Overby (2008, 2012) defines this transition from a physical process to a virtual process as process virtualization. As an emerging phenomenon in information systems, Overby (2008, 2012) proposed PVT to explain the process virtualization. The focus of PVT is to understand how and to what extent a process that operates in a face-to-face physical environment can be implemented in a contactless virtual environment (Alarabiat et al., 2023). Overby (2008) defined the expected effect or actual result obtained by promoting process virtualization as process virtualizability and presented a dependent variable (process virtualizability) to explain and predict how suitable a process is to be performed in a virtual environment.

Overby (2008) first proposed that the PVT includes two groups of independent variables. The first is the psychological requirements group of the business process, and the second is the IT characteristic group, which affects process virtualizability. The psychological requirements group has sensory, relationship, synchronism, and identification and control requirements. Sensory requirements refer to the fact that participants in a virtual process need to be able to enjoy the whole sensory experience in the traditional physical environment. Synchronism requirements refer to the requirements that participants in a virtual process need to communicate with other participants with the lowest delay. Relationship requirements refer to the perceived needs of participants in a virtual process to interact, socialize, and establish friendships with other process participants. Identification and control requirements refer to the requirements of participants in a virtual process to identify other participants in the same process. These psychological requirements negatively impact process virtualizability. The IT characteristic group has monitoring capability, reach, and richness. Monitoring Capability refers to the ability of an information system to provide high-quality information. Reach refers to the ability provided by information systems to break through time and distance constraints. Richness refers to the ability of an information system to provide rich and diverse information. The IT characteristics can mitigate the negative impact of process requirements on process virtualizability.

In subsequent studies, the definition of process virtualizability is slightly different depending on the researcher. Overby and Konsynski (2010) defined process virtualizability as the extent to which they can be easily performed between people or between people and objects without physical interaction. Balci and Rosenkranz (2014) defines process virtualizability as the degree to which it is possible to succeed in converting a process from a physical environment to a virtual environment. Alarabiat et al. (2023) defines process virtualizability as the adaptability of transferring from the physical environment to the virtual environment at any time.

Furthermore, more researchers demonstrated the advancement of PVT application in different scholarly contexts (Overby and Konsynski, 2010; Barth and Veit, 2011; Balci and Rosenkranz, 2014; Mburu and Oboko, 2018; Agrawal et al., 2020; Alarabiat et al., 2023). Czarnecki et al. (2010) propose a framework for planning the processes virtualization intended to assist the decision-maker in prioritizing the processes to be virtualized: highlighting their demand for changes at different levels of the organization. Based on the theoretical foundation of PVT, Graupner and Maedche (2015) conducted a quantitative study to identify the factors influencing intended digital process use in retail banking. The results indicate that relationship, sensory, and control requirements inhibit intended digital process use in retail banking. The results not only contribute to a better understanding of the factors that influence intended digital process use in retail banking but also provide an empirical examination of a theory that has been largely untested. Jasper (2015, 2019) integrates developed new product features into a joint virtual reality (VR)-based environment for measuring consumer preferences during product innovation development. Virtualization of the new product demonstration process offers the potential for early customer integration in the new product development process. To assess whether IT-enabled process virtualization capabilities impact organizational Green IT initiatives, Tomás et al. (2018) proposed a conceptual model that combines three theories: technology-organization-environment framework, PVT, and diffusion of innovation theory. PVT can explain whether processes are suitable for migration into virtual environments, such as those enabled by information technology.

In this study, we hypothesize that psychological requirements (Sensory requirements, Synchronism requirements, and Relationship requirements) negatively impact business process virtualizability based on the basic PVT theoretical model. The following is the basis for our derivation of this hypothesis.

Sensory requirements are that organization members must be able to enjoy the whole sensory experience of traditional work processes (e.g., seeing and hearing other colleagues or leaders in the workplace and having a feeling of being in the office). In the virtual environment, the communication and interaction between people rely on IT technology and information systems, so it is impossible to fully experience the visual and olfactory sensory experience in the virtual environment. Therefore, if the sensory demands of the work to be performed are high, process virtualizability may be low. According to The Law of Mehrabian, if the importance of linguistic elements in communication is 7%, non-verbal elements such as gestures, tone, and intonation is 93% (Mehrabian, 1971). While group members can converse in a near-face-to-face feel via video conferencing, there are limits to capturing each other’s facial expressions, speech, and attitude changes in detail. Therefore, if members of the organization have high sensory demands on the work handled in the virtual environment, it may lead to poor communication and poor work processing (Overby and Konsynski, 2010; Balci and Rosenkranz, 2014). According to related research, higher sensory demands lead to lower process virtuality (Overby, 2008, 2012; Overby and Konsynski, 2010; Balci and Rosenkranz, 2014; Graupner and Maedche, 2015; Agrawal et al., 2020; Alarabiat et al., 2023). The present study builds on the hypothesis that sensory demands impair process virtuality based on previous studies on sensory demands.

Relationship requirements refer to the degree to which process participants need to interact socially with others involved in the process to meet professional needs (Overby, 2008; Overby and Konsynski, 2010; Balci and Rosenkranz, 2014; Graupner and Maedche, 2015; Alarabiat et al., 2023). This social interaction can foster friendships and knowledge sharing among organizational members, enhancing team creativity (Wang et al., 2021). The formation and development of relationships among organizational members or supervisors in the workplace is the focus of many scholars’ research. Fay and Kline (2012) believe that organizational members with high-quality friendships have higher organizational commitment and identification. In addition, high-quality supervisor relationships create cognitive and social resources that can satisfy the members’ sense of belonging (Bono and Yoon, 2012). However, telecommuting has changed how people communicate, damaging the quality of relationships between colleagues and supervisors. The reason is that this relationship is weaker in virtual environments than in face-to-face physical environments (Mesch and Talmud, 2006; Lima et al., 2017). According to related studies, higher relational needs lead to lower process virtualizability (Overby, 2008, 2012; Overby and Konsynski, 2010; Barth and Veit, 2011; Balci and Rosenkranz, 2014; Graupner and Maedche, 2015; Mburu and Oboko, 2018; Alarabiat et al., 2023). For example, Barth and Veit (2011) confirmed in a study on e-government services that the more demanding users need to establish relationships with responsible public officials, the lower the acceptance of online services. Graupner and Maedche (2015) found that bank customers’ intention to use online banking is lower if they have higher intentions to build relationships and develop friendships with bank staff. Alarabiat et al. (2023) confirmed in their research on online learning that if students need to interact and build relationships with their classmates or educators in a physical space, students are less satisfied with online learning platforms and decrease their intent to continue using online learning. Based on these results, an information system-based virtual environment may not be suitable if people involved in performing work need to build friendships and trust and accumulate the knowledge necessary for performing work. So, we propose the following hypothesis.

Synchronism requirements are the level at which members of an organization can communicate immediately with other people involved in a business process that needs to be processed immediately with minimal delay (Overby, 2008; Overby and Konsynski, 2010; Graupner and Maedche, 2015; Alarabiat et al., 2023). Teamwork is playing an increasingly important role in many jobs, and teamwork quality has always been a critical factor in project success and organizational performance (Gander et al., 2020). A high-quality team needs synchronous and efficient communication to achieve consensus within the team. According to the Media Synchronicity Theory proposed by Dennis et al. (2008), when two or more people are in the same business process, there is synchronization between them. Media Synchronicity Theory has verified five factors that affect the level of synchronization, including Transmission speed, Parallel processing, Symbol diversity, Rehearsal, and Re-processing. Communication among teleworkers using corporate social media can be broken down into two processes: conveyance and convergence. The conveyance process means a large amount of information is exchanged between workers, and the convergence process involves consensus on these messages. Higher synchrony among team members means smoother social interaction, which is conducive to improving the cohesive social capital of employees, all of which positively impact work efficiency (Jong et al., 2021). In the face-to-face physical space, organization members can see colleagues and communicate anytime, quickly exchanging the necessary information to meet people’s synchronization needs.

In contrast, in the environment of process virtualization, the organization’s members need to meet the synchronization requirements through means based on IT communication systems such as enterprise SNS. However, compared to a face-to-face physical environment, the synchronization process facilitated by IT may be subject to limitations due to factors that impact the level of synchronization, such as transmission speed, parallel processing capabilities, diversity of symbols, and opportunities for rehearsal. The presence of these limitations may impede the synchronization process. Consequently, as the level of synchronization required increases, the feasibility of process virtualization decreases. Overby and Konsynski (2010) showed that synchronization demand had a negative effect on process virtualizable demand in online auction research. Balci and Rosenkranz (2014) identified the synchronization requirements as the biggest impediment to process virtuality in Germany’s online banking study. Agrawal et al. (2020) verified that converting the working process to a virtual environment would be difficult as the synchronization requests of organizational members increased. In this study, we derived the following research hypothesis by referring to previous research results and the characteristics of telecommuting: synchronism requirements hurt the virtuality of the process.

Information overload refers to a situation where people receive more information than they can process and utilize the information, leading to anxiety or a negative sense of failure (Granero-Gallegos et al., 2021). Information overload affects information-intensive industries across society, including government, education, family life, online consumption, and citizenship (Bawden and Robinson, 2009; Huang and Wang, 2022). In the digital information age, individuals and organizations are overwhelmed by too much information. Because there is too much information to read and process, which reduces productivity and performance, and is not conducive to business innovation (Bawden and Robinson, 2009; Jackson and Farzaneh, 2012). Based on the media richness theory, Hwang et al. (2020) demonstrated that the information overload caused by instant messaging affects civil servants’ work engagement. Yu et al. (2018) explained how excessive use of social media at work affects individual work performance through an extended stressor–strain–outcome research model. Its results prove that information overload is a vital stress factor affecting social media burnout, and social media burnout further significantly reduces individual job performance.

In our study, organization members worked from home through communication technology-based information systems, and business processes eliminated physical interaction, which led to a substantial increase in the information exchanged online. In this virtualized work environment, when teleworkers search for information through collaborative software to process business and make decisions, individuals can obtain deeper and broader information than necessary (Landers and Schmidt, 2016). This process too much information requires individuals to spend a lot of time and effort to extract the information that is useful to them. Thus, we speculate that information overload prevents the business process virtualizability.

Communication overload refers to the network’s communication demand exceeding the individual’s communication ability (Barrett et al., 2022). Too many emails, phone calls, text messages, instant messages, and social media notifications can interfere with people’s everyday life and work (Lin et al., 2021). Through collaboration software like DingTalk, team members in the virtual workplace will inevitably receive more notifications. Organization members frequently interrupt their work due to various notifications, making it difficult for them to concentrate, which can cause organization members to become tired or anxious (Granero-Gallegos et al., 2021). Not only that, but it forces organization members to multitask because the effort is spent contacting people via different methods while working, which can be distracting (Reinecke et al., 2017). The distractions threaten the work output of members of the organization. According to a University of California Irvine study, “it takes an average of 23 min and 15 s to get back to the task.” Even if members of an organization eliminate distractions, they will not be as productive as they were before the distractions (Mark et al., 2008). Therefore, we deduce the following hypothesis.

We review the PVT literature and propose a model suitable for studying business processes. To understand the relationship between several variables, we proposed and examined several hypotheses in Figure 1.

This study selected the fully mandatory teleworkers during the COVID-19 pandemic as study subjects. To prevent the spread of the epidemic, China has effectively implemented government-led regional isolation and social distancing policies. Therefore, many members of the organization began to use collaboration software (DingTalk) to work remotely, which provided an opportunity for our research. Unlike previous hybrid telecommuting, the physical interaction between teleworkers during the epidemic is wholly eliminated, which benefits this study of business process virtualization.

Due to the restrictions of the quarantine policy, starting from October 2021, 1 month period, we conducted a questionnaire survey online using a random sampling method. The questionnaire survey was managed through China’s “Wen Juan Wang” online survey platform. A total of 380 samples were collected from teleworkers. After excluding 37 invalid samples, the remaining 343 samples were used for data analysis. Those who participated in the survey were telecommuters using Alibaba’s DingTalk system. They are mainly engaged in e-commerce, dealing in food, clothing, and so on related to infants and young children. According to “The 47th China Statistical Report on Internet Development 2021” published by the China Internet Network Information Center (CNNIC), there were approximately 5 million teleworkers in China at the time of this survey. Based on the calculation of the number of respondents and population size, the margin of error in this paper is 5.29%.

Table 1 reflects that males accounted for 42.86%, and females accounted for 57.14%. People in their 20 s accounted for the most, accounting for 63.55%, followed by people in their 30 s, accounting for 24.78%. Among the number of employees working in the company, 58.35% had less than 50 employees, 25.19% had 50–99 employees, and 16.47% had more than 100 employees.

The study used a five-point Likert scale (1 = very unimportant to 5 = very important) to record responses. First, we conducted a test before the formal investigation to verify the questionnaire’s feasibility. After analyzing and discussing the test results with relevant experts, we modified the questions, wording, and ambiguity of the questionnaire and finally formed a 21-item questionnaire. The questionnaire was distributed as a formal questionnaire (Table 2). Sensory, relationship, synchronism requirements, and business process virtualizability were measured using 14 items from Overby (2008) and Alarabiat et al. (2023). Information overload and communication overload were measured using seven items from Granero-Gallegos et al. (2021).

To ensure the robustness and validity of the theoretical hypotheses proposed in this study, a structural equation modeling (SEM) approach was employed to analyze the data. This method allows for the examination of complex relationships among multiple variables, which is particularly useful in light of the multiple factors that may influence the model under examination.

To facilitate this analysis, the SmartPLS 4 program was utilized. This software is widely used in various fields and employs partial least squares (PLS) SEM to analyze structural equations (Ali Qalati et al., 2020). The program is known for its comprehensiveness, with an intuitive graphical interface that allows for the observation of interrelationships among multiple variables (Hair et al., 2011, 2019).

Furthermore, the measurement model and structural model were evaluated using an improved metric proposed by Hair et al. (2011). This new metric not only assesses internal consistency and discriminative validity but also effectively validates the predictive ability of the structural model (Sarstedt et al., 2022).

The present study employs SEM and PLS techniques as outlined in Sarstedt et al. (2022) to examine the measurement models. The results, presented in Table 2, demonstrate that the constructs possess high levels of reliability, as evidenced by the Cronbach’s Alpha and rho_A values, which are greater than or equal to 0.7. Additionally, the Mean-Variance Extracted values (AVE) are greater than or equal to 0.5, indicating strong convergent validity (Sarstedt et al., 2022). Furthermore, in accordance with the guidelines provided by Franke and Sarstedt (2019), this study evaluates the discriminant validity of the constructs by utilizing the heterotrait-monotrait ratio (HTMT) method, as presented in Table 3. The results indicate that all HTMT critical values are less than 0.9, indicating that the constructs possess strong discriminant validity.

The value of R² represents the explanatory power of the PLS structural model. The explained variance is shown in Figure 2. All independent variables of this study can explain 17.6% of business process virtuality. The value of R² ranges from 0 to 1, and the larger the value, the stronger the explanatory ability of the structural model. According to Raithel et al. (2012)’s claim, the acceptable R² value is above 0.1. The coefficients for the structural relationships among all variables were obtained by computing a series of regression equations. In addition, we increased the value of Q² (Table 2) in this study to evaluate the prediction accuracy of the PLS path model. A higher Q² value means higher prediction accuracy (Pangarso et al., 2022). Before assessing structural relationships, collinearity must be checked to ensure that it does not affect the results obtained from the regression equation. The collinearity of all formed indicator sets can be assessed by calculating the variance inflation factor (VIF).

Hair et al. (2019) as shown in Table 4, the VIF values of this survey are all less than 3, which proves that there are no collinearity issues. This study calculated the Model fit of the model using SRMR (Hair et al., 2021). Table 5 shows that the model structure is a fit model with the SRMR, d_ULS and d_G values already at the 95% confidence interval. This study uses SmartPLS 4 to calculate the path coefficients and test the hypotheses. Table 6 shows the significant results using normalized path coefficients (β-values) and t-values of path coefficients. Among them, H1 (β = −0.218, t = 4.177), H3 (β = −0.172, t = 3.852), and H5 (β = −0.240, t = 4.691) passed the significance test. The results can be interpreted to conclude that sensory requirements, synchronism requirements, and communication overload significantly negatively impact process virtualizability. However, the path coefficient of H2 was −0.039 (t = 0.666), and H4 was −0.085 (t = 1.319), which proves that relationship requirements and information overload are not the influencing factors that hinder business process virtualization.

This study applies the PVT to telecommuting and contributes to expanding the application. Moreover, we proved that some variables in the prior research are no longer applicable to business processes, providing new evidence for business-related researchers to study business processes. The core contribution of this study is that we reveal the harmful effects of information systems on business processes that have been neglected in previous studies based on PVT and effectively enhance the explainability of PVT on business processes.

Additionally, this research provides a timely contribution. Organization members have been working remotely for extended periods due to COVID-19. Unlike previous occasional remote or hybrid telecommuting, mandatory isolation policies keep members of the organization in a virtual space with no physical interaction, providing a better environment for virtualizing the research process. Excluding the influence of other factors provides a better look at what hinders the migration of business processes from physical to virtual space. The findings could help future researchers study how organizations implement telecommuting.

This study reveals the factors that hinder the business process virtualization and provides a basis for enterprises to implement teleworking. To promote enterprises’ digital transformation and increase business process virtualization. Governments and businesses can develop guidance and policies on using information systems and clarify when information is exchanged between team members. Find a balance that ensures efficient collaboration without letting too much communication disrupt their work. Managers can control information access and distribution by integrating knowledge exchange processes and the social activities of teleworkers in virtual environments. On an individual level, teleworkers can create an office-like environment to satisfy their senses. Schedule management through collaboration software, prioritize communication with teams, and reduce distractions caused by notification reminders.

Although our research reveals essential discoveries, there are also limitations. Our research only collected data on the use of DingTalk for telecommuting. DingTalk is the most used collaboration software among Chinese enterprises, so a single type of collaboration software limits the applicability of our results to different information systems. Future research can adopt a cross-country, cross-cultural approach to study business process virtualization based on different information systems. Improve the accuracy of research results by expanding the scope of the survey and increasing the sample size.