Agile work methods, job autonomy and innovative work behaviour: examining their interplay in organisational contexts

In today’s rapidly evolving business landscape, characterised by technological advancements and global competition, innovation is crucial for organisational survival and competitive advantage (Teece et al., 2016). Within organisations, innovation involves the development of new products, technologies, business models and solutions arising from employees’ spontaneous initiatives (De Jong and Den Hartog, 2010; Janssen, 2000).

Agility has emerged as a key concept encompassing organisational flexibility, adaptability and resilience (Dhanpat, 2024; Junker et al., 2023; Serrador and Pinto, 2015). Despite the absence of a unique interpretation of the concept, agility centres on collaborative teamwork and customer-centric practices (Petermann and Zacher, 2022). Research links agility to improved job performance and satisfaction, indicating its potential to nurture employees’ innovative work behaviour (IWB) (Janssen, 2000; Salmen and Festing, 2021). However, most of the research has focused on the impact of agile work practices (AWPs). These refer to agile organisational routines and denote the recurring implementation of agility at the workplace level (Junker et al., 2023), while agile work methods (AWMs) capture the underlying principles of agility as expressed in individual employees’ ways of working. Specifically, with AWMs we refer to the application of agile principles in the execution of employees’ daily work tasks.

By fostering cooperation and knowledge sharing, AWMs can enhance the emergence of innovative ideas (Franco and Landini, 2022). However, previous research shown mixed findings regarding the relationship between agility and innovation. The systematic literature review by Nguyen et al. (2024) reveals that a non-negligible amount of research papers reports mixed effects for this relationship. These mixed findings depend on innovation type (radical vs incremental), contextual factors and implementation quality, creating a theoretical gap that justifies the examination of mediating mechanisms to understand when and how AWMs can influence innovation. This is also urgent from a practical standpoint, as contemporary organisations face unprecedented challenges including increasing market volatility, accelerating innovation cycles and more employee demands for workplace autonomy (Teece et al., 2016; Dhanpat, 2024). Moreover, improper or standardised implementation of agile principles may have detrimental effects within organisations, leading to higher employees’ stress and workload (Meier and Kock, 2023). This may lead to resource draining and exhaustion for some employee categories, highlighting the relevance of defining boundary conditions for the implementation of AWMs. These pressures make understanding the relationship between AWMs and IWB critical for organisational competitiveness in today’s dynamic business environments.

In this scenario, job autonomy plays a critical role in fostering innovation by enhancing productivity and work quality (Johannsen and Zak, 2020). Defined as employees’ capacity to make independent work-related decisions, the Job Demands-Resources (JD-R) theory identifies autonomy as a key job resource that enhances motivation and well-being (Xanthopoulou et al., 2007). In agile work settings, autonomy enables employees to execute tasks independently while benefitting from teamwork support (Gupta et al., 2019). Nevertheless, research has also put into evidence potential drawbacks due to workplace autonomy, such as increased stress, perceived lack of support and continuous thinking about work (De Spiegelaere et al., 2016). Thus, autonomy can serve as a crucial variable linking agility and IWB (Černe et al., 2017; Lai et al., 2021; Sherehiy and Karwowski, 2014) but could also have a negative or null effect on employee performance (Muecke and Iseke, 2019).

Building on the JD-R, we examine the impact of AWMs on IWB by considering the mediating role of autonomy. Using survey data from Italian multinationals collected in 2024, this study advances research on agility and innovation by providing new insights for scholars and firms seeking to foster employee IWB through agility. The investigation is timely in Italy, where the adoption of agile principles is expanding. Beyond new practices and organisational changes, this raises issues concerning job and personal resources linked to work environments and individual skills (Rusconi, 2024). Italian employees generally enjoy greater autonomy, but also work more under time pressure than employees in other EU countries (Eurostat, 2020). Against this backdrop, while AWPs and AWMs may offer useful solutions for firms, assessing the effect of AWMs on IWB becomes particularly relevant. Moreover, we selected autonomy as our focal mediator because: (1) apart from the JD-R theory, it represents a fundamental psychological element in other job-related frameworks, such as the self-determination theory (SDT); (2) agile methodologies leverage on employee self-organisation and decision-making freedom; (3) contemporary workforce research has consistently identified autonomy as a top employee priority, particularly after the Covid-19, by also highlighting the double-edged effects that it may have on their performance; and (4) autonomy has served as a proximal mechanism through which AWMs can directly influence employee IWB.

The paper addresses calls for empirical studies on the multi-dimensional aspects of agility (Ulrich and Yeung, 2019), specifically by proposing a new concept and measurement scale for AWMs. Furthermore, the paper contributes to academic literature on the antecedents to IWB and the interactionist perspective on creativity. Moreover, it provides further evidence on the impact of autonomy on IWB. Despite the small mediating effect, the results contribute to clarifying the by-now inconclusive relationship between autonomy and employees’ performance (Muecke and Iseke, 2019), specifically IWB, and the antecedent role of AWMs. Despite agility may imply higher degrees of autonomy in the workplace, we corroborate the idea that the two are distinct concepts. As the implementation of AWMs may make employees feel more and less autonomous at the same time (Meier and Kock, 2023), we contribute to clarify that the relationship between them is positive and their impact on IWB is positive, supporting our hypotheses.

The paper provides theoretical insights for organisational theory, human resources and innovation management, alongside practical guidance for innovation-driven workplaces. The paper is structured as follows: Section 2 outlines the theoretical framework, Section 3 presents the methodology, Section 4 shows the findings, while Sections 5 and 6 discuss the results and conclude the study.

IWB refers to the processes through which individuals generate, promote and implement ideas within organisations (Janssen, 2000). It encompasses activities that contribute to the development and application of novel ideas, processes or products (De Jong and Den Hartog, 2010). IWB is essential for firms’ sustained competitive advantage, as it requires innovative efforts across all organisational levels (Simonton et al., 1992).

Research identifies multiple antecedents of IWB (Amabile, 1983; Anderson et al., 2014; Mathisen et al., 2012). Hammond et al. (2011) classify antecedents as individual differences, motivation, job characteristics and contextual variables. While individual differences include personality and demographics, motivation encompasses intrinsic engagement and extrinsic incentives (Schaufeli and Bakker, 2004).

Previous research has associated firm-level innovation with physical workplace interactions (Amabile, 1983). However, the shift to remote work following the Covid-19 pandemic has reshaped scholarly discourse and employees’ conditions for IWB due to the use of digital tools and hybrid work (Armstrong et al., 2024; Garlatti Costa et al., 2022; Zia et al., 2024). This is particularly relevant in the context of Italy, where our empirical study was conducted. Currently, the proportion of remote workers stands at approximately 10.3% – a decrease from 13.6% in 2020, yet still more than double the 4.7% recorded in 2019 (Assolombarda, 2025).

According to the JD-R model, job resources play a crucial role in shaping IWB (Bos-Nehles et al., 2017). Thus, the following section examines how AWMs can trigger job resources which, in turn, influence IWB.

AWPs have significant implications for modern work organisations (Sherehiy and Karwowski, 2014; Teece et al., 2016). Despite the absence of a unique definition, agility can be attributed to manufacturing processes, organisations and employees (Chung et al., 2014; Pitafi et al., 2023). However, a critical yet frequently overlooked aspect concerns the distinction among these various levels of analysis (Salmen and Festing, 2021). Prior research has often conflated organisational-level agile practices with individual-level agile behaviours, assuming rather than empirically examining how agility at one level translates to another (Werder and Maedche, 2018). Previous research has described AWPs as workplace routines composed of repetitive organisational and team practices that facilitate change-oriented behaviours (e.g. sprints or stand-up meetings; Junker et al., 2023). This paper defines AWMs as the underlying principles of AWPs, manifested in employees’ individual ways of working. This conceptual distinction directly addresses the aforementioned gap: while organisational-level agility encompasses strategic flexibility and dynamic capabilities for sensing and responding to environmental changes (Teece et al., 2016), individual-level AWMs concern how employees internalise and enact agile principles in their daily work. Our approach moves beyond the assumption that implementing organisational agile practices automatically produces individual agile behaviours, rather examining the psychological mechanisms through which individual-level agility relates to innovation outcomes.

Previous research has shown a positive relationship between agility and job performance (Petermann and Zacher, 2022; Uraon et al., 2024). The recursive nature of AWPs can foster employees’ continuous updating and experimentation, which are essential for IWB (Petermann and Zacher, 2020). Thus, we argue AWMs can favour the mobilisation of job resources such as knowledge sharing and teamwork. From a JD-R perspective, AWMs can stimulate valuable job resources – such as collaboration and iterative feedback (Dhanpat, 2024) – that support employees in managing work demands and fostering innovation (Park and Cho, 2022; Schaufeli and Bakker, 2004). These can facilitate idea generation and innovation, while reducing risks of emotional exhaustion and burnout (Anderson et al., 2014; Bakker et al., 2023; Gupta et al., 2019).

AWMs can stimulate out-of-the-box thinking and problem-solving, leading to higher creativity (Amabile and Pratt, 2016). Learning can ensure updated knowledge implementation and real-world testing of new ideas. Later, by fostering knowledge sharing and teamwork, AWMs can ease the idea diffusion stage. In turn, AWMs can foster synergistic efforts, mutual learning and a fast experimentation approach that motivate employees to proceed with the idea implementation stage. Here, responsiveness and speed can enhance cognitive flexibility, a key element for the elaboration of innovative ideas and the restyling of existing ones (Amabile et al., 1996; Sherehiy et al., 2007).

Additionally, AWMs stimulate personal growth and job satisfaction – which are key job resources, pre-conditions for IWB (Petermann and Zacher, 2020; Sherehiy and Karwowski, 2014). This is particularly relevant, given evidence that the mere adoption of AWPs, without adequate agile principles and employee internalisation, can undermine the success of agility implementation (Nguyen et al., 2024; Rusconi, 2024). Prior research has shown that agile workforces experience greater well-being not directly from the application of agile practices, but indirectly through the outcomes – such as improved task organisation and enhanced teamwork – resulting from the implementation of agile organisational routines (Rietze and Zacher, 2022). Similarly, in line with interactionist scholars (Amabile and Pratt, 2016; Woodman et al., 1993), the job resources triggered through the implementation of AWMs can lead to superior innovation-related performance.

However, the link between AWMs and IWB may not be as straightforward. Rietze and Zacher (2022) highlight agility leads to higher employee well-being only indirectly, by reducing the workload and increasing job resources such as autonomy. Excessive autonomy without organisational support may also lead to stress and exhaustion (Hodgson and Briand, 2013; Muecke and Iseke, 2019). Moreover, AWMs are not a panacea for all corporate problems. This is truly relevant in Italy, where only 23% of the companies implementing agile management report being completely satisfied (Rusconi, 2024), likely because of the absence of adequate background conditions for agility. Last, several studies show mixed results on the impact of agility on performance (Nguyen et al., 2024). These contradictions stem from theoretical tensions inherent to agility: the relationship between environmental turbulence and agility benefits may be curvilinear rather than linear, with agility enhancing adaptiveness under moderate uncertainty but potentially becoming counterproductive under extreme volatility, where stability provides greater value (Teece et al., 2016). Additionally, implementation quality moderates outcomes significantly. As noted earlier, mere adoption of AWPs without adequate internalisation can undermine agility’s success – superficial application without cultural alignment often yields null or negative effects. Finally, temporal trade-offs distinguish between immediate operational stability and long-term viability: Vaculík et al. (2018) find that while organisational agility is a precondition for sustained competitive advantage, it renders firms vulnerable in the short term, often necessitating the termination of innovation projects to realign resources. This highlights the link between AWMs and IWB could be far from intuitive or straightforward. The agility effects on innovation are mediated by individual-level factors, making job autonomy a theoretically relevant mechanism for understanding these relationships.

Nevertheless, based on extensive literature, we posit that AWMs positively affect IWB.

Job autonomy refers to employees’ independence in scheduling tasks and deciding how to execute them (Hackman and Oldham, 1976). According to the Job Characteristics Model (JCM), autonomy is a core dimension of work design, while constituting a fundamental psychological need for the SDT (Deci and Ryan, 2000). Within the JD-R model, autonomy functions as a job resource enhancing motivation and performance (Johannsen and Zak, 2020; Orth and Volmer, 2017).

Studies highlight autonomy as a key predictor of IWB, as it fosters commitment and self-responsibility, enabling employees to experiment with new ideas and approaches (Bos-Nehles et al., 2017). As a job resource, autonomy supports various motivational mechanisms, including engagement and the ability to navigate workplace challenges (Bakker and Demerouti, 2017; Franco and Landini, 2022). Autonomy reinforces employees’ sense of responsibility and proactiveness, encouraging them to exceed workplace expectations (Deci and Ryan, 2008; Giebels et al., 2016). It facilitates problem-solving, role flexibility and adaptability, reinforcing employees’ openness to organisational change (Petermann and Zacher, 2022). Workplace autonomy allows employees to break routines, discover alternative work methods and craft new solutions (Giebels et al., 2016), stimulating creativity and higher performance (Bakker et al., 2023). These dynamics can enhance job satisfaction and overall well-being – critical factors for innovation (Bakker et al., 2023; Rietze and Zacher, 2022; Tims et al., 2014). Recent research has highlighted digital autonomy is also highly relevant to employee creativity (Huu, 2023). Conversely, low autonomy stifles innovation by limiting individual initiative and reducing perceptions of self-efficacy and sense of responsibility (De Spiegelaere et al., 2016). Ensuring autonomy is thus fundamental to fostering creativity and IWB (Jo and Hong, 2022).

Despite this, evidence is not conclusive. Some scholars suggest autonomy primarily enhances the implementation rather than the generation of new ideas (Hammond et al., 2011). These mixed findings align with theoretical perspectives suggesting curvilinear rather than linear autonomy effects. Following Warr’s (1987) vitamin model and the “too-much-of-a-good-thing” effect (Pierce and Aguinis, 2013), autonomy may benefit performance up to an optimal threshold, beyond which additional autonomy can become counterproductive. Additionally, the relationship between autonomy and job performance (including IWB) may be more complex than assumed, as some scholars report marginal or anecdotal effects (Dysvik and Kuvaas, 2011; Sørlie et al., 2022). While autonomy generally benefits structured workplaces, it may also introduce uncertainty and ambiguity in more flexible contexts (Muecke and Iseke, 2019). These contradictions suggest autonomy’s effects depend on boundary conditions such as organisational support, task structure and individual capabilities. Under high job demands and without adequate organisational support, autonomy alone may not be sufficient to drive innovation. This complexity underscores why examining autonomy as a mediating mechanism – as our study does – provides valuable understanding for when and how autonomy translates into innovation benefits. It is therefore important to carefully assess the specific workplace conditions to understand the impact of employee autonomy on IWB.

Nevertheless, building on a large body of literature, we argue for a positive relationship between job autonomy and IWB. We thus formulate the following hypothesis:

AWMs promote employees’ autonomy. In the JD-R model, job autonomy is a key resource that empowers employees to manage workloads, boosting motivation and enabling them to handle job demands individually (Xanthopoulou et al., 2007). According to the job demand-control model, perceived control over job situations reduces stress, enhances well-being and improves performance, organisational commitment and task enjoyment (Karasek, 1979).

Scholars highlight how organisational environments shape employees’ perceptions of autonomy (Orth and Volmer, 2017). AWMs support self-organisation and promote collaboration, favouring the integration of individual tasks into team efforts (Forner et al., 2020). These, in turn, can foster motivation, engagement and a greater sense of control (Junker et al., 2022; Sheldon et al., 2022).

By providing employees with flexibility and independence in structuring their work and decision-making, we argue AWMs strengthen perceived job autonomy. This significantly influences engagement, motivation and job satisfaction (Joo et al., 2010; Rich et al., 2010). Therefore, we argue for the following hypothesis:

Given appropriate contextual conditions (Park and Kang, 2025), job autonomy enables employees to explore creative solutions and adapt to evolving work demands (Karasek, 1979). Under AWMs, employees gain greater autonomy in prioritising tasks, collaborating and experimenting, which increases their likelihood of engaging in IWB. Research has shown employees in agile teams benefit from wider margins for independent decision-making, work methods and scheduling, making job autonomy a key resource for innovation (Hoegl and Parboteeah, 2006; Morgeson and Humphrey, 2006). Given the relevance of IT technologies and digital tools for work, nowadays employees benefitting from wider digital autonomy are more likely to engage in IWB (Huu, 2023).

AWMs also foster intrinsic motivation by satisfying autonomy, enhancing proactivity and self-efficacy (Junker et al., 2022; Moran et al., 2012). In high-demand environments, autonomy helps reduce exhaustion and fosters commitment, problem-solving, engagement and innovation (Gagné and Deci, 2005; Petermann and Zacher, 2022). The SDT also suggests autonomy-supportive conditions, such as AWMs, enhance intrinsic motivation and performance (Moran et al., 2012), which also includes IWB. Therefore, we argue AWMs’ effect on IWB is mediated by job autonomy. AWMs provide the background principles and ways of working for autonomy, which in turn empowers and motivates employees to innovate. In other words, while AWMs establish structured yet flexible ways of working, the autonomy they promote supports IWB.

However, this relationship is not universally positive. Autonomy’s impact on performance depends on contextual factors (Kubicek et al., 2017; Muecke and Iseke, 2019; Park and Kang, 2025). Uncontrolled autonomy may lead to isolation and may signify a lack of managerial or team support, which stifle innovation. Autonomy may also mean low coordination, which could be either beneficial to improve collaboration, or create divergent interpretations and negative reactions among team members (Beretta and Smith, 2023). Moreover, under conditions of high job demands, autonomy may contribute to work exhaustion and reduced IWB (Demerouti et al., 2001).

However, job autonomy remains influential across work contexts. On these bases, we argue for the following hypothesis:

The study adopted a quantitative approach through the collection of survey-based data from three multinational corporations in Italy, selected through convenience sampling. This sampling approach specifically targeted organisations that have demonstrated commitment to and practical experience with agile methodologies in their operations. The firms belong to different sectors, namely frozen food, steel and micro-mechanics, providing diverse contexts for examining AWMs adoption across varying industrial settings. The selection of companies from distinct sectors allows for exploring how AWMs are adapted and implemented across different manufacturing environments.

We sent an online survey to the HR managers of each company. We gave instructions to HR managers on the survey targets, namely employees in different job positions and working units. Then, they had the responsibility to electronically send the questionnaire to their employees. Surveys were sent electronically and the data collection was conducted in the first semester of 2024. Overall, we collected and analysed 744 questionnaires. None of these was removed because of high percentages of missing data.

The questionnaire was initially developed in English and translated into Italian using a back-to-back translation process (Brislin, 1986). Before starting the survey, participants were informed their data would remain confidential, anonymous, collected and stored for research purposes only. Ethics approval for the survey was not required by our institution, in line with Italian norms.

To assess AWMs as the independent variable, we developed and validated a five-item custom scale (Table 1). While prior studies conceptualised AWPs as organisational routines such as sprints, stand-up meetings or retrospective sessions (Junker et al., 2023; Rietze and Zacher, 2022), our operationalisation of AWMs refers to the agile principles that employees mobilise at the individual level through their working modes. Specifically, the scale items capture personal attributes which are necessary for effectively applying AWPs in daily routines and ensuring alignment with them.

A preliminary validation study with Italian professionals (n = 98) representing various organisational roles and industry sectors supported its applicability across industries. Respondents shown a balanced gender distribution (50% women) and most of the respondents (55%) were aged 25–34. More than half (54%) held university degrees.

An exploratory factor analysis (EFA) was conducted to examine the underlying structure of the five items measuring AWMs. The analysis used principal axis factoring (PAF) as the extraction method. The Kaiser–Meyer–Olkin (KMO) measure attested the sampling adequacy (KMO = 0.72), and Bartlett’s test of sphericity was significant (χ²(10) = 129.19, p < 0.001), indicating that the data were suitable for factor analysis. The analysis identified a single underlying factor, with an eigenvalue of 2.09 explaining 41.76% of the total variance. Both the scree plot and the Kaiser criterion supported the one-factor solution. All five items loaded satisfactorily on this factor, with loadings ranging from 0.47 (close to the 0.5 threshold) to 0.74, suggesting that the items reliably measure a common underlying construct, labelled as AWMs (see  Appendix).

After that, we examined construct validity through principal component analysis (PCA), supporting the scale’s unidimensional structure. Last, we used confirmatory factor analysis (CFA) to test model fit, which provided additional evidence of construct validity. Psychometric analysis (Table 1) showed strong reliability, with a Cronbach’s alpha of 0.77. The Omega Hierarchical and Omega Total coefficients both reached 0.78, indicating substantial variance explained by a general factor. The Omega Asymptotic was 1, reflecting maximum theoretical reliability. CFA showed acceptable fit (Chi-Square = 15.31, df = 5, RMSEA = 0.145, RMR = 0.07).

IWB, the dependent variable, was measured using Janssen’s (2000) scale, with responses ranging from 1 (never) to 7 (regularly) (α = 0.92). Job autonomy, the mediation variable, was assessed using Morgeson and Humphrey’s (2006) scale, rated from 1 (strongly disagree) to 7 (strongly agree) (α = 0.93).

The correlation matrix among the variables of interest can be seen in Table 2. Given the relatively strong correlation between AWMs and IWB (r = 0.52), we conducted additional analyses to assess the distinctiveness of the IWB construct, as it can be seen in Table 3. Reliability analysis indicated high internal consistency (Cronbach’s α = 0.92). Both the PCA and the EFA supported a clear one-factor structure, with all items loading strongly on a single factor (loadings ranging from 0.81 to 0.94) and acceptable sampling adequacy (KMO = 0.74; Bartlett’s test p < 0.001).

Conceptually, these findings are consistent with viewing AWMs as a structural and procedural context that enables IWB, rather than as an overlapping manifestation of the same construct. Thus, although the two constructs share meaningful variance, they remain empirically and theoretically distinguishable.

Control variables included gender, age, education, caregiving responsibilities, position and contract type, given their potential influence on IWB. For instance, the level of education could play a role in stimulating out-of-the-box thinking, while younger personnel tend to be more innovation-oriented. Moreover, people with higher-level positions may feel freer to engage in IWB, as well as those holding longer-term contracts. We also collected data on caregiving responsibilities, as situations at home can affect workplace situations.

The main sample characteristics are presented in Table 4.

Although the correlation between agility and IWB is higher than 0.5 (Table 2), the VIF coefficients obtained through linear regression were below the 5 threshold, suggesting low room for multicollinearity issues (Hair et al., 2010). To test our hypotheses, we conducted a mediation analysis using Model 4 of the PROCESS macro v.4.2 for SPSS (Hayes, 2017). We grand mean-centred variables (Aiken et al., 1991) and opted for standardised coefficients. Results are presented in Figure 1 and Table 5.

The results indicate path a from AWMs (X) to job autonomy (M) is significant, β = 0.08, SE = 0.04, p < 0.05. Path b from job autonomy to IWB is also significant with β = 0.28 (SE = 0.03, p < 0.001).

The direct effect of AWMs on IWB (path c’), controlling for job autonomy, shown a coefficient β = 0.41, with a 95% CI of [0.34, 0.52]. This effect is significant (p < 0.001), too.

The indirect effect of AWMs on IWB through job autonomy is β = 0.02, 95% CI [0.01, 0.05], which is also significant.

The total effect of AWMs on IWB (path c) corresponds to β = 0.43, with a 95% CI [0.41, 0.55]. The total effect is significant (p < 0.001). Results support Hypotheses 1, 2, 3 and 4, by providing evidence for the mediating role of job autonomy on the path between AWMs and IWB, where the direct paths of both variables on IWB are positive and significant.

We conducted a Sobel test to assess whether the indirect effect of autonomy on the relationship between AWMs and IWB was significant. The Sobel test statistic coefficient was 2.19, with a p-value of 0.03 (two-tailed), indicating that the mediation effect is significant.

To address concerns regarding alternative causal explanations, we tested a reverse mediation model in which IWB was specified as the predictor, AWMs as the dependent variable and job autonomy played the role of the mediator. Results indicate that the indirect effect was not statistically significant (β = −0.02, SE = 0.02, 95% CI [−0.049, 0.01], p = 0.21). While the direct effect of IWB on AWMs remained strong and significant (β = 0.45, SE = 0.03, p < 0.001), the path from IWB to autonomy was not significant (β = −0.04, SE = 0.03, p = 0.21), indicating that autonomy does not function as a meaningful mediating mechanism in this alternative causal ordering. These findings provide no empirical support for reverse causation and lend further credibility to the theoretically proposed model in which AWMs precede autonomy and IWB.

Beyond statistical significance, we examined the magnitude of the effects. In the theorised mediation model, the indirect effect accounted for approximately 5% of the total effect, which – while modest – is consistent with prior organisational research examining distal outcomes such as IWB. Notably, in the alternative reverse mediation model, the indirect effect was both smaller (3.74%) and non-significant, further supporting the theoretical plausibility and practical relevance of the proposed model.

This study contributes to the fields of innovation management and organisational studies by examining key aspects of job design, employee conditions and innovative performance. It advances the interactionist perspective (Woodman et al., 1993) on organisational and individual conditions for IWB, providing new insights into agility and workplace innovation.

First, our findings reinforce the applicability of the JD-R model in agile work contexts, showing that AWMs can create autonomy-supportive conditions enhancing employee motivation and engagement (Dysvik and Kuvaas, 2011). However, we acknowledge the dual nature of AWMs – while they can serve as triggers for job resources, excessive autonomy or poorly implemented agile practices may also lead to work stress, coordination challenges and role ambiguity (Nguyen et al., 2024; Rietze and Zacher, 2022). Although the indirect effect coefficient appears modest (β = 0.02), its practical relevance should be evaluated within organisational research contexts, where even small mediation effects can represent meaningful workplace changes across large employee populations. We acknowledge that organisational dynamics are complex and that our linear model captures only part of this complexity, with alternative explanations such as organisational culture, leadership support and team dynamics potentially contributing to the observed relationships (Amabile and Pratt, 2016; Hernaus et al., 2022). In line with the JD-R, properly implemented AWMs fulfil employees’ psychological need for autonomy, boosting intrinsic motivation while strengthening extrinsic incentives (Schaufeli and Bakker, 2004). This can create a resource-rich environment that promotes employee well-being and enhances work performance (Ryan and Deci, 2000). Given that, future research could proceed to investigate the conditions under which AWMs can thrive, at both personal and contextual level.

Second, our study contributes to the innovation management literature by empirically supporting that AWMs positively influence IWB through autonomy (Amabile et al., 1996). While the role of agility in corporate performance was already acknowledged, we extend this understanding by responding to calls for empirical research on its workplace impact (Petermann and Zacher, 2020; Salmen and Festing, 2021). Our findings support the notion that AWMs establish the necessary conditions for IWB, aligning with previous literature on the interplay of personal and contextual factors in fostering innovation (Schaufeli and Bakker, 2004; Xanthopoulou et al., 2007). Additionally, we contribute to current literature on organisational agility and human resource management by emphasising the relevance of a comprehensive notion of agility to achieve successful IWB. Nevertheless, our AWMs measure should be tested again and refined to be consistent and reliable.

Third, we provide empirical evidence on the relationship between autonomy and performance, particularly IWB. While prior research has yielded mixed findings, we find autonomy to be a job resource that enhances performance across diverse work arrangements (Kubicek et al., 2017; Muecke and Iseke, 2019), despite not being so determinant per se. Our focus on autonomy as a key resource is justified as most empirical evidence shows positive effects of properly implemented agile methods (Nguyen et al., 2024). The findings support the idea that employee autonomy serves as a key resource across different organisational contexts, supporting the transition from AWMs on IWB. This underscores the necessity of further exploring the conditions under which autonomy could play a greater role.

Overall, this study bridges multiple disciplines – including organisational and innovation management, and psychology – to offer a coherent framework for understanding the interplay between agility, IWB and autonomy. Our multidisciplinary approach provides a robust foundation for addressing evolving workplace conditions and innovation-oriented organisational dynamics.

The findings of this study offer valuable insights for managers and practitioners seeking to enhance employees’ well-being and innovative performance.

First, our results highlight the significance of job conditions for employees’ psychological empowerment and motivation (Sheldon et al., 2022). Workplaces that prioritise flexibility, mutual support and knowledge sharing contribute to employee satisfaction, self-efficacy and well-being, reinforcing individual commitment to innovative solutions.

Second, we emphasise the importance of strengthening employee autonomy, although its effects on IWB in presence of AWMs are evidently not determinant. The relatively small mediating effect observed in this study suggests that autonomy-enhancement interventions should not be interpreted as a primary lever for stimulating IWB in agile contexts. Rather than acting as a dominant mechanism, autonomy appears to function as a complementary job resource that amplifies the innovation-related benefits of AWMs when these are already in place. While agile organisations inherently promote autonomy, we acknowledge that different organisational contexts, roles and job positions require varying levels of it (Kubicek et al., 2017; Stiglbauer and Kovacs, 2018). Organisations should carefully assess their readiness for agile implementation by evaluating existing autonomy levels, stress factors, collaboration capabilities and cultural readiness before adoption. Neither standalone agility nor autonomy ensure higher IWB per se (Werder and Maedche, 2018). Rather, successful AWMs implementation requires a balanced and structured approach tailored to organisational needs (Rich et al., 2010; Teece et al., 2016). In this respect, the modest role of autonomy as a mediator indicates that managerial interventions should primarily aim at strengthening AWMs themselves, rather than focusing exclusively on increasing individual discretion. Practices such as cross-training, mentoring and coaching should therefore be understood as mechanisms that reinforce shared understanding, coordination and learning – core principles of AWMs – while providing bounded and supported forms of autonomy. In these cases, managers should support gradual transitions to ensure workforce alignment. To this end, we recommend targeted initiatives that foster agility at the individual level (Pitafi et al., 2023). For instance, cross-training can enhance flexibility and teamwork (Sherehiy et al., 2007), while mentoring and coaching programs can strengthen responsibility and self-efficacy (Forner et al., 2020). Team-building activities further support collaboration and trust, driving continuous self-development.

Job redesign strategies should focus on boosting autonomy while ensuring sufficient job resources to sustain motivation and self-efficacy (Joo et al., 2010). From a practical standpoint, these findings caution against autonomy-enhancement initiatives implemented in isolation. Increasing discretion without simultaneously reinforcing agile coordination, feedback mechanisms and managerial support may yield limited innovation returns and, in some cases, exacerbate role ambiguity or work-related strain. Empowering employees with more challenging tasks can enhance motivation, buffering job demands and fostering innovation (Oldham et al., 1976; Xanthopoulou et al., 2007). However, autonomy must be paired with strong managerial and team support to prevent work exhaustion and health impairment (Gagné and Deci, 2005; Hernaus et al., 2022).

Finally, organisations should implement initiatives to encourage IWB (Muduli, 2017). Open communication channels and collective brainstorming sessions, particularly in remote settings, can stimulate idea exchange and collaboration. Leadership is instrumental in fostering a culture of innovation, creating work environments where creativity thrives (Mathisen et al., 2012; Mumford et al., 2002; Xu et al., 2024).

By integrating these strategies, organisations can cultivate agile, autonomy-supportive workplaces that enhance innovative initiatives and employee well-being.

Despite its contributions, this study has several limitations that should be acknowledged. First, the cross-sectional design prevents causal inference, and the reliance on self-reported data introduces the potential for response-related biases. Although procedural precautions were taken and common method bias does not appear to be a major concern, as indicated by the VIF values reported in the Results section, future research would benefit from designs that further reduce such risks through temporal separation or multi-source data.

Second, the use of a non-probability convenience sample drawn from three companies may have introduced selection bias and limit the generalisability of the findings. While efforts were made to ensure representativeness across industries, the geographical scope of the study remains constrained. Moreover, given the heterogeneity in organisational structures among the sampled firms, the results should be interpreted primarily at an aggregate level, rather than as firm-specific effects. The modest size of the indirect effect may also be partly attributable to the specific organisational context examined in this study. Data were collected from large multinational manufacturing firms operating in Italy, all of which had already implemented agile practices to some extent. In such relatively mature and structured agile environments, baseline levels of autonomy may already be institutionalised, thereby reducing its salience as a differentiating job resource in explaining IWB.

Third, although the newly developed AWMs scale demonstrated acceptable internal consistency and was validated across participants representing diverse organisational roles and industry sectors, some model fit indicators suggest room for improvement. In particular, the RMSEA value exceeds commonly recommended thresholds, indicating that construct validity should be further examined. Additional validation across different organisational cultures, company sizes and international contexts is therefore necessary to strengthen the robustness and broader applicability of the scale.

Fourth, while the proposed linear mediation model was statistically supported, the indirect effect was modest, suggesting that the relationships under scrutiny may not fully capture the complexity of organisational dynamics. Future research should consider alternative model specifications, including reverse causality, spurious relationships, nonlinear effects and interaction terms, as well as report additional effect size measures beyond standardised beta coefficients to facilitate a clearer assessment of practical significance.

Finally, this study has highlighted the positive effects of AWMs, while acknowledging that agile ways of working may also introduce increased job demands, such as heightened pace, role ambiguity or uncertainty. A more balanced perspective that simultaneously examines both enabling and straining effects of agility would provide a more comprehensive understanding of its consequences for employees.

Future research can address these limitations in several ways. First, longitudinal and experimental designs are needed to establish causal relationships and examine the temporal dynamics of AWMs. Second, broader cross-cultural and geographical validation of the AWMs scale, combined with multi-rater or objective data sources, would further strengthen measurement quality and reduce bias. Third, scholars should investigate organisational culture and entrepreneurial orientation as potential moderating factors shaping the effectiveness of AWMs, particularly in determining the optimal balance between autonomy and structure. Fourth, future studies should explicitly explore the dual positive and negative effects of agile practices, identifying boundary conditions under which each is more likely to emerge. In particular, qualitative or mixed-methods research designs could help to unpack when and why autonomy matters more or less in the relationship between AWMs and IWB. In-depth interviews, case studies or ethnographic approaches could provide richer insights into how employees experience autonomy within distinct phases and configurations of agile implementation. Finally, multi-level and intersectoral research designs would deepen understanding of how AWMs operate across organisational levels and industries.

This study supports a positive relationship between AWMs and IWB, highlighting the mediating role of job autonomy. Our findings contribute to innovation management and organisational studies by identifying contextual conditions that foster IWB. Practically, we offer insights into job design factors that enhance IWB, emphasising autonomy as a key job resource that, when supported by AWMs, stimulates employees’ innovative performance (Muduli, 2017). This research provides organisations with actionable insights for implementing agile methodologies while maximising their innovation potential. For practitioners, our findings suggest that successful agile transformation requires careful attention to autonomy and contextual factors, rather than one-size-fits-all approaches. For theory, we advance understanding of how job design factors operate in contemporary work environments characterised by increased flexibility and digital transformation.

An exploratory factor analysis (EFA) was conducted on the five items (Q1–Q5) to examine their underlying structure. The extraction method used was principal axis factoring (PAF), with no rotation performed as only one factor was extracted.

The KMO value above 0.70 and the significant Bartlett’s test of sphericity indicate that the dataset was suitable for factor analysis.

Communalities after extraction ranged between 0.22 and 0.54, suggesting a moderate amount of shared variance among the items. One item (Q5) exhibited a relatively low communality (0.22), but was retained due to its theoretical relevance to the construct.

A single factor emerged with an eigenvalue greater than 1, explaining 41.76% of the total variance.

All items loaded positively on the single factor, with loadings ranging from 0.47 to 0.74, indicating satisfactory convergence of the items on one latent construct.

The EFA results supported a unidimensional structure for the five items. The single-factor model demonstrated adequate sampling adequacy (KMO = 0.724), significant inter-item correlations (Bartlett’s test p < 0.001) and acceptable item loadings (>0.45), supporting the reliability of this latent construct for further analysis.