogenous variable. Nevertheless, it is not advisable to evaluate models exclusively based on the R² metric. Hence, the calculation of Q² is performed to assess the predictive relevance of the structural framework. Chin (1998) recommended that the Q² score should be greater than zero. For the current study, the Q² scores for all the endogenous variables were above 0, indicating the predictive relevance of exogenous variables for the outcome variables. Cohen (1992) divided the f² score into robust, moderate and small categories, ranging from 0.35, 0.15 and 0.02, respectively. Our findings demonstrated the effect size (f²) of IWP to be small (0.08) on DI, moderate (0.27) on DC and large (0.36) on IC. Both mediators (DC and IC) exhibited a moderate effect on DI with scores of 0.15 and 0.19, respectively. Following Hair et al. (2019), model fit was evaluated using standardized root mean square residual (SRMR), d_ULS and d_G, with SRMR < 0.08 as the threshold. The results indicated satisfactory fit (SRMR = 0.06; d_ULS = 0.57; d_G = 24), with nonsignificant d_ULS and d_G values confirming robust model fitness.

To assess out-of-sample predictive performance, PLSpredict was conducted using 10-fold cross-validation with ten repetitions, focusing on the key target construct, DI. As reported in Table 5, all DI indicators show positive Q²_predict values, indicating meaningful predictive relevance. In addition, the PLS-SEM model consistently yields lower RMSE values than the linear model benchmark across all indicators, as reflected by the negative PLS-SEM minus LM RMSE differences (Shmueli et al., 2019). Together, these results demonstrate that the proposed model exhibits adequate out-of-sample predictive capability for DI, supporting its predictive as well as explanatory validity.

A summary of the hypotheses testing is given in Table 6. Hypotheses were evaluated using the path coefficient (β), t-value and p-value obtained via SmartPLS bootstrapping with 5,000 resamples. For H1, the results indicate a strong positive association between IWP and DI (β = 0.69, t = 7.35, p < 0.001), suggesting that organizations implementing more IWP tend to experience higher levels of DI. This finding supports contingency theory by highlighting that the effectiveness of IWP is contingent upon enabling organizational conditions, reinforcing the theoretical link between work practices and innovation outcomes. This study hypothesized that the relationship between IWP and DI is mediated by DC (H2) and IC (H3). Consistent with methodological guidelines (Hair et al., 2019), direct associations were first confirmed. The mediation analysis specified that both DC (β = 0.24, t = 3.13, p = 0.002, 95% CI [0.14, 0.33]) and IC (β = 0.19; t = 2.71, p = 0.007, 95% CI [0.11, 0.26]) partially mediated the relationship between IWP and DI. The mediation mechanism suggests that DC enhances DI by improving organizational flexibility, resource integration and responsiveness to market changes, whereas IC fosters a culture that encourages experimentation, knowledge sharing and employee-driven innovation. From a contingency theory perspective, these findings indicate that IWP operate through multiple, distinct internal contingencies – structural (DC) and cultural (IC) – rather than through a single uniform pathway. Thus, H2 and H3 stand supported. This suggests that IWP is associated with DI both directly and indirectly by strengthening organizational capabilities and fostering a supportive IC, consistent with contingency theory (e.g. Yusuf et al., 2023). By unpacking these parallel mechanisms, the findings extend contingency theory beyond a simple fit argument and provide a more process-oriented explanation of how IWP are associated with DI. These findings offer interpretive insight, highlighting how specific organizational processes and cultural enablers act as pathways through which IWP influence DI outcomes.

Table 7 presents the results of the moderation and moderated mediation analyses using bias-corrected bootstrapping with 5,000 resamples, with statistical inferences based on 95% confidence intervals. The endogenous variable is DI. Results indicate that IWP (β = 0.22, t = 3.29, p < 0.001), TL (β = 0.28, t = 5.71, p < 0.001) and the interaction between IWP and TL (β = 0.61, t = 9.43, p < 0.001, 95% CI [0.43, 0.78])) strengthen the positive relationship between IWP and DI. Hence, the authors conclude that H4 stands supported, indicating that TL moderates the positive association between IWP and DI. At lower TL levels, the association between IWP and DI is weaker, while at higher TL levels, the association is stronger, indicating that TL conditions the association between IWP and DI. This is consistent with leadership theory, emphasizing TL’s role in providing guidance, motivation and a supportive environment that enables employees to leverage innovative practices. Importantly, this finding identifies TL as a critical boundary condition, demonstrating that the association between IWP and DI is contingent upon leadership behaviors that mobilize and legitimize innovation-oriented actions. As illustrated in Figure 2, the positive association between IWP and DI is stronger at higher levels of TL. Similarly, results indicate that IWP (β = 0.17, t = 2.24; p < 0.025), TL (β = 0.31, t = 4.55, p < 0.001) and the interaction between IWP and TL (β = 0.52, t = 7.37, p < 0.001, 95% CI [0.38, 0.65]) strengthen the positive relationship between IWP and DC. Thus, H5 is also supported, indicating that TL moderates the positive association between IWP and DC. These findings underscore TL’s critical enabling role, showing how leadership fosters the development of capabilities that support innovation. Together, the moderation results refine contingency theory by demonstrating that both IWP and capability development are not universally effective but depend on leadership-driven contextual conditions. Figure 3 shows that the IWP–DC relationship becomes stronger as TL increases.

As shown in Figure 2, TL strengthens the positive association between IWP and DI, such that this relationship is stronger at higher levels of TL and weaker at lower levels of TL.

As shown in Figure 3, TL strengthens the positive association between IWP and DC, such that this relationship is stronger at higher levels of TL and weaker at lower levels of TL.

This study set out to examine how IWP translate into DI-driven growth under specific organizational contingencies in Pakistan’s banking sector. Drawing on contingency theory, the findings demonstrate that IWP do not operate as a universally effective innovation mechanism; rather, their disruptive potential depends on the presence and alignment of internal contextual conditions, namely, DC, IC and TL (Osborn and Marion, 2009; Talukder and Wang, 2023).

The strong direct association between IWP and DI confirms that employee-centered innovation practices remain a primary driver of disruptive outcomes even in highly regulated service contexts (Lei and Le, 2021; Pemer and Werr, 2023). This finding extends prior DI research, which has predominantly emphasized technological investments, by showing that human-centric work practices constitute a foundational mechanism for disruption in service organizations (Amabile and Pratt, 2016; Brunswicker and Chesbrough, 2018). Importantly, the persistence of a significant direct effect alongside mediated pathways suggests that IWP contribute to DI both through formal organizational processes and through less codified behavioral experimentation (Felin et al., 2015). From a contingency perspective, this indicates that not all innovation value is captured through structured capabilities or culture alone. Some disruptive potential emerges from localized, practice-based experimentation that bypasses formal routines (Osborn and Marion, 2009; Kivimaa et al., 2021).

The mediation results provide deeper insight into how IWP are converted into disruptive outcomes. The partial mediation through DC indicates that IWP enhance DI by strengthening firms’ ability to sense market shifts, reconfigure resources and respond adaptively to technological turbulence (Teece, 2014; Karimi and Walter, 2015; Zahra et al., 2022). This finding refines DC theory by positioning DC not as an independent predictor of innovation, but as a contingent mechanism that translates employee-level innovation practices into organizational-level disruption (Volberda et al., 2021; Pihlajamaa, 2023). In parallel, IC emerges as a complementary social mechanism that enables experimentation, legitimizes risk-taking and sustains innovation momentum (Hogan and Coote, 2014; Arsawan et al., 2022; Barros and Ramos, 2023). The coexistence of DC and IC as parallel mediators advances contingency theory by demonstrating that structural and cultural contingencies operate simultaneously rather than sequentially in shaping DI (Felin et al., 2015; Yusuf et al., 2023).

The moderation results further strengthen the contingency logic of the model. TL significantly amplifies both the IWP–DI and IWP–DC relationships, indicating that leadership behavior conditions whether innovative practices evolve into capabilities and disruptive outcomes (Shafi et al., 2020; Lim and Moon, 2022). Notably, the stronger moderating effect of TL on the IWP–DC pathway suggests that leadership plays a more critical role in capability formation than in innovation output alone (Teece, 2014; Carvalho et al., 2023). This finding refines leadership theory in innovation contexts by showing that transformational leaders do not merely inspire creativity, but actively enable the organizational learning and resource integration processes that underpin DC development. Collectively, these findings reposition TL as a contextual enabler rather than a substitute for innovative practices, reinforcing the core premise of contingency theory that effectiveness arises from alignment rather than isolated excellence.

This study makes several theoretical contributions to the innovation and leadership literature. First, it extends contingency theory by empirically demonstrating that the relationship between IWP and DI is neither linear nor universal, but contingent upon internal organizational configurations (Osborn and Marion, 2009; Talukder and Wang, 2023). By modeling DC, IC and TL within a unified moderated-mediation framework, the study moves beyond static fit arguments and offers a process-oriented explanation of how disruption emerges in service organizations (Felin et al., 2015; Volberda et al., 2021).

Second, the findings contribute to the microfoundations of innovation by clarifying how employee-level practices are transformed into firm-level disruptive outcomes (Felin et al., 2015; Amabile and Pratt, 2016). The identification of DC and IC as parallel mediators highlights that innovation depends on both adaptive processes and cultural legitimacy, challenging studies that privilege one mechanism over the other (Teece, 2014; Arsawan et al., 2022). This dual-pathway logic refines existing innovation models by showing that disruption requires simultaneous investment in capability-building and cultural support.

Third, the study advances leadership theory by demonstrating that TL functions as a boundary condition that shapes the effectiveness of innovation practices. Rather than exerting a direct influence on DI, TL strengthens the translation of IWP into both capabilities and innovation outcomes (Shafi et al., 2020; Tian et al., 2023). This insight deepens theoretical understanding of leadership as a contextual force that enables alignment among practices, culture and capabilities.

The findings offer clear, evidence-based implications for managers in financial institutions and other regulated service sectors. First, organizations seeking disruptive outcomes should prioritize the design and institutionalization of IWP rather than relying solely on top-down innovation strategies (Lei and Le, 2021; Buhalis et al., 2023). The strong direct effect of IWP on DI indicates that empowering employees through experimentation, collaboration and knowledge sharing is a critical starting point for disruption.

Second, managers should recognize that innovative practices alone are insufficient unless supported by enabling capabilities and culture. The mediating roles of DC and IC suggest that firms must deliberately invest in learning routines, resource reconfiguration processes and cultural norms that legitimize experimentation (Teece, 2014; Attar and Abdul-Kareem, 2020; Bansal et al., 2023). Without these contingencies, IWP may generate ideas that fail to scale or translate into disruptive outcomes.

Third, leadership development emerges as a strategic lever for innovation effectiveness. The moderating role of TL indicates that transformational behaviors such as intellectual stimulation, individualized consideration and vision articulation are essential for converting innovative practices into adaptive capabilities (Ismail, 2023; van Dun and Kumar, 2023). Managers are therefore advised to integrate leadership development with innovation initiatives rather than treating them as separate domains.

Within Pakistan’s banking sector, these insights are particularly salient given the dual pressures of regulatory rigidity and digital disruption. The results suggest that banks can balance compliance and innovation by adopting IWP such as cross-functional innovation teams, data-driven experimentation and agile project structures, while relying on transformational leaders to legitimize deviation from entrenched routines. This alignment enables banks to pursue DI without undermining operational stability.

Despite its contributions, this study has limitations that offer avenues for future research. First, the empirical context is limited to Pakistan’s banking sector, which may constrain generalizability. Future studies could replicate the model across industries and national contexts to examine how different institutional environments shape contingency effects. Second, the cross-sectional and self-reported design limits causal inference (Hair et al., 2019). Longitudinal, experimental or multisource studies would strengthen understanding of how IWP, DC and IC evolve over time to sustain DI (Kraus et al., 2023; Uddin et al., 2023). Third, future research could explore additional boundary conditions, such as regulatory intensity or technological turbulence, to further refine contingency-based innovation models. Finally, qualitative or mixed-methods approaches could deepen insight into how employees and leaders enact innovative practices in practice (Felin et al., 2015), enriching the process-level understanding of DI.

This study advances understanding of DI by demonstrating that IWP generate disruptive outcomes only when aligned with supportive organizational contingencies. By integrating DC, IC and TL into a unified framework, the findings extend contingency theory and offer a nuanced explanation of how disruption emerges in service organizations (Osborn and Marion, 2009; Teece, 2014; Volberda et al., 2021). The results underscore that innovation effectiveness is not driven by isolated practices or leadership styles, but by their strategic alignment within specific organizational contexts. In doing so, the study provides a theoretically grounded and empirically validated framework that informs both scholarship and managerial action in innovation-driven environments.