Influence of Employee Well-Being and Work Flexibility on Innovative Work Behavior and Job Performance: A Comparative Study of Full-Time and Gig Workers in Digital Business

Abstract

This study investigates the impact of employee well-being, work flexibility, and innovative work behavior on job performance among full-time and gig workers in digital businesses. A comparative analysis was conducted to examine potential differences between the two groups. A structured questionnaire was administered to 201 full-time employees in digital business system development and 199 gig workers from the IT Support Thailand group on Facebook using convenience sampling. The data were analyzed using multiple group structural equation modeling (MG-SEM) via partial least squares (PLS). The findings reveal that work flexibility boosts innovative work behavior, with gig workers showing greater adaptability than full-time employees. Innovative work behavior is positively linked to job performance, underscoring creativity’s role in organizational success. However, employee well-being and work flexibility did not demonstrate a significant direct effect on job performance. This study employed a sample of full-time and gig workers in Thai digital businesses, which may limit the generalizability of our findings to other industries or sectors. To enhance external validity, future research is recommended, including comparative studies across diverse employment forms and industries. Moreover, the adoption of a mixed-methods approach is encouraged to provide a more comprehensive understanding and broaden the scope of inquiry across multiple national contexts. Our findings underscore the need for policies that promote flexibility, well-being, and innovation to boost job performance. Digital business managers should foster adaptability, creativity, and support for both full-time and gig workers. An inclusive, balanced work environment can enhance performance, innovation, and satisfaction, helping organizations stay competitive in fast-changing markets. This study contributes to digital business research by examining the interplay between employee well-being, work flexibility, and innovative work behavior in determining job performance across different employment types.

1. Introduction

The temporary economy has emerged as a key issue in the process of reviewing modern employment practices. Taylor et al. (2017) identified the temporary economy as comprising platform, on-site delivery, physical and remote delivery, and exchange work. Huws et al. (2016) estimated that approximately 70 million workers worldwide are registered on online labor platforms that facilitate remote work. Heeks (2017) reported that the index of online labor platform usage is increasing at a rate of 26% per year (Kässi & Lehdonvirta, 2018). However, despite the large number of people employed in the temporary economy, there are concerns about its impact on future work. Piecework creates insecurity and changes the standard employment relationship, which should be carefully considered when developing future economic policies and human resource management practices in the temporary economy (Wood et al., 2019). Most previous studies have focused on permanent employees; however, researchers are now increasingly interested in temporary work (Oliveira et al., 2021). Policies that address employee attitudes and behaviors focus on flexibility and developing skills that are in line with current and future changes to enable flexible work in today’s business and social worlds (Brauner et al., 2021). Research in Thailand has found a growing trend of temporary work, with a 170% increase in temporary job applicants, with digital graphic work accounting for a major proportion of this. There has been an increase in online business work, according to research by the National Council for Higher Education, Science, Research, and Innovation Policy in 2021. Changes in the free market economy in Thailand have affected the trend of temporary work, especially part-time work, causing high levels of stress and job uncertainty, which affect employees’ mental and physical health. Learning about employee well-being can help organizations manage and promote employees, reduce emotional exhaustion, and increase their mental energy and self-esteem. This, in turn, affects well-being (Gabriel et al., 2020). Other research also suggests that it is important to create an environment that protects employees’ mental and emotional well-being (Montgomery et al., 2022), leading to long-term job performance and success (Alvarez-Torres & Schiuma, 2022; Li et al., 2021).

Work flexibility affects employees’ well-being by enabling them to meet their own mental and physical health needs and improve their quality of life, both at work. Work flexibility is a key factor in today’s workplace that promotes innovative work behaviors, which are important for organizations that want to lead and grow in today’s competitive business environment (Afrin et al., 2022; Jain, 2022; Muhamad et al., 2023). Work techniques and support are also important factors in promoting creative work behaviors (Koroglu & Ozmen, 2021). Innovative work behaviors affect job performance by enhancing employees’ ability to solve problems and innovate in their work. It makes job performance more creative and effective according to the needs of the organization (Saridakis et al., 2020). Job performance reflects an employee’s ability to achieve organizational goals (Ángeles López-Cabarcos et al., 2022; Deng et al., 2022; Wibowo et al., 2022).

A review of the literature on employee well-being, work flexibility, and innovative work behavior found that job performance is directly affected by employee well-being and work flexibility, which, in turn, affect innovative behavior in the organization. Blau’s (1964) Social Exchange Theory describes the relationship between employees and organizations based on mutual benefit exchange. While a substantial amount of the literature examined employee well-being (EWB), work flexibility (WF), and innovative work behavior (IWB), most studies have focused on permanent employees and neglected the role of temporary employees in the digital economy. Furthermore, limited studies have addressed these variables within Thailand’s cultural context, where trust and personal relationships significantly influence workplace dynamics. The differentiation between permanent and temporary employees in terms of factors affecting IWB in digital business environments remains underexplored. Therefore, this study explicitly aims to compare full-time and gig workers in Thailand’s digital business sector to identify potential differences in how employee well-being and work flexibility influence innovative work behavior and job performance. This comparative approach addresses an important gap in the literature and provides insights into workforce management across different employment types. This study seeks to explore several key research questions. It investigates whether employee well-being positively influences innovative work behavior and whether work flexibility also contributes to enhancing such behavior in digital business contexts. In addition, the study examines the extent to which innovative work behavior impacts job performance among both full-time and gig workers. Beyond these indirect pathways, the study further inquires as to whether employee well-being and work flexibility have direct effects on job performance. These research questions are designed to provide a comprehensive understanding of how well-being, flexibility, and innovation interact to shape performance outcomes in the rapidly evolving digital business environment.

Additionally, Social Exchange Theory (SET) is applied to explain the interplay between EWB and WF, on the one hand, and job performance (JP) and IWB, on the other hand, in diverse employment contexts. This study bridges these gaps by analyzing the effects of EWB and WF on IWB and JP among permanent and temporary employees in the digital sector, offering insights for effective workforce management.

2. Literature Review

2.1. Theoretical Background

Social Exchange Theory (SET), which originated in the 1920s, has significantly influenced various disciplines, including social psychology (Homans, 1958) and sociology (Blau, 1964). This theory explains interpersonal behavior through the exchange of resources and expectations of returns. Homans (1969) further developed SET in the context of stock market studies, highlighting behavior, commitment, trust, fairness, and coalition formation. More recently, SET has been expanded to incorporate the role of emotions in social exchanges, both in controlled environments and real-world settings. From an economic perspective, one study (Zychlinski et al., 2021) has emphasized the relevance of SET in contemporary business operations, with Cropanzano et al. (2017) noting that social exchanges are essential not only within organizations but also in daily life, such as in family and peer relationships. Prior research, including studies by Kmieciak (2020), Koroglu and Ozmen (2021), Liu et al. (2023), and Ridwan Maksum et al. (2020), has applied SET to examine employee behavior and job performance, particularly in relation to innovative work behavior. This study posits that when employees perceive a positive psychological and social environment and are supported by flexible work arrangements, they are more likely to exhibit innovative behavior, ultimately leading to optimal job performance.

2.2. Employee Well-Being and Innovative Work Behavior

Interpersonal relationships within organizations enhance employee well-being by fostering trust and a positive work climate, which, in turn, encourages innovative behavior. Nazir et al. (2019) emphasized that employee well-being promotes innovative work behavior when supported by organizational cultures that encourage information sharing, reward systems, and social networking. Rasool et al. (2021) found that organizational support and well-being reduce negative work environments, whereas self-efficacy boosts both well-being and innovation performance (Singh et al., 2019). Despite a growing interest in this area, Afsar et al. (2020) and Nguyen et al. (2019) noted that there is limited research on enhancing employee’s overall motivation and well-being. Miao and Cao (2019) identified a positive link between well-being and innovative behavior, while AlSuwaidi et al. (2021) highlighted that well-being also motivates pro-environmental actions. Effective management strategies, such as promoting psychological capabilities (Mishra et al., 2019), providing job flexibility and safety during crises (such as COVID-19) (Azizi et al., 2021), and fostering supportive leadership (Cai et al., 2018), are essential for driving innovative behavior and job performance. W. Kim and Park (2017) further stressed the role of employee engagement in enhancing performance. This study proposes that Social Exchange Theory (SET) explains the relationship between employee well-being and innovative work behavior. Therefore, this study aimed to explore this relationship to develop policies and strategies that promote well-being and stimulate innovation in the workplace. Based on this discussion, we propose the following hypothesis:

H1: 

Employee well-being has a positive influence on innovative work behavior.

2.3. Work Flexibility and Innovative Work Behavior

Work flexibility refers to the balance between work and personal life, including flexible working hours, location, and communication methods that are tailored to individual needs. The COVID-19 pandemic has forced many organizations to restructure their arrangements. Allowing employees to choose their work style, location, and communication technologies can alleviate these challenges, because flexible working supports better adaptability. Telework, when supported by proper facilities and effective communication systems, is a viable alternative (Aidla et al., 2022). Work flexibility is closely linked to innovative work behavior, with psychological empowerment and knowledge sharing serving as key enablers (Wen et al., 2021). Ethical leadership and self-esteem positively influence innovation (Yasir et al., 2021). Rafique et al. (2022) found that leader member exchange (LMX) and schedule flexibility significantly impact employee empowerment and innovative behavior. Similarly, knowledge sharing plays a vital role in fostering innovation (Anser et al., 2020). Chua et al. (2022) highlighted the importance of open-plan workspace designs and a robust IT infrastructure for adopting flexible work arrangements. In addition, smart technologies, digital applications, and modern IT systems can enhance workplace management and operations. Aura and Desiana (2023), emphasized that flexible working styles, where employees manage their own time and workplace, reflect adaptive human resource practices. Autonomy promotes psychological empowerment and enhances innovation performance. Based on this discussion, we propose the following hypothesis:

H2: 

Work flexibility has a positive influence on innovative work behavior.

2.4. Innovative Work Behavior and Job Performance

Innovative work behavior and job performance are closely interconnected. Innovation enhances job performance, whereas a strong performance fosters an environment that supports it. Organizations can benefit from promoting both aspects. Alikaj et al. (2020) found that task-oriented leadership sustained performance during crises. Promoting work autonomy and team cohesion also positively affects performance by encouraging innovation through adjusted human resource practices. Shanker et al. (2017) highlighted that innovative behavior mediates the relationship between the organizational climate and performance, with innovation creating a positive atmosphere that boosts performance. M.-S. Kim and Koo (2017) further emphasized that job performance is significantly influenced by innovative behavior. Employee identity also plays a role in stimulating innovation performance (Frare & Beuren, 2021). Based on this discussion, we propose the following hypothesis:

H3: 

Innovative work behavior has a positive influence on job performance.

2.5. Employee Well-Being and Job Performance

Employee well-being is crucial for organizations, as it enhances productivity and positively impacts both individual and organizational performance (Darvishmotevalia & Alib, 2020). Employees with high PsyCap scores are better equipped to cope with challenges and sustain their job performance over time. During the COVID-19 pandemic, Jarosz (2021) found that tools improving well-being helped managers and teams manage stress effectively. Additionally, M. Kim and Kim (2020) confirmed that well-being is linked to higher job performance, highlighting the importance of strong manager–employee relationships. Quade et al. (2020) reported that general creativity is positively associated with all dimensions of job performance, but not directly with well-being. However, Zhang et al. (2020) found that higher creativity correlated with increased well-being and performance. Overall, employee well-being strongly influences job performance, requiring organizational support to foster a work environment that promotes long-term well-being and productivity. Based on this discussion, we propose the following hypothesis:

H4: 

Employee well-being has a positive influence on job performance.

2.6. Work Flexibility and Job Performance

Work flexibility in terms of time and location enhances employee satisfaction, reduces stress, and boosts productivity by enabling better management of personal and work lives. Davidescu et al. (2020) found that flexible working hours and employee development improved job satisfaction and performance. Chatterjee et al. (2022) further noted that remote work flexibility positively impacts organizational performance, as management support makes employees feel valued. Ererdi et al. (2022) highlighted that flexible work arrangements reduce turnover intentions by boosting motivation and minimizing work–family conflicts, leading to greater engagement. Similarly, Halinski and Duxbury (2019) showed that workplace flexibility helps employees manage work–life imbalances. Ray and Pana-Cryan (2021) emphasized that flexibility and employee development are key to sustainable job and organizational performances. Overall, work flexibility significantly affects employees’ mental health and job satisfaction, particularly in remote work contexts, which is crucial for enhancing organizational performance (Chatterjee et al., 2022). Based on this discussion, we propose the following hypothesis:

H5: 

Work flexibility has a positive influence on job performance.

Our theoretical model is illustrated in Figure 1. In addition to using sufficiency logic to conceptualize causal relationships between variables, this study applies necessity logic to identify the “must-have elements” influencing performance through innovative work behaviors. Following Dul’s (2015) recommendation, we propose the hypotheses presented above to conduct our study, focusing on employee well-being, work flexibility, and innovative work behaviors in relation to job performance. Given the study’s focus on both full-time and gig workers, each hypothesis (H1–H5) is examined comparatively across the two groups to determine whether the strength and significance of the proposed relationships differ based on employment type.

3. Methodology

3.1. Measurement

We reviewed the literature to determine this study’s components, working definitions, and variable structures. A questionnaire, comprising four sections covering key variables, was developed. The first measures employee well-being and was adapted from Kuriakose and Sreejesh (2023), Jaiswal et al. (2022), and Pradhan and Hati (2019). Section 2 assesses work flexibility and was adapted from Chatterjee et al. (2022), Hartner-Tiefenthaler et al. (2022), and Aura and Desiana (2023). Section 3 focuses on innovative work behavior and was adapted from Parnitvitidkun et al. (2024). Section 6 evaluates job performance and was adapted from Sanlioz et al. (2023), Santos et al. (2018), and Ochoa Pacheco et al. (2023). A 5-point Likert scale was used (1 = least, 5 = most) (Bell et al., 2022), where higher scores indicate greater awareness and lower scores indicate less awareness. The data collection questionnaire was approved by the Human Research Ethics Exemption Review (Code HE673232) of the Human Research Ethics Committee of Khon Kaen University. Prior to the main data collection stage, a pilot test was conducted with 30 respondents who shared similar characteristics with the target population. The objective was to examine the clarity, reliability, and validity of the questionnaire items. The results indicated satisfactory internal consistency, with Cronbach’s alpha values exceeding the recommended threshold of 0.70 for all constructs (EWB = 0.855; WF = 0.868; IWB = 0.843; JP = 0.890). Based on the feedback from the pilot test, minor wording adjustments were made to improve clarity and contextual relevance before administering the final survey.

3.2. Data Collection

This study was conducted in Thailand with a sample of 201 full-time employees working in digital business system development (Digital Economy, 2023). Participants were male or female, aged 20–60 years, and had at least two years of work experience, ensuring that they could provide relevant insights into employee well-being, work flexibility, innovative work behavior, and performance (Garg & Dhar, 2017). The questionnaire was distributed online, using a web-based platform. Additionally, gig workers from the IT Support Thailand public Facebook group, who occasionally worked in digital business system development, were included. As the exact population size was unknown, Cochran’s (1977) formula was applied, suggesting a sample size of 384 at a 95% confidence level and a 5% error margin. Convenience sampling was also conducted. However, only 199 valid responses were received, likely because of concerns about data security and privacy associated with online surveys. Despite this, the response rate exceeded 50%, which Hoonakker and Carayon (2009) considered acceptable for online surveys and sufficient for data analyses. Therefore, the final sample size was deemed appropriate. The relevant details are presented in

Table 1. Population demographics.

Respondent Profile	Categories	Frequency	Percent
Gender	Male	206	51.50
	Female	194	48.50
Age	>25 years	35	8.75
	25–30 years	112	28.00
	31–40 years	133	33.25
	41–50 years	85	21.25
	51–60 years	27	6.75
	<60 years	8	2.00
Education level	Below bachelor’s degree	37	9.25
	Bachelor’s degree	256	64.00
	Above bachelor’s degree	107	26.75
Nature of work	Digital services	108	27.00
	Software and software services	194	48.50
	Digital content	63	15.75
	Other (please specify)	35	8.75
Employment type	Full-time	201	50.25
	Gig worker	199	49.75
Average monthly Income	>THB 30,000	118	29.50
	THB 30,000–40,000	101	25.25
	THB 40,001–50,000	88	22.00
	THB 50,001–60,000	52	13.00
	THB Over 60,000	41	10.25
Work experience in the digital business sector	1–2 years	128	32.00
	3–4 years	148	37.00
	<5 years	124	31.00

.

3.3. Validity and Reliability Testing

The analysis was conducted in two phases. First, multiple group structural equation modeling (MG-SEM) using GSCA Pro 1.2.1 (a free licensed software, developed by Hwang and available for download from the developer’s official website) was used to examine the variables and generate factor scores (Hwang et al., 2023a; Manosuthi et al., 2021). This approach improves validity. It should be noted that the present study focused solely on examining the direct relationships among employee well-being, work flexibility, innovative work behavior, and job performance, and convergent validity was confirmed with factor loadings > 0.7 (strong) or >0.6 (acceptable), average variance extracted > 0.50, and CR > 0.70 (Hair et al., 2020). According to our results the factor loadings of individual items ranged from 0.500 to 0.773, which are within the acceptable range, indicating that each item adequately reflects its underlying latent construct in accordance with theoretical expectations. Regarding reliability, the Cronbach’s alpha (α) values for all constructs ranged from 0.800 to 0.903, the Rho_A values ranged from 0.816 to 0.907 and the Composite Reliability (CR) values ranged from 0.816 to 0.907, thereby all exceeding the recommended threshold of 0.70, which suggests good internal consistency and stability of the measurement model. At the same time, the AVE values for all constructions ranged from 0.417 to 0.485, which are slightly below the widely accepted threshold of 0.50. However, as recommended by Lam (2012), convergent validity can still be considered acceptable when the Composite Reliability (CR) exceeds 0.60. Additionally, following Bagozzi’s (1993) guidelines, the AVE values were found to be greater than the squared correlations between construct pairs, further supporting the adequacy of convergent validity, despite the marginally low AVE values. Each item was considered essential, as all were theoretically aligned with the research objectives and contributed to fully capturing the dimensions of the studied constructs. To further ensure the robustness of the measurement model, multicollinearity among the independent variables was assessed using the Tolerance values and Variance Inflation Factor (VIF) indices. The results showed that the VIF values ranged from 1.512 to 2.084, all well below the threshold of 10, while Tolerance values ranged from 0.480 to 0.611, exceeding the minimum acceptable threshold of 0.20 (Hair et al., 2020). These results indicate that multicollinearity was not a concern in this study, as shown in

Table 2. Tolerance and VIF of variables.

Variable	Collinearity Statistics
	Tolerance	VIF
EWB	0.483	2.072
FW	0.611	1.512
IWB	0.480	2.084

.

Discriminant validity was assessed using the HTMT (<0.85) (Fornell & Larcker, 1981). To ensure the robustness of the findings and mitigate concerns regarding common method variance (CMV), procedural and statistical remedies were employed. Anonymity and confidentiality were assured during data collection to reduce social desirability bias. In terms of statistical remedies, the discriminant validity was assessed using the HTMT criterion, with all values being below the recommended threshold of 0.85 (Fornell & Larcker, 1981), indicating that CMV was unlikely to bias the results. This supports the robustness of the measurement model and reduces concerns about method bias, as shown in

Table 3. Discriminant validity analysis using Heterotrait Monotrait (HTMT) method.

Constructs	EWB	FW	IWB	JP
EWB
FW	0.765
IWB	0.768	0.725
JP	0.642	0.608	0.823

.

For the FIT model, the ideal value should be close to 1 (Hwang et al., 2023b), with SRMR < 0.08 and GFI > 0.9 (Benitez et al., 2020; Hair et al., 2020; Manosuthi et al., 2021).

4. Results

First, MG-SEM was used to examine all variables and generate component scores (Manosuthi et al., 2021). This study employed MG-SEM using GSCA Pro 1.2.1 to examine the direct relationships among employee well-being, work flexibility, innovative work behavior, and job performance for both full-time and gig workers. The proposed research framework was evaluated using a fully informed estimator and integrated generalized structural component analysis to ensure robust model fit and parameter estimation, in accordance with the guidelines. Finally, the interaction effects among all the variables were assessed using component scores. Convergent validity was used to evaluate the average variance extracted (AVE) of the EWB, WF, IWB, and JP variables for both permanent and temporary employees, with an AVE value of at least 0.50. However, an AVE value of 0.40 was considered acceptable, if the CR was above the level of adequate reliability (0.6) considered acceptable (Lam, 2012; Maruf et al., 2021), as this suggests convergent validity according to the criteria presented for MG-SEM in

Table 4. Verification of the research framework.

Construct Indicator		AVE	CR	wi	CIwi	$\widehat{\mathit{\lambda }}$i	$\mathit{C}\mathit{I}\widehat{\mathit{\lambda }}$i
EWB	I consistently value your work.	0.417	0.865	0.190	[0.1660, 0.2232]	0.715	[0.6470, 0.7620]
	I work Job performance inspires.			0.172	[0.1477, 0.1928]	0.645	[0.5500, 0.7180]
	I am enthusiastic about developing skills.			0.176	[0.1548, 0.2030]	0.662	[0.5860, 0.7260]
	I am involved in teamwork with colleagues to achieve goals.			0.166	[0.1488, 0.1897]	0.625	[0.5620, 0.7030]
	I have a way of building and maintaining good social relationships with people around.			0.142	[0.1186, 0.1577]	0.532	[0.4040, 0.6360]
	I am able to adapt to changes in life on a daily basis.			0.167	[0.1501, 0.1870]	0.629	[0.5010, 0.7070]
	I feel confident in using your expertise and experience in work.			0.184	[0.1653, 0.2150]	0.692	[0.6170, 0.7700]
	I feel happy at work.			0.172	[0.1517, 0.2041]	0.647	[0.5830, 0.6970]
	I am able to manage and control emotional reactions.			0.174	[0.1580, 0.1912]	0.652	[0.5630, 0.7440]
WF	I can work whenever want.	0.469	0.812	0.331	[0.3009, 0.3591]	0.777	[0.7130, 0.8380]
	I can manage working hours to help be more productive.			0.286	[0.2615, 0.3072]	0.672	[0.6060, 0.7440]
	I can set own working style.			0.320	[0.2940, 0.3461]	0.751	[0.6840, 0.8140]
	I can choose own workplace.			0.305	[0.2754, 0.3412]	0.716	[0.6560, 0.7980]
	I can use technology to communicate effectively.			0.198	[0.1624, 0.2383]	0.464	[0.3590, 0.5810]
IWB	I find new ideas for your work to solve problems for I service recipients	0.454	0.892	0.155	[0.1405, 0.1697]	0.703	[0.6520, 0.7560]
	I studied the feasibility of using new methods or solutions to benefit work.			0.155	[0.1396, 0.1713]	0.701	[0.6480, 0.7640]
	I can analyze the real problems of your work.			0.131	[0.1174, 0.1479]	0.596	[0.5150, 0.6760]
	I meet and discuss with colleagues to exchange new ideas for your work.			0.142	[0.1310, 0.1571]	0.642	[0.5740, 0.7210]
	I am proactive in learning and seeking up-to-date knowledge.			0.147	[0.1310, 0.1635]	0.665	[0.5860, 0.7620]
	I collect information or opinions from service recipients and related persons.			0.151	[0.1366, 0.1673]	0.686	[0.6220, 0.7420]
	I have new ideas for providing services.			0.155	[0.1411, 0.1771]	0.702	[0.6410, 0.7640]
	I evaluate the results after implementing new methods or new problem-solving methods			0.156	[0.1424, 0.1723]	0.710	[0.6430, 0.7680]
	I apply new methods of work to solve unresolved problems.			0.146	[0.1317, 0.1642]	0.661	[0.5960, 0.7180]
	I present or recommend to your colleagues to implement the new methods have created.			0.146	[0.1322, 0.1588]	0.662	[0.5920, 0.7300]
JP	I complete the assigned tasks within the specified time frame.	0.485	0.903	0.151	[0.1388, 0.1624]	0.732	[0.6480, 0.8050]
	I can efficiently perform the tasks specified in the job description.			0.158	[0.1396, 0.1763]	0.766	[0.7100, 0.8200]
	I can perform the tasks in accordance with the assigned position, according to standards and with quality.			0.145	[0.1308, 0.1660]	0.703	[0.6450, 0.7720]
	I work plays an important role in the success of the project or assigned task.			0.146	[0.1322, 0.1629]	0.706	[0.6180, 0.7780]
	I can efficiently and successfully complete special assignments.			0.152	[0.1369, 0.1702]	0.735	[0.6440, 0.8110]
	I can efficiently perform and achieve the set goals.			0.149	[0.1355, 0.1649]	0.721	[0.6460, 0.7850]
	I have achieved the quality criteria set for work.			0.159	[0.1411, 0.1748]	0.772	[0.6790, 0.8330]
	I have completed more work than the set target.			0.105	[0.0935, 0.1204]	0.510	[0.4160, 0.6220]
	I have achieved higher quality work than the set standard.			0.116	[0.1028, 0.1282]	0.563	[0.4680, 0.6400]
	I have achieved the results of work that are responsible for.			0.146	[0.1339, 0.1624]	0.708	[0.6370, 0.7830]

Note: W_i = estimated; CI_wi = 95% confidence interval of estimated weights; $\widehat{\lambda }$_i = estimated loadings; CI$\widehat{\lambda }$_i = 95% confidence in the interval of estimated loadings with 100 bootstrap samples; employee well-being = EWB; work flexibility = WF; innovative work behavior = IWB; job performance = JP.

. In addition, to examine the discriminant validity, this study adopted an advanced heterogeneous univariate correlation ratio. This method is popular, because it can estimate the relationship between latent variables without bias, unlike the traditional method (Roemer et al., 2021). The results indicated that all values were lower than the recommended criterion (0.85), confirming the discriminant validity (Roemer et al., 2021). No significant conflicting issues were found, as all VIF values were less than 5 (Belsley et al., 2005). Additional measures were used in this study, as recommended by Hwang et al. (2023a). The results showed that the proposed model is suitable for theoretical data (FIT = 0.870, GFI = 0.951, SRMR = 0.088) (Benitez et al., 2020; Hair et al., 2020; Hwang et al., 2023b; Manosuthi et al., 2021). This met the criteria for our analysis, which focused on assessing the appropriateness of the GFI and SRMR. When N > 100, researchers may choose a GFI cutoff of 0.93 and an SRMR cutoff of 0.08. In this case, no one index is chosen over another, or a combination of indices is used instead of using each index separately. The recommended cutoff values of each index may be used separately to assess the fit of the model, that is, GFI ≥ 0.93 or SRMR ≤ 0.08 indicates acceptable fit (Cho et al., 2020). Therefore, the research model was reliable and valid. We then studied the relationship between the research constructs. Multiple path coefficients showed significant and positive relationships between identified causes and desired outcomes, as shown in

Table 5. Multiple group SEM analysis and hypothesis testing.

Relationship	Path Coefficient		t-Value (One-Tail)		p-Value		Std. Error		95% CI (L, U)		Result
	G1	G2	G1	G2	G1	G2	G1	G2	G1	G2
H1: EWB -> IWB	0.716 *	0.543 *	4.318	5.271	0.000 *	0.000 *	0.083 *	0.089 *	[0.534, 0.832]	[0.359, 0.699]	Supported
H2: WF -> IWB	0.159 *	0.294 *	3.765	5.959	0.000 *	0.000 *	0.080 *	0.094 *	[0.042, 0.355]	[0.145, 0.495]	Supported
H3: IWB -> JP	0.715 *	0.751 *	8.302	7.612	0.000 *	0.000 *	0.202 *	0.170 *	[0.203, 1.006]	[0.327, 1.051]	Supported
H4: EWB -> JP	0.135	0.065	0.065	0.850	0.948	0.395	0.205	0.199	[−0.175, 0.670]	[−0.310, 0.450]	Not Supported
H5: WF -> JP	−0.048	0.003	1.219	0.753	0.223	0.451	0.087	0.144	[−0.233, 0.116]	[−0.180, 0.264]	Not Supported

Note: R² = IWB = 0.649 (G1), 0.614 (G2); JP = 0.700 (G1), 0.648 (G2). * = p < 0.001significance level; employee well-being = EWB; work flexibility = WF; innovative work behavior = IWB; job performance = JP, (G1 = full-time; G2 = gig workers).

. The main concepts for full-time and gig workers are employee well-being, work flexibility, innovative work behaviors, and job performance. The results show a positive influence, as hypothesized in H1, H2, and H3, whereas the results for H4 and H5 show a negative relationship between the identified causes and desired outcomes.

In addition, we examined the relationships between employees’ well-being, work flexibility, innovative work behaviors, and job performance. First, Table 4 shows that employee well-being (EWB) has a positive and statistically significant effect on innovative work behavior (IWB) in both groups (G1: β = 0.716 ***, p < 0.001; G2: β = 0.543 ***, p < 0.001). This indicates that improvements in EWB significantly enhance IWB among both full-time employees and gig workers. Thus, H1 is supported. Second, work flexibility (WF) also exhibits a positive and statistically significant effect on IWB (G1: β = 0.159 ***, p < 0.001; G2: β = 0.294 ***, p < 0.001). This suggests that increased flexibility in working conditions significantly strengthens innovative behavior across both types of employment. Therefore, H2 is supported. Third, the analysis reveals that IWB has a strong and statistically significant effect on job performance (JP) in both groups (G1: β = 0.715 ***, p < 0.001; G2: β = 0.751 ***, p < 0.001). This demonstrates the crucial role of innovative behavior in significantly enhancing employee performance. Thus, H3 is supported.

However, the direct effects of EWB on JP (G1: β = 0.135, p = 0.948; G2: β = 0.065, p = 0.395) and WF on JP (G1: β = −0.048, p = 0.223; G2: β = 0.003, p = 0.451) were not statistically significant. These findings indicate that while EWB and WF contribute indirectly to JP through their positive effects on IWB, they do not have significant direct effects on JP. Thus, H4 and H5 are not supported.

Finally, the model explained 64.9% of the variance in IWB for full-time employees (R² = 0.649) and 61.4% for gig workers (R² = 0.614). For job performance, the explained variance was 70.0% for full-time employees (R² = 0.700) and 64.8% for gig workers (R² = 0.648), indicating substantial explanatory power (Hair et al., 2020).

5. Discussion

Our study on full-time gig workers in digital businesses supports three of our key hypotheses. Our investigation of H1 confirms that employee well-being positively influences innovative work behavior (β = 0.716, 0.543). Full-time workers benefit from job security and career growth, fostering motivation and commitment, whereas gig workers leverage flexibility for creative problem solving. Despite these differences, both groups rely on organizational support and leadership to enhance well-being and innovation (Nazir et al., 2019; Rasool et al., 2021; Singh et al., 2020). Our results regarding H2 reveal that work flexibility positively impacts innovative work behavior (β = 0.190, 0.294). Although full-time workers enjoy stability, their limited flexibility may hinder creativity. By contrast, gig workers benefit from flexible schedules, adapt quickly, and generate new ideas. However, job insecurity can influence morale among gig workers, contrasting with the relatively secure environment of full-time employees. Organizations should balance stability and flexibility to support innovation (Anser et al., 2020; Wen et al., 2021; Yasir et al., 2021). In terms of H3, this study establishes that innovative work behavior enhances job performance (β = 0.715, 0.751). Full-time workers thrive in terms of stability and development, whereas gig workers gain diverse experiences but struggle with job insecurity. Despite these contrasts, both groups ultimately depend on innovation as a critical driver of performance. Organizations must support both groups to drive innovation and long-term performance (Alikaj et al., 2020; Shanker et al., 2017). However, our results for H4 found that employee well-being did not significantly affect job performance (β = 0.135, 0.065). This finding is consistent with the Job Demands–Resources (JD-R) model (Lesener et al., 2020), which posits that job resources such as well-being may not directly improve performance unless they are channeled through motivational processes such as innovative work behavior. Similarly, Social Exchange Theory suggests that employees may not reciprocate organizational support in the form of well-being with improved performance unless they perceive meaningful career prospects or reciprocal benefits (Blau, 1964; Cropanzano et al., 2017).

Similarly, our investigation of H5 indicates that work flexibility does not significantly influence performance, this result further supports the argument that work flexibility primarily enhances performance through indirect pathways (β = −0.048, 0.003). According to prior research, flexibility creates conditions that facilitate adaptability and creativity, which in turn drive performance (Irawan & Sari, 2021; Taibah & Ho, 2023). In rigid organizational contexts, however, these potential benefits may not materialize directly, highlighting the need to view innovative behavior as a mediating mechanism especially in rigid organizational cultures. Overall, organizations must tailor their strategies to support both full-time and gig workers, fostering a culture that balances well-being, flexibility, and innovation for sustainable job performance, while also recognizing the unique needs, strengths, and challenges of each group.

This study employs MG-SEM between employee well-being and work flexibility. The findings indicate that while both factors are crucial for fostering creative work behavior and performance, they do not support a direct positive impact of well-being on job performance. However, employee well-being is essential for creative work behavior, which drives performance (Fakfare et al., 2024). Although work flexibility does not directly enhance job performance, it is a necessary condition for creativity, a key driver of high job performance. This finding underscores the need for a work culture that nurtures innovation. Ultimately, this study revealed that, without creative work behavior, high performance may not materialize, despite work flexibility and well-being. Notably, while flexibility fosters well-being and supports long-term performance in both groups, the mechanisms differ: full-time workers benefit more from structured support and career development, whereas gig workers depend on adaptability and diverse work experiences. Ultimately, organizations must embrace differentiated strategies that acknowledge these group-specific dynamics to sustain innovation and performance.

6. Conclusions

This study examines how organizations can better support employees by adapting strategies to different workforce groups. A comparative analysis was conducted to identify potential differences between full-time and gig workers in Thailand’s digital business sector. The findings reveal that organizations must meet specific conditions to achieve high performance. These results align with the Job Demands–Resources (JD-R) model, highlighting that resources such as well-being and flexibility primarily enhance performance through motivational pathways such as innovative work behavior. Social Exchange Theory further underscores that employees translate organizational support into improved outcomes when they perceive fairness and reciprocity. This study has explicitly identified its findings to demonstrate how they contribute to the existing body of knowledge in the field. By incorporating both full-time and gig workers in the underexplored Thai digital business context, it extends the application of Social Exchange Theory and the Job Demands–Resources model, filling critical gaps in prior research and providing a foundation for future studies.

Companies in the digital business sector can use these findings to improve workplace well-being, increase job flexibility, and promote innovation, leading to sustainable growth and a competitive edge. By acknowledging the different needs of full-time and gig workers, organizations can design inclusive strategies that foster both stability and adaptability, thereby sustaining innovation and long-term competitiveness in digital business contexts.

6.1. Theoretical Contribution

This study contributes to the theoretical understanding of employee well-being, work flexibility, and innovative work behavior in digital businesses. It provides empirical support for the hypothesis that employee well-being positively influences innovative work behavior (H1), highlighting the critical role of well-being factors such as physical and mental health, a supportive work environment, and work–life balance. These findings align with those of previous studies, reinforcing the importance of fostering an organizational culture that prioritizes employee well-being to enhance creativity and innovation (Nazir et al., 2019; Rasool et al., 2021). Additionally, the study confirms that work flexibility positively impacts innovative work behavior (H2), demonstrating that flexibility in time and location can stimulate creativity, particularly among casual employees. This supports research in this area by emphasizing the necessity of flexible work arrangements to foster creative thinking and problem-solving skills (Wen et al., 2021; Yasir et al., 2021). Furthermore, the study reveals that innovative work behavior significantly influences job performance (H3), with both permanent and casual employees benefiting from innovative practices. This finding aligns with those of other studies, underscoring the importance of promoting innovative behavior to enhance overall job performance (Alikaj et al., 2020). However, the study also found that employee well-being did not significantly influence job performance (H4) and that work flexibility did not significantly influence work performance (H5), challenging the conventional wisdom that well-being and flexibility are directly correlated with performance. These insights suggest that while well-being and flexibility are crucial for fostering innovative work behavior, they do not directly translate to improved job performance without the mediation of creative behaviors (Shanker et al., 2017).

While addressing critical gaps in the literature, this study also incorporates both full-time and gig workers, providing a comparative perspective within the underexplored Thai context. By applying multiple group structural equation modeling (MG-SEM), it advances the application of Social Exchange Theory and the Job Demands–Resources model, revealing how employee well-being and work flexibility influence job performance indirectly through innovative work behavior.

6.2. Practical Implications

The practical implications of this study are multifaceted and provide valuable insights for managers and policymakers in digital business. Organizations promoting employee well-being should invest in comprehensive well-being programs that address both physical and mental health, provide a supportive work environment, and promote work–life balance. This investment is essential for enhancing innovative work behavior, as demonstrated by the positive correlation between well-being and creativity. Employers should consider implementing wellness initiatives, offering mental health support, and creating a culture that values employee well-being. Given the positive impact of work flexibility on innovative behavior, organizations should explore flexible work arrangements, such as remote work options and flexible schedules. These arrangements can help casual employees leverage their adaptability and creativity while also providing permanent employees with the flexibility that is needed to stimulate innovative thinking. Organizations should develop policies that support flexible work practices and provide the necessary tools and resources for effective teleworking. They should also create an environment that encourages creativity and idea generation to enhance job performance through innovative work behavior. This can be achieved by providing opportunities for skill development, fostering open communication, and recognizing and rewarding innovative contributions. Leaders should focus on creating a supportive and inclusive culture that empowers employees to experiment and be innovative.

In practice, our findings provide actionable insights for managers and policymakers in the tourism and hospitality sector. Organizations can design inclusive policies that balance stability for full-time employees with flexibility for gig workers, thereby fostering adaptability and creativity. Such strategies not only enhance innovative work behavior but also sustain long-term job performance and competitiveness in fast-changing digital markets. These contributions highlight the importance of tailoring workforce strategies to different employment types to achieve sustainable organizational growth.

6.3. Limitations and Future Studies

This study has several limitations. First, the study only employed quantitative methods, utilizing structural equation modeling for the analysis. The lack of qualitative approaches limits the depth of insights that can be obtained. Incorporating qualitative methods, such as interviews or focus groups, could provide a more nuanced understanding and actionable insights into the factors influencing employee well-being, work flexibility, and innovative behavior. Employing fuzzy set Qualitative Comparative Analysis (fsQCA) (Manosuthi et al., 2021) could also be advantageous, as it provides both causal and non-causal explanations, enhancing theory development and testing. Second, while the study identifies significant relationships between employee well-being, work flexibility, and innovative work behavior, it does not delve deeply into the specific aspects of these variables. A more detailed examination of elements, such as different types of flexible work arrangements, specific well-being initiatives, and their respective impacts on innovation, would provide a richer understanding. Additionally, exploring the role of organizational culture and leadership styles in these dynamics could offer valuable insights. Finally, this study focuses primarily on employees within the digital business sector in a specific geographical region. This limits the generalizability of our findings. Broadening the research to include more diverse industries and regions would enhance the applicability of the results. Furthermore, this study’s cross-sectional design captures data at a single point in time, restricting its ability to infer causality.

Future studies should address these limitations by using different qualitative methods. Qualitative approaches, such as interviews and focus groups, can provide deeper insights into the factors that influence employee well-being and work flexibility. Employing techniques such as fsQCA can also enhance the robustness of theoretical models. An in-depth examination of the specific components of work flexibility and well-being initiatives is needed to better understand their individual and combined effects on innovative behavior and job performance. Exploring the impacts of various leadership styles and organizational cultures on these relationships would further enrich the findings. Additionally, expanding the scope of this study to include a wider range of industries and geographical regions would improve the generalizability of the results. Longitudinal studies should be conducted to establish causal relationships and observe changes over time. Research focusing on different generational segments and their perceptions of work flexibility and well-being could also provide valuable insights to enhance the overall understanding of these dynamics in the workplace.