Sustained Learning as a Dynamic Capability for Digital Transformation: A Multilevel Quantitative Study on Workforce Readiness and Digital Services in Healthcare

Abstract

In the context of the digital transformation of healthcare organisations, this study investigates the critical role of sustained learning, employee readiness, and supportive learning conditions to enable digital service offerings. Drawing on dynamic capabilities theory, we conceptualise and empirically test a multilevel model, exploring how sustained learning behaviour and mindset shape the Ability–Motivation–Opportunity (AMO) framework at the individual level. Furthermore, we analyse how workplace learning mediates the relationship between AMO on service outcomes at an organisational level, with sector affiliation as a moderating factor. Data were collected from 856 participants with online surveys and analysed with PLS-SEM. The results confirmed that sustained learning significantly enhances individual readiness (ability, motivation, and opportunity), which in turn positively influences digital services. Workplace learning was found to be a potent mediator, and sector affiliation significantly moderated the relationship between workforce enhancement and digital service outcomes. These findings underline the importance of embedding an employee sustained learning mindset and behaviour as an organisational capability, beyond technical implementation. The results suggest that a successful digital transformation hinges on cognitive and behavioural learning engagement, supported by supportive learning structures and context-specific strategies.

1. Introduction

Service-oriented sectors such as healthcare are witnessing numerous innovations in digitally transformed environments. Digital technologies should enhance processes and ensure that healthcare is accessible and affordable in increasingly ageing societies [1,2]. Patients can benefit from digital service offerings, for which various factors, such as data integration, interoperability, technological compatibility, platforms, and IT security, must be managed [3]. Despite the anticipated benefits, healthcare faces limited adoption and reluctance to implement digital technologies [4]. Only providing technological opportunities and legislative frameworks is insufficient if individual actors, along with their willingness and readiness to adapt and integrate digital technologies into their daily routines successfully, are overlooked [5].

The digital transformation in healthcare is not only technological, organisational and cultural changes in processes must be initiated [6]. Organisations operating in healthcare must increasingly develop dynamic capabilities to navigate these changes and harness their benefits [6,7]. As a fundamental capability, knowledge and learning play a critical role in this human-centric sector [8].

Previous research highlighted the relevance of knowledge, learning and workforce readiness [9,10]. Since the impact of individual factors is crucial [11,12]; the ongoing disruptive changes necessitate fundamentally transforming individual capabilities and the readiness to adopt innovation [13,14].

Particularly in the healthcare sector, which faces unique challenges due to various stakeholders and diverse competencies developing competencies on a continuous and sustainable basis is essential for overcoming limited adaptation, resource constraints, and technological capabilities [15].

However, while prior studies have examined technology adoption and the relevance of lifelong learning, continuous professional development, or learning orientation [16,17,18], few have explicitly addressed the behavioural and cognitive mechanisms that enable employees to continuously reconfigure competencies during transformation.

This study fills that gap by introducing the novel concept of sustained learning, distinguishing it from existing constructs by combining behavioural engagement with aa learning-focused mindset as a measurable, dynamic capability.

Grounded in dynamic capabilities theory, we create a conceptual model to address the existing knowledge gaps, explaining the relationships and interdependencies on these two levels. We empirically test the presumed effect of SL and AMO as individual-level factors on digital service offerings in healthcare, mediated by WPL as an organisational supporting condition.

The rest of the paper is structured as follows. Section 2 reviews the literature and develops hypotheses. Section 3 outlines the methodology, Section 4 presents the results, Section 5 discusses implication, concluding in Section 6 with contributions, limitations and future research directions.

2. Literature Review and Hypothesis Development

Contemporary workplaces are characterised by uncertainty and disruptive changes due to digital transformation. As there is no standard definition of digital transformation, this paper adopts the summary proposed by Vial [19], which defines digital transformation as a process aimed at enhancing services or processes by instigating significant changes through the implementation and use of technology. This shift demands new digital competencies and effective knowledge management. Learning and knowledge are vital for fostering innovation and enabling continuous transformation [20]. To adapt, organisations require dynamic capabilities—abilities to integrate and reconfigure resources in rapidly changing environments [21].

In healthcare, ongoing technological advances, regulatory shifts, and evolving patient needs make knowledge acquisition and adaptation essential for sustainability. Research shows that sustained learning strengthens capabilities critical to transformation and adaptability [22,23]. This paper examines how sustained learning functions as a dynamic capability in healthcare, and how structured learning can support digital integration into everyday practice.

2.1. Dependent Variable: Digital Service Offerings

Digital transformation involves more than simply converting analogue processes into digital ones. Digital services require a comprehensive redesign based on user needs, or the creation of entirely new digital services [6].

The success of digital service offerings relies on the interplay between a robust IT infrastructure and a readiness to embrace innovation at both individual and organisational levels. Investing in technology alone is inadequate; organisational capabilities, shifts in mindsets, and adjustments in processes are vital for progress [24]. An organisation’s ability to assess its internal digital infrastructures and investigate new solutions is essential for harnessing the potential of digital transformation [25].

2.2. Independent Variables

Digital services can realise their full potential only if investments are made in a robust and secure IT infrastructure while also promoting a culture of innovation and continuous learning that encompasses individual motivation and the relevant skills of employees [26]. Therefore, this paper investigates the effects of sustained learning (SL) as the capability to sense, seize, and reconfigure digital competencies in response to evolving demands, structured employee empowerment by enhancing their motivation, abilities and opportunities to participate, and accessible learning opportunities provided by structured organisational workplace learning concepts.

2.2.1. Sustained Learning Behaviour

In the context of ongoing transformation, sustained learning serves as a crucial individual-level prerequisite for developing dynamic capabilities [20]. We define sustained learning as a behavioural pattern rooted in a learning-oriented mindset that facilitates continuous development or competence reconfiguration through autonomous, proactive problem-solving and skill acquisition [27]. Therefore, the multidimensional construct of sustained learning consists of two interconnected yet distinct dimensions: sustained learning mindset and sustained learning behaviour.

Rooted in organisational learning [28], sustained learning behaviour involves observable actions that show an individual’s active engagement in learning activities. We argue that this behaviour is crucial at the individual level. It improves capability by building relevant knowledge and skills, increases motivation through proactive problem-solving and continuous self-directed learning, and creates opportunities by encouraging active interaction with available learning resources and participation mechanisms. Based on the above considerations, we formulate the first set of hypotheses:

Hypothesis 1a (H1a). 

An individual’s sustained learning behaviour positively affects the enabling set of skills and knowledge (ability).

Hypothesis 1b (H1b). 

An individual’s sustained learning behaviour positively affects the willingness to engage in specific actions (motivation).

Hypothesis 1c (H1c). 

An individua’s sustained learning behaviour positively affects the appropriate resources as enabling circumstances (opportunity to participate).

2.2.2. Sustained Learning Mindset

A sustained learning mindset refers to an individual’s lasting cognitive orientation towards growth, curiosity, and improvement. It includes beliefs about learning as a continuous, self-directed process and reflects the psychological readiness to engage with new knowledge over time. This concept is grounded in the growth mindset theory [29], which emphasises the motivational and attitudinal foundations of learning engagement. A sustained learning mindset enhances individual capabilities by encouraging openness to acquiring complex knowledge; it boosts motivation by grounding learning in intrinsic values; and it promotes opportunity by increasing the chances that individuals will recognise and seize learning moments within their work environments. Therefore, we propose the second set of hypotheses:

Hypothesis 2a (H2a). 

An individual’s sustained learning mindset positively affects the facilitating skills and knowledge (ability).

Hypothesis 2b (H2b). 

An individual’s sustained learning mindset positively affects the willingness to engage in specific actions (motivation).

Hypothesis 2c (H2c). 

An individual’s sustained learning mindset positively affects the appropriate resources as enabling circumstances (opportunity to participate).

2.2.3. Workforce Enhancement: AMO Framework

Previous research emphasises the importance of motivation, as well as the opportunity to engage in and acquire new skills during digital transformation [30]. Individual growth and sustained development are the most crucial criteria for employability amid technological change [31]. The Ability–Motivation–Opportunity (AMO) framework is a well-established model in human resource management that highlights three key elements: ability, motivation, and opportunity to [32]. According to this framework, consistent individual performance depends on the systematic development of essential skills and knowledge (ability), the willingness to undertake specific behaviours (motivation), and the suitable environmental conditions that facilitate action (opportunity to participate) [33] In contrast to sustained learning, which is the dynamic ability that arises from the continuous use of individual open attitudes and active behaviours in the context of digital change, the AMO model describes the basic capabilities (skills, motivation, opportunities) required for employee participation and performance. While sustained learning describes the “how” and “why” factors, AMO explains the “what” of the continuous adaptation process in the digital age.

The AMO concept provides a valuable perspective for understanding how workforce dynamics influence technological innovation and service delivery. We argue that improving individual skills can enhance digital services when employees are properly motivated and have genuine opportunities for learning and participating in change processes. Thus, we hypothesise:

Hypothesis 3 (H3). 

Enhancing the workforce’s independent, long-term ability and willingness, including the necessary resources and opportunities for participation at an individual level, positively affects digital service offerings as a desired organisational outcome.

2.3. Mediating Variable: Workplace Learning

Workplace learning (WPL) refers to structured and unstructured processes through which individuals acquire, apply, and refine their knowledge and skills within their occupational context [34]. It encompasses formal training programmes, informal learning-by-doing, peer learning, mentorship, and reflective practices embedded in daily work activities [35]. Importantly, it supports ongoing competence development in response to changing organisational and technological demands.

In the disruptive change caused by digital transformation, structured workplace learning is crucial in transforming individual capabilities into meaningful organisational outcomes, including enhanced digital service delivery and innovation. While ability and willingness are prerequisites for learning engagement, workplace learning offers the institutional and procedural support that sustains skill development over time and ensures alignment with organisational goals.

Structured workplace learning bridges the individual-level readiness for change with the systemic realisation of innovative digital services. It facilitates the conversion of individual potential into collective, digitally enabled performance by embedding learning processes into routine healthcare operations.

Hypothesis 4 (H4). 

Structured workplace learning positively mediates the effect of enhancing the workforce’s independent, long-term ability and willingness, including the necessary resources and opportunities for participation at an individual level, on digital service offerings.

2.4. Moderating Effect of Sector-Specifics

Previous research indicates that learning abilities vary across various types of organisations. Accordingly, we assume that private sector organisations are more likely to pursue improvements through learning and digital service enhancement than those in public healthcare [36]. Studies examining the influence of HR management and strategies for digital service integration highlight the importance of assessing context-specific potential moderators to understand sector-specific traits [3,37].

Hypothesis 5 (H5). 

The sector in which the organisation operates influences the strength of the effect on digital services.

Figure 1 displays the research model.

3. Materials and Methods

Five variables were established based on existing research and theories to examine the relationship among the constructs in the conceptual model. A questionnaire was created to assess the measures and test the hypothesis, which was distributed via a link to an online survey targeting various healthcare professionals. Additionally, data collected through the crowdsourcing platform Prolific were included.

Table 1. Purposive sampling.

Sample	Reference
All employees of a statutory health insurance company (app. 3500 employees)	[39]
Mailing List of the Rhineland-Palatine Hospital Association (app. 90 hospitals)	[40]
All employees of a nursery home (app. 400)	[41]
Mailing List of the Rhineland-Palatine General Practitioners (GP) Association (app. 3000)	[42]
Mailing List of the city local Health Innovation Hub (app. 50)	[43]
Additional data from full-time or part-time employed participants via Prolific for an international diversified sample	[44,45]

provides an overview of the sampling and references of applications in previous research. This purposive sampling is appropriate as it captures core stakeholder groups of the healthcare system across diverse roles and organisational contexts. This heterogeneity ensures ecological validity and allows testing how the supposed mechanisms manifest under different workplace learning conditions [38]. The addition of an international Prolific sample further enhances generalizability by enabling comparisons between the healthcare sector and other industries.

3.1. Operationalisation of the Variables

Given the conceptual breadth of this construct, our measurement items are aligned closely with validated scales in prior literature, ensuring comparability and conceptual clarity. Items to measure the dependent variable of digital service offerings are derived from the work of Arranz et al. [46] and Kwilinski et al. [25], who investigate IT infrastructure, organisational innovation, and digital services. We operationalised digital service offerings through measurable indicators for the extent of digital services, alongside participants’ perceptions of the organisation’s adoption of innovation and IT infrastructure adequacy. These items are appropriate for measuring the desired outcome, as previous literature highlights the importance of IT infrastructure and innovation adoption for the successful implementation and further development of digital services. Digital transformation is not merely a technological challenge; it is also a human and organisational endeavour [3].

The measures for the two subdimensions of the independent variable, sustained learning, are based on dynamic capabilities theory and theoretical foundations that cover the cognitive and behavioural aspects of sustained learning [47,48,49]. The authors developed the construct and dimensions, validated and tested the items initially, and demonstrated their distinctiveness from similar constructs [27].

The items assessing ability, motivation, and opportunity to participate are based on those used in previous studies [50,51,52]. The items in the survey assess factors considered relevant at the individual level in research.

The proposed mediating variable, workplace learning, encompasses items that evaluate a supportive infrastructure, communication, and both formal and informal learning. Prior literature has underscored these as vital components for effective workplace learning [34,53]. These items were utilised as an appropriate measure within a human-centred approach [54,55].

To investigate the possible moderating effect of the sector in which the participants are employed, we asked the prolific participants about the sector in which their current organisation operates. The data collected in the first five samples (see Table 1) were coded as the healthcare sector.

3.2. Survey Design

Different scales were used for psychological separation to avoid common method bias: a 5-point Likert scale for digital competencies and a 7-point Likert scale for other items [56]. As a second step, the measures of dependent and independent variables were methodologically separated by using different scales and psychologically differentiated by inserting socio-demographic questions in between [56]. The questionnaire comprised 34 questions based on existing scales used in previous studies; the construct of sustained learning was developed by the authors based on established concepts and theories. The survey was distributed in German to the selected sample shown in Table 1, ensuring proper representation of healthcare actors and providing additional data to validate the results. Further data was gathered via the crowdsourcing platform Prolific in English from fully or partly employed individuals to create a more valid and generalizable database and to check whether results can be specifically attributed to the health sector.

The quality of the translation was discussed with experts. Sample questions with references to the sources are displayed in

Table 2. Excerpt of sample survey questions.

Construct	Sample Questions	Sources
Ability–Motivation–Opportunity to Participate 12 items	I look for digital self-service options and know how to use them. I regularly spend time for trainings to extend my digital knowledge I feel a sense of satisfaction when I realise ideas, possible improvements and innovation.	[49,57,58,59]
Workplace learning 5 items	There are sufficient and easily accessible offers for training in the context of my work. In our work environment we communicate and take action to stay up to date with new technology.	[53,54,55]
Digital service offerings 3 items	To what rate is digital transformation present at your workplace? How would you characterise the digitalisation of services provided by your organisation? How would you characterize your organisation ability to be innovative?	[46,60,61]
Sustained learning mindset 3 items	If there are conversations and discussions about new technologies in my environment, I am interested in them.	[27]
Sustained learning behaviour 3 items	When I learn something new about digital technologies, I rethink my previous actions and try to develop them further with the new knowledge.	[27]

; additional material is available on request. The statistical analysis was conducted using SmartPLS software Version 4.1.1.5 with standardised data.

4. Results

443 healthcare professionals responded to the survey, complemented by 413 respondents collected on the platform Prolific. Discontinued surveys and bot responses were filtered out [62]. The remaining data sets contain random missing answers [63]. We decided to keep these responses to avoid distortions caused by the systematic deletion. Since only a small proportion of values are missing, the mean replacement method is suitable for handling missing data when performing the partial least squares structural equation modeling (PLS-SEM) algorithm [64]. PLS-SEM using SmartPLS was chosen because it is particularly suitable for exploratory models, non-normally distributed data, and theory development objectives. Alternative methods, such as CB-SEM, were deemed less appropriate as they assume normality and are better suited for theory confirmation [65]. Assuming that the sample size should be at least ten times the number of free parameters in the model, the sample size of 856 respondents exceeds the recommended thresholds for PLS-SEM, ensuring statistical power [66].

The sample included 383 (44.7%) male and 458 (53.5%) female participants. The age groups from 25 onwards are evenly distributed, each accounting for around 20%. Approximately 10% of participants were under 25, and about 6% were over 60. Most participants (nearly 60%) are working in the healthcare sector, the second-largest group in the services sector (approx. 21%).

Table 3. Socio-demographic background of the participants.

Element	Total Number = 856	Occurrence	Percentage
Gender	Male	383	44.7%
	Female	458	53.5%
	Divers	4	0.5%
	Missing data	11	1.3%
Age group	Under 25	93	10.9%
	25–30	165	19.3%
	31–40	163	19.0%
	41–50	180	21.0%
	51–60	196	22.9%
	Over 60	49	5.7%
	Missing data	10	1.2%
Sector	Healthcare	502	58.6%
	Academia/Education	35	4.1%
	Manufacturing	53	6.2%
	Finance	3	0.4%
	Services	179	20.9%
	IT	27	3.2%
	Sales	48	5.6%
	Government	4	0.5%
	Missing data	5	0.6%
Tenure	Less than 1 year	58	6.8%
	1–5 years	265	31.0%
	5–10 years	144	16.8%
	11–20 years	97	11.3%
	More than 20 years	285	33.3%
	Missing data	7	0.8%

provides an overview of the respondents’ socio-demographic analysis.

In the first step, the software Jamovi Version 2.4.8.0 was used to assess the data. The survey data was collected from the same respondents at one point in time; therefore, we evaluated the potential impact of common method bias on the data. Harman’s single-factor test (HSF) was employed as an ad hoc measure. This analysis assumes that the dominance of a single factor indicates common method bias if, when all items are combined into one factor, the percentage of variance explained by this factor exceeds the threshold of 0.5 [67]. The resulting variance of 0.32 is below the threshold of 0.5, indicating that the item characteristics differ and common method bias is unlikely to be present in this study [68]. Confirmatory factor analysis (CFA) was conducted to evaluate the model fit. The chi-square statistic measures the difference between the expected and observed data; a lower chi-square value indicates a better model fit. The results are presented in

Table 4. Summary of the model fit indices.

	Test for Exact Fit			Fit Measure
X2	CFI	df	p	RMSEA
1354	0.859	215	<0.001	0.0787

. The chi-square test suggests no exact fit, which is understandable given the number of items [69]. RMSEA shows a reasonable model fit. A CFI of 0.95 is recommended as an indicator of a good fit, an RMSEA of less than 0.05 indicates a good fit, 0.08 indicates a reasonable fit, and an RMSEA of over 0.1 indicates a poor fit [70]. The resulting CFI of 0.859 and RMSEA of 0.0787 indicate a reasonable model fit.

Next, the data distribution was assessed with the Shapiro–Wilk test. Since p was <0.01 for all items, the test is significant, and the data are not normally distributed. Therefore, SmartPLS 4 was used to test and analyse the results, applying PLS-SEM, as a normal distribution is not a requirement. PLS-SEM is suitable for multivariate analysis and is widely used in business research [71].

4.1. Outer Model Results

The model was constructed and further tested in Smart PLS software, evaluating the outer model for reliability. Items with weaker loadings that do not meet the recommended threshold for indicator reliability should only be deleted to enhance composite reliability or internal consistency [71].

In our sample, we deleted three items of the construct AMO: two from the Ability dimension (COMP1, EDU2) and one item from the Opportunity dimension (PERS2). Although these items were removed due to weak loadings, the retained indicators sufficiently represent the construct dimensions, preserving content validity. Cronbach’s alpha and Composite reliability for all variables are higher than 0.7. For convergent validity, the average variance extracted (AVE) should exceed 50%. These assessments are achieved after this elimination, thus ensuring internal consistency and item reliability [72]. Figure 2 visualises the model, the results are displayed in

Table 5. Outer model results with composite reliability and convergent validity for all variables.

	Cronbach’s Alpha	Average Variance Extracted (AVE)	Composite Reliability (rho_a)	Composite Reliability (rho_c)
Ability–Motivation–Opportunity to Participate AMO 9 items	0.826	0.429	0.837	0.868
Workplace learning WPL 5 items	0.846	0.620	0.851	0.890
Digital service offerings DigServ 3 items	0.764	0.620	0.851	0.890
Sustained learning behaviour SUSB 3 items	0.769	0.683	0.771	0.866
Sustained learning mindset SUSL 3 items	0.846	0.765	0.868	0.907

.

As a following criterion for outer model validity, discriminant validity was assessed using the heterotrait–monotrait (HTMT) ratio of correlations [73]. Each indicator’s factor loading is more significant than that of all other constructs (see

Table 6. Matrix of heterotrait–monotrait (HTMT) ratio.

AMO	DigServ	SUSB	SUSM	WLEAR	SEC × AMO
AMO
DigServ	0.621
SUSB	0.634	0.702
SUSM	0.659	0.307	0.808
WPL	0.821	0.796	0.390	0.379
SEC × AMO	0.090	0.144	0.078	0.052	0.045

), except for self-determined motivation/innovation adoption, which is acceptable since both constructs are conceptually similar [64]. This suggests that the items assess distinct constructs. Since these standards are fulfilled, the outer model’s validity is given.

The variance inflation factor (VIF) was checked for all items for collinearity statistics. Since VIF is lower than 3 for all items, multicollinearity is unlikely [74]. The results for all items are provided in Appendix A.

4.2. Inner Model Results

Furthermore, the inner model is tested for relationships between the variables. R² measures the extent to which the independent variables predict the dependent variable. Values above 0.75 are described as substantial, 0.5 as moderate, and 0.25 as weak [71]. The inner model results are displayed in

Table 7. Predictive power, path coefficients and significance of the pathways as results of the inner model.

Path	R2	Path Coefficient	p-Value	T-Value
DigServ	0.445
AMO → DigServ		0.103 *	0.003	2.947
WPL → DigServ		0.537	0.000	17.231
AMO → WPL		0.693	0.000	35.428
SUSB → Ability		0.250	0.000	6.022
SUSB → Motivation		0.212	0.000	6.028
SUSB → Opportunity to participate		0.199	0.000	4.980
SUSM → Ability		0.336	0.000	8.191
SUSM → Motivation		0.402	0.000	10.747
SUSM → Opportunity to participate		0.312	0.000	7.674

* p-value < 0.01.

. The results indicate that the model has moderate predictive power.

The hypothesised positive effects on developing skills and knowledge (ability), individual willingness (motivation), and active involvement (opportunity to participate) caused by sustained learning behaviour (H1a–H1c) and a sustained learning mindset (H2a–H2c) are confirmed. There is also empirical evidence supporting the positive impact of these enhancements on the workforce’s ability, willingness, and participation opportunities in digital service offerings (H3).

The mediation effect was calculated using Baron and Kenny’s approach [75]. AMO significantly impacts the dependent variable DigServ when the mediator WPL is not present. Adding the hypothesised mediating variable WPL in step 2 shows a significant connection between AMO and WPL. Adding this mediating factor, the effect between the independent and dependent variables is significantly decreased [76], as presented in the third step in

Table 8. Mediating effect of Workplace Learning on Digital Services.

Construct	(1) DigServ (WLEAR Absent)		(2) WLEAR		(3) DigServ (WLEAR Present)
	Path Coeff.	p-Value	Path Coeff	p-Value	Path Coeff	p-Value
AMO	0.486	0.000	0.689	0.000	0.100	0.003

.

Based on these steps, the proportion of the variance of a dependent variable that is explained by a mediation relationship (Variance Accounted For = VAF) is calculated (see

Table 9. Results of the mediation analysis of digital competence.

Direct Effect	Indirect Effect	Total Effect	VAF	Mediation
0.100 *	0.689 × 0.575 = 0.396	0.496	0.798	Partial mediation

* p < 0.01.

). According to Hair et al. [65], a partial mediation of digital competence is observed for the relationship between AMO and DigServ, confirming hypothesis H4.

To test the effect of the sector-specific factor, an independent sample t-Test was calculated (

Table 10. Impact of the sector on the outcome DigServ.

	Test-Statistic	df	p	Effect Size
t-Test	−10.3	833	<0.001	Cohen’s d	−0.720
Mann–Whitney-U	52,856		<0.001	Rank biserial correlation	0.376

). The participants were divided into two groups: healthcare (n = 502) and all other sectors (n = 349). The mean values of the result variables for digital services differ significantly between the two groups (2.89 for healthcare, 3.44 for others).

The t-value is a measure of the statistical significance of a difference between two group means or a relationship between variables. The negative sign of the t-value indicates the direction of the effect. In this comparison of the two groups [77]. The results show a highly significant difference between the significantly lower value of the first group compared to the second group, with a very low probability that this difference occurred by chance [78]. The effect size quantifies the magnitude of the observed effect or relationship, regardless of the sample size. For correlations, an effect size of >0.3 is considered strong [79]. The current effect size (−0.72) indicates a strong correlation and means that the differences are not only statistically significant but also meaningful. Since the result of the Shapiro–Wilk test (W = 0.952, p < 0.001) shows that there is no normal distribution, the Mann–Whitney U test was also performed. The test statistic value also shows a systematic difference between the two groups. The rank-biserial correlation of 0.376 indicates a moderate effect. Thus, the probability that a randomly selected person from the “others” group has a higher digital service score than a randomly selected person from the healthcare sector is significantly above 50%.

It is evident that the healthcare sector has a substantial impact on digital services. Participants from the healthcare sector score noticeably lower on average (Mean 2.89) than participants from other sectors (Mean 3.44) and are significantly less intensive users of digital services than those in different industries, as shown in the plot in Figure 3.

Therefore, hypothesis 5 is confirmed, as we see a relevant impact of the sector on the outcome.

5. Discussion

This study examined the connections between sustained learning behaviour and mindset as a dynamic capability in digitally transformed workplaces, focusing on workforce development based on the Ability-Motivation-Opportunity to participate concept, and the preparedness of digital workplaces for expanded digital service offerings. It also explored the mediating role of formal and informal workplace learning, especially within the healthcare sector, including the sector as a context-specific potential moderator to improve generalisability and validity. Our findings confirmed hypotheses H1a–H1c and H2a–H2c, demonstrating that both sustained learning behaviour and mindset significantly enhance ability, motivation, and opportunities to participate. H3 was supported, indicating that enhanced workforce readiness contributes positively to digital service offerings. H4 was validated through the mediating role of workplace learning, while H5 was confirmed, showing sector-specific moderation effects.

Our findings provide empirical support for the fundamental role of ongoing learning engagement, which we call sustained learning (SL) and workforce development initiatives at the individual level in driving digital transformation. Using the theoretical perspective of dynamic capabilities [7], sustained learning as a behavioural pattern combined with a learning-focused mindset positively influences the dimensions of the AMO construct. Continuous engagement in proactive learning activities (Sustained learning behaviour) is positively linked to perceived ability (H1a: path coefficient 0.250; p-value 0.000), motivation (H1b: path coefficient 0.212; p-value 0.000), and opportunities to participate (H1c: path coefficient 0.199; p-value 0.000). These findings are consistent with prior research, which suggests that behavioural engagement in learning forms a microfoundation for developing adaptive capabilities [17,49]. This microfoundations movement in strategy and organisation theory examines how individual actions and interactions explain phenomena in companies [80]. Our results reinforce that digital transformation extends far beyond technological implementation. It requires the strategic integration of digital technologies to reshape structures, processes, and workforce dynamics, thereby redefining identity, value propositions, and competitiveness. In line with the microfoundations movement, our evidence highlights that this transformation is deeply rooted in individual behaviours, competencies, performance, and interpersonal interactions, which together underpin the dynamic capabilities needed to adapt in complex, globalised environments.

Similarly, the cognitive component of a sustained learning mindset was found to significantly influence all three AMO factors: perceived ability (H2a: path coefficient 0.336; p-value 0.000), motivation (H2b: path coefficient 0.402; p-value 0.000), and perceived opportunity to participate (H2c: path coefficient 0.312; p-value 0.000). These results align with research on growth mindset theory [29] but go further by contextualising mindset within a digital transformation framework and showing its tangible impact on individual readiness factors.

Both dimensions of sustained learning, behaviour and mindset highlight the importance of cognitive and proactive learning mechanisms. Since previous studies often lack a conceptual integration [81,82], we offer a more nuanced and operational framework for understanding digital knowledge development to build and reconfigure human resources from the perspective of dynamic capabilities theory. Theoretically, the study contributes to dynamic capabilities theory by introducing sustained learning as a measurable, two-dimensional construct that goes beyond existing notions of lifelong learning or learning orientation. Unlike these broader concepts, sustained learning specifically captures the behavioural and cognitive mechanisms that enable ongoing competence reconfiguration in dynamic digital environments.

The posited positive effect of enhanced AMO factors, such as well-developed individual competencies, motivation, and participatory conditions, on digital service offerings was supported by the data (H3: path coefficient 0.103; p-value 0.003). We therefore reinforce the view that digital transformation cannot be achieved through technological investment alone [15,83]. Human-centred strategies that build adaptive capabilities through motivation and skill development are decisive [31].

Addressing the mediating role of individual enhancing measures through organisational workplace learning, the results confirmed that structured and informal learning within the workplace partly mediates the relationship between AMO factors and digital service outcomes (H4: VAF 0.798). This mediation is consistent with prior research [53,54] which emphasises the importance of organisational learning infrastructure, integrating this factor into a multilevel model. Our findings support the argument that without institutionalised learning processes, individual readiness does not reliably translate into digital service outcomes. Thus, workplace learning serves as a transmission mechanism, converting individual potential into organisational capability.

Finally, we investigated the sector-specific context as a moderating factor to answer the call for empirical evidence on that [3]. The significant effect suggests that the digital service outcomes vary depending on whether the organisation operates in the healthcare or other sectors (Healthcare mean: 2.89; others mean: 3.44). This result adds nuance to previous findings [36], who noted the institutional inertia of public organisations like healthcare. Our data suggest that healthcare companies are significantly less developed or less intensive users of digital services than other industries. However, the effectiveness of workforce development strategies is not predetermined by sector alone; it is amplified or constrained by the underlying learning culture and organisational autonomy.

Our findings challenge traditional top-down models of digital transformation that emphasise leadership, technology, or funding as primary levers [4,84,85]. Instead, our results suggest that readiness for digital service innovation is grounded in integrated concepts to boost learning engagement. Employees who continuously learn and adapt, supported by learning-enabling organisational conditions, become the true carriers of transformation. Moreover, the sector-specific differences imply that no universal blueprint for digital transformation exists. Contextual contingencies must be accounted for when designing learning architectures, particularly in public-sector organisations that operate under higher regulation and limited agility. Our study, therefore, offers a more holistic understanding of the interplay between individual learning, organisational fostering conditions, and sector-specific context. It reconceptualises digital service readiness as a dynamic, sustainable, co-created learning process. The findings are adaptable to similar, strictly regulated environments, like public administration.

6. Conclusions

Digital transformation is an ongoing process that requires organisations to redesign interconnected frameworks for learning, innovation, and adaptation. Based on our findings, we reflect on the current transitional state of digital work environments, where the uncertainty of new technologies challenges motivation as a key factor for learning in organisational settings. This emphasises the need for change management that incorporates structured learning opportunities and support mechanisms, prompting practitioners to reconsider learning strategies and adopt a holistic model that integrates individual, organisational, and technological factors. Organisations need to develop their human capital by acquiring knowledge and adopting new mindsets through sustained learning. This study examined the effects and relationships of learning, empowerment, and facilitating structures as relevant aspects in digital transformation and offers an empirically grounded contribution to the further development of dynamic capabilities theory by introducing the concept of sustained learning. The results confirm the hypothesised mechanism that sustained learning is a prerequisite for developing workforce enhancement concepts, which in turn affects digital service outcomes, mediated by structured workplace learning. Our data suggests that the healthcare sector engages less in digital services than others. It is therefore worthwhile to take a comparative look at other sectors to develop appropriate measures for individual development, increased adaptation and organisational adaptability. Consequently, we advise a cautious approach to generalisation, involving context-sensitive modifications to workforce development and learning strategies.

This paper concludes with several contributions, summarised in three main points. First, it advances dynamic capabilities theory by introducing the concept of sustained learning as an essential success factor for digital transformation. Our findings underscore the relevance of intentionally investing in the cognitive and behavioural learning development of employees. We also provide validation for the novel measurement of this two-dimensional construct, offering additional methodological innovation. The readiness of employees, shaped by their attitudes and prior experiences, is thus a decisive enabler in the change processes. Extending the microfoundation perspective, our research demonstrates that sustained behavioural learning acts as a catalyst for mobilising individual resources essential for managing digital change.

Second, the study bridges the research streams of HR, learning, and digital transformation, examining the impact of the established AMO concept on enhanced digital service offerings. We therefore provide evidence for essential mechanisms that enable the current workforce to participate in digital transformation. We demonstrate that workplace learning operates not just as a complementary factor, but as a necessary organisational condition for realising digital services.

Third, we proved that external factors, such as organisational requirements and environmental conditions, are influential. There are sector-specific constraints in healthcare, like regulatory complexity and strict data protection requirements, that might explain why the perceived level of digital services lags behind other industries. Moreover, legacy IT infrastructures, heterogeneous system landscapes, and fragmented stakeholder interests impede the seamless integration of digital tools into care processes. These structural and cultural factors must be addressed to keep up with other, more commercially driven sectors.

The findings have practical implications, advising managers to prioritise low-risk digital initiatives to build trust while gradually enhancing digital capabilities. Organisations should involve employees and design training methods that promote informal peer-to-peer learning, thereby supporting the autonomous development of competencies needed to use digital technologies effectively and confidently. For policymakers, the findings point to the need for supportive regulatory frameworks that enable innovation without compromising patient safety, targeted funding for infrastructure modernisation, and information and knowledge-building campaigns.

Certainly, several limitations must be acknowledged. The cross-sectional data collection captures participants’ perceptions at a specific point in time. Since learning in digital transformation is a dynamic process, a longitudinal study would provide deeper insights into the learning mechanisms over time at different maturity stages. Additionally, factors such as leadership styles or team support may also influence digital service outcomes, but were not examined in our study. Future research could explore these additional factors to provide a more comprehensive understanding of the evaluated model. This would allow for the identification of developmental trajectories in workforce adaptation, which are particularly relevant in dynamic digital transformation contexts. As we have proven the relevance of sector-specifics, our findings may not apply to other, highly competitive contexts. Empirically, using a more diverse sample in terms of various sectors, regional specifics, and culture would enhance external validity and determine whether the identified mechanisms will operate similarly in other sectors.

Overall, this study highlights a vital insight: digital transformation succeeds when learning becomes an ongoing, employee-driven, and organisationally embedded practice. Beyond the dynamic capabilities’ perspective, these findings could be further interpreted through complementary lenses. Expectancy theory may reveal how situated learning and work routines enable or hinder the translation of sustained learning into tangible service innovations. Integrating additional theoretical angles could yield a richer, multi-layered understanding of capability emergence in complex organisational environments.

In conclusion, sustained learning serves as a dynamic capability essential for driving digital transformation in healthcare. By embedding continuous learning at both individual and organisational levels, healthcare organisations can overcome sectoral barriers and achieve sustainable innovation.