Original Article
Mobile Group Address and Knowledge Sharing: Evidence from Activity-Based Working in a Japanese Firm
2026 年 5 巻 2 号 論文ID: oa.25-013
詳細
2026 年 5 巻 2 号 論文ID: oa.25-013
This study examines how office space and its utilisation policies influence knowledge sharing among employees in a Japanese company, focusing on the unique operational model of introducing mobile Group addresses, which involves forcibly changing the location of Group addresses in activity-based working (ABW). Group addressing is a unique approach commonly seen in Japanese companies that have adopted open-plan office layouts, allowing employees to freely select their seats within a broadly designated workspace assigned to their department (group). A new office policy was implemented involving the introduction of a mobile Group address for 587 employees at the headquarters of a large Japanese firm (Company A). Two questionnaire surveys were conducted before and after the introduction of the new policy, and responses were obtained at both time points from 224 employees. Statistical analyses using t-tests and paired t-tests were employed to test the hypotheses on knowledge sharing. The frequency of inter-group knowledge sharing during a Group address in the ABW system varied depending on the use of communication spaces, which are open areas that can be used without prior booking for meetings or discussions involving a few people. While the introduction of a mobile Group address did have a measurable effect on inter-group knowledge sharing, the effect was limited and dependent on prior communication space utilisation patterns.
activity-based working
EGKSinter-group knowledge sharing
HRhuman resources
IGKSintra-group knowledge sharing
This study examines the effect of office space and its utilisation policy on knowledge sharing among employees in Japanese companies. We focus on the influence of introducing mobile Group addresses in ABW—an office environment that promotes knowledge sharing and is idiosyncratic to Japanese companies. Specifically, we clarify that employees working in Group addresses in ABW differ in their degree of knowledge sharing depending on the frequency of their use of communication spaces within the office. We further examine the effect of introducing a ‘mobile’ Group address in the ABW system, which involves forcibly changing the location of the Group address, on employees’ knowledge sharing. Group addressing discussed in this study is a unique approach commonly seen in Japanese companies that have adopted open-plan office layouts, allowing employees to freely select their seats within a broadly designated workspace assigned to their department (group).
Introducing flexible work arrangements while preventing workplace isolation and mental health issues has come to be recognised as a key challenge in occupational health (Kawakami et al. 2025). However, recent research suggests that initiatives designed to enable flexible work arrangements may increase employees’ feelings of loneliness and hinder their willingness to share knowledge (Cheng et al. 2023). Examining the impact of organisational policies in the design of workplace environments that support flexible work arrangements, facilitate knowledge sharing, and maintain employee connections is likely to contribute significantly to industrial health practice. However, to date, limited research exists on the relationship between the introduction of office utilisation policies that enable flexible work arrangements within the office and employee knowledge sharing. Therefore, in this study, although somewhat indirectly, we focus on knowledge sharing as a variable that captures employee collaboration in the workplace and examine the impact of implementing office usage policies on this aspect.
The organisational factors that promote knowledge sharing include flexible work arrangements (Cheng et al. 2023), human resource policies (Andreeva and Sergeeva 2016), leadership (Morinaga et al. 2023; Xue, Bradley and, Liang 2011) and organisational culture (Intezari, Taskin, and Pauleen 2017; Schepers and van den Berg 2007). However, the effect of office space design and utilisation policy has not been sufficiently examined.
ABW-type offices, where employees can choose where and how to work according to the characteristics of their tasks (Engelen et al. 2019), are an evolution of open-plan offices. They overcome the flaws of cell-type offices, such as difficulties in collaboration within the organisation, and promote knowledge sharing and creativity among employees through the use of communal spaces (de Been, Beijer, and den Hollander 2015). Such ABW-style offices have been considered suitable for employees who are expected to demonstrate creative performance through collaboration with others (de Been, Beijer and den Hollander 2015). A study by Danielsson and Bodin (2008) reported that approximately 67% of respondents worked in open-plan offices, whereas employees working in ABW-style offices—a variant of this model—accounted for approximately 12%. At the time of this study, the proportion of people working in ABW-style offices was by no means high. In particular, technical professions tended to utilise cubicle-style offices. In occupations or industries with low levels of interdependence, the disadvantages of open-plan offices—such as noise concerns and insufficient privacy—are emphasised. Hoendervanger et al. (2022) reported that when employees are engaged in highly concentrated individual tasks, enclosed workspaces foster a greater sense of fit among employees, compared to open workspaces. However, in ABW-style offices, spaces for concentration and communication are separate, allowing employees to move between them as required. This approach can simultaneously address both the challenges of collaboration in cubicle-style offices and noise issues and privacy concerns associated with open-plan offices (de Been, Beijer and den Hollander 2015). With their introduction in Japan, academic evidence of their effectiveness is starting to accumulate (Tanaka et al. 2023; Tokumura et al. 2023). Although transitioning to an ABW-type office improves productivity (Tanaka et al. 2023), the effect of these offices on knowledge sharing is poorly understood. Furthermore, the impact of using communication spaces—where discussions with other groups or informal meetings can be held without prior reservation—on knowledge sharing needs to be explored. In addition, examining the effect of office space utilisation on knowledge sharing is necessary. The mere introduction of ABW-type offices does not necessarily lead to collaborative office utilisation by employees (Hoendervanger et al. 2016; Irving, Ayoko and Ashkanasy 2020). The freedom of movement and in-space utilisation in open-plan offices suggest a differentiated effect of space on individual employees. In ABW-type offices, the frequency of use of spaces intended for interaction often varies among employees, and these differences may affect the degree of interaction with others as well as the degree of knowledge sharing. However, there has been insufficient research on the relationship between office space utilisation and knowledge sharing. As a result, this may be the reason why consistent results on the relationship between ABW-type offices and employee human relations or communication have not been found across studies (Marzban et al. 2023).
Moreover, a detailed discussion on the effect of changes in office space based on their historical context is necessary. In both open-plan and ABW-type offices (the latter being an evolution of the former), knowledge sharing increases through greater chance encounters (Allen 1977; Allen and Gerstberger 1973). This effect is expected to be weaker in Japanese (vs. Western) companies, which traditionally follow a taiko-shimagata (‘taiko island-type’) office layout. In this widely adopted layout, employees gather in groups based on their affiliations within an open office (Naka 2019). Notably, opportunities for intra-group interaction and communication were already abundant in Japanese companies that had adopted the ‘taiko island-type’ office layout.
The introduction of ABW-type offices in Japanese companies is expected to promote knowledge sharing by increasing opportunities for interaction with employees outside one’s team. However, there have been concerns that the effect on IGKS may be limited or negative. In fact, Tokumura et al. (2023) found that although such offices increased communication with employees outside one’s group, they decreased communication within groups. This evidence suggests that ABW-type offices may not be as effective in Japan as they are in Western companies at promoting organisational collaboration, prompting Japanese companies to introduce ‘group-address ABW’. In this method, employees freely select their seats within a broadly designated workspace assigned to their department (Kokuyo Furniture Co., Ltd. 2022), which allows both IGKS and the freedom to move seats to other workspaces as needed. This hybrid layout can increase inter- and intra-group collaboration and communication. It is an effective management strategy that modifies the Western ABW-type settings and applies them to a Japanese context.
To examine the effectiveness of introducing group address-based ABW in Japanese companies, we separately consider knowledge sharing among employees within the organisational unit and outside it. We focus on the differences in the utilisation of communication spaces by employees working in group-address ABW and clarify the effect of these differences on the degree of IGKS through data analysis.
We further examine the effects of experimentally introducing ‘mobile Group address’ into this system. Here, ‘mobile Group address’ refers to an office-utilisation method in which the organisation uniformly changes the designated Group-address locations of employees and changes the departments adjacent to the Group address for a certain period. In Group-address ABW, the departments adjacent to one’s department Group address are always the same departments. As a result, while close interaction is formed with adjacent groups, interaction with distant groups is difficult. Mobile Group address can promote knowledge sharing by intentionally creating opportunities to interact with different groups, thereby facilitating long-overdue interaction and communication. Examining the effects of introducing mobile Group addresses in group-address ABW will provide valuable insights into office spaces that promote knowledge sharing at the organisational level and the policies for utilising such spaces.
Knowledge sharing is defined as the provision or reception of task information, know-how and feedback related to products or procedures (Cummings 2004). It positively influences corporate performance through intellectual capital (Wang, Wang, and Liang 2014), financial performance (Son, Phong, and Loan 2020) and corporate innovation capabilities (Lin 2007), and is a driving force of innovative outcomes within organisations. Individual factors such as high educational attainment and extensive work experience (Constant, Kiesler, and Sproull 1994); organisational-level factors such as a cooperative organisational culture (Schepers and van den Berg 2007) and human resource management practices that enhance employee motivation (Andreeva and Sergeeva 2016); and empowerment-style (Srivastava, Bartol, and Locke 2006), ethical (Bavik et al. 2018) and inclusive (Morinaga et al. 2023) leadership especially promote knowledge sharing.
Non-territorial open offices, including ABW-type offices that promote serendipitous encounters and information exchange, are known to contribute to knowledge sharing and creativity (Allen 1977; Allen and Gerstberger 1973). However, Oldham and Brass’s study (1979) states evidence to the contrary. This may be because the effect of environmental design, such as office design, is influenced by context and varies by employee. In particular, office environments are not always used as intended by office designers or managers. For example, even when a collaborative office environment is provided, some employees avoid contact with others and try to use the office in ways that minimise interaction (Irving, Ayoko, and Ashkanasy 2020). Even when working in the same ABW-type office, the amount of communication within the office and the frequency of knowledge sharing are likely to vary by employees’ working space preferences.
Similar phenomena are likely to occur in group-address ABW. That is, even among employees working in group-address ABW offices, the degree of knowledge sharing is expected to vary by how spaces are utilised. Specifically, the higher the degree of communication space utilisation, the higher the frequency of EGKS and vice versa. Notably, in group-address ABW, IGKS occurs within the Group address. Accordingly, differences in the degree of knowledge sharing within the group due to differences in the degree of communication space utilisation are not expected. Therefore, the following hypothesis is set.
Hypothesis 1: Among employees working in a group-address ABW office layout, the degree of inter-knowledge sharing between groups will differ depending on the utilisation of communication spaces. In other words, employees who utilise communication spaces more frequently prior to the intervention are likely to engage in EGKS more frequently before the intervention than those who do not.
2-2. Spatial proximity and knowledge sharingThe mobile Group address introduced in this study can be understood as a practice in which organisations intentionally redesign knowledge-sharing networks within the office by changing spatial proximity. In this regard, Salazar Miranda and Claudel (2021) demonstrate that spatial proximity in offices promotes collaboration among researchers and that the effects of proximity diminish over a period of about 5 years. Thus, introducing a mobile Group address in an ABW office, where the Group address was previously fixed, and changing the location of the Group address will introduce changes to the spatial proximity between employees and promote chance encounters between employees (Kabo 2017). This layout is expected to promote new communication and knowledge sharing between employees outside their own groups.
Hypothesis 2a: EGKS among employees working in a group address-based ABW office layout will increase with the introduction of a mobile Group address.
However, the effect is unlikely to be uniform. In ABW spaces where the degree of freedom in office use is high, employees are thought to actively utilise the office (Roskams et al. 2021). Some employees may choose to come to the office at times when it is easier to interact with other employees or actively work in communication spaces. In such cases, incidental communication increases, which further promotes EGKS. The converse may be observed as well. While knowledge sharing among employees belonging to different groups is expected to take place in office communication spaces, those who do not normally need to use these spaces will be unable to take full advantage of the opportunities for spontaneous communication, even if mobile Group addresses are introduced. As a result, the effect on EGKS is likely to be limited. Therefore, we propose the following:
Hypothesis 2b: The effect of introducing a mobile Group address on EGKS among employees will vary depending on whether employees utilise the communication space.
In other words, the introduction of a mobile Group address will promote EGKS to a greater extent for employees who use communication spaces frequently compared to employees who do not use communication spaces, in which case the effect will be weaker.
The survey was conducted among employees working at the headquarters of Company A, a major Japanese corporate group. Company A is a core manufacturing company with more than 10,000 employees across the group. Company A has multiple bases throughout Japan, and its headquarters building in Tokyo’s 23 wards is 1 of them. The headquarters building has 4 floors and houses 587 employees across 18 departments, including planning, sales, production technology and general affairs.
There are 2 reasons why this study focuses on the intervention results at Company A’s headquarters. First, Company A has widely adopted group address-based ABW, which has recently gained attention in Japanese companies. Offices where a sufficient sample size for statistical analysis can be secured within a specific company or office are rare. Second, it is experimentally introducing a mobile Group address system, an advanced initiative to overcome the challenges of Group address. This case study is limited to employees working at Company A’s headquarters. While limiting the scope to Company A may reduce external validity, it allows us to examine the effects of operating a group address-based ABW system under the same human resources system and talent strategy. Additionally, only a few companies have newly introduced mobile Group addresses.
Accordingly, this study is a valuable investigation into the effects of introducing mobile Group addresses. Company A introduced an ABW-type office in 2019. Like many Japanese companies, it has operated as a group-address ABW because there are many departments where fixed seats are necessary owing to the use of specific equipment within the office, and many departments prefer to sit together to some extent for the sake of information sharing within the department. The office consists of (1) concentrated seating, (2) high tables/bench seats, (3) worktables and (4) communication spaces. A worktable refers to a standard office desk typically used for daily work. For inter-group discussions or brief meetings, communication spaces may be used; these spaces do not require prior reservation for meetings or information exchange. Conversely, when employees want to work without being disturbed by others, they can move to concentrated seating. When several employees need to work together, they can move to the high tables/bench seats to work together.
As over 4 years have passed since the introduction of group-address ABW, concerns have emerged that communication paths between departments may have become somewhat fixed.
3-2. Background of intervention design and overview of interventionPrior to this survey, a preliminary survey (N = 55) conducted by Company A’s General Affairs Department revealed that communication within the company is quite active. In particular, 83.7% of the respondents (4.27 out of 5 points) rated information sharing with employees within their own group as ‘very active’ or ‘somewhat active’. By contrast, only 58.2% of respondents (3.65 out of 5) rated information sharing with employees outside their own group as ‘very active’ or ‘somewhat active’.
We introduced a mobile Group address system that involved changing the designated locations of some Group addresses within the head office and having all departments change their addresses to adjacent departments for a certain period. Although reassignments are typically random, they were adjusted to account for practical constraints such as the number of seats on each floor and the number of staff in each department. Following this change, we conducted 2 rounds of questionnaire surveys to assess the effect. The mobile Group address system was implemented from 13 November 2023 to 15 December 2023, for a total of 33 days. The duration of the intervention was primarily determined by Company A’s operational constraints. Specifically, the plan was to conduct the intervention during a relatively stable period when there were no personnel transfers and when workloads did not fluctuate drastically—such as during the year-end, New Year holidays, or the fiscal year-end. The short duration of the intervention could mean that changes in the office utilisation policy did not capture medium- to long-term behavioural changes among employees. However, we believe it allowed for an analysis that, to some extent, mitigated the impact of other factors affecting knowledge sharing, such as personnel changes—including changes in supervisors—and shifts in organisational culture. Although the period was brief, we believe that it was sufficient to foster chance encounters within the office and the resulting interactions.
In this study, we focus on the group address-based ABW model described above, which is unique to Japan. To examine the effects of the highly unusual intervention involving relocation of these group-address locations, we employ a quasi-experimental design. This method of organisational research is used to examine the impact of changes in key independent variables occurring in the field; it is employed when researchers cannot establish a control group or control key variables, as in contrast to field experiments. Although internal validity is somewhat compromised compared with experimental studies, external validity is considered to be enhanced in the sense that it facilitates examination of the impact of changing office utilisation policies on employee behaviour under specific HR practices (Grant and Wall 2009). The 2 questionnaire surveys were conducted among all employees (587) who were the target of the intervention. The General Affairs Department sent an email to employees requesting their cooperation, and the survey was conducted after clearly stating its purpose on the first page of the questionnaire. Participation in the survey was voluntary. Respondents were informed that their responses would be processed statistically, that individual responses would not be identified and that they would not face any disadvantage or unfavourable treatment even if they chose not to respond; by submitting their responses, they were deemed to have given their consent.
Responses regarding office utilisation and knowledge sharing were requested before (t1) and after (t2) the intervention. The t1 questionnaire survey was conducted from 25 October to 10 November 2023, and the t2 questionnaire survey was conducted from 18 to 29 December 2023.
Three hundred and twenty-four employees responded to the t1 questionnaire survey, but only 224 of them responded to the second survey (141 men, 82 women, and 1 other). The analysis thus focused on the 224 respondents with matching data. The average age of respondents was 45.2 years (SD = 10.2), and the average length of service at the head office building, which was the survey target office, was 9.4 years (SD = 9.1). In terms of job position, 94 respondents were managers at the section chief level or above, and 130 were ‘not managers’.
3-3. Survey itemsAll items were measured using a 5-point Likert scale in Japanese (1–5). Two members from Company A’s General Affairs Department were asked to check the translations and wording for any awkwardness, and the first author revised the Japanese expressions to ensure they were natural. The number of measurement items was carefully selected to reduce the burden on respondents. Except for knowledge sharing, which is a core measurement item, each dimension was measured using a single item.
3-3-1. Actual use of the officeRespondents were queried about their actual use of the office, including the frequency of use of the 4 components that make up the floor: concentration desks, high tables/bench seats, worktables and communication spaces. First, Company A’s General Affairs Department reviewed the original draft prepared by the lead author to ensure that the wording was understandable to the employees. As a result, revisions were made to align with Company A’s context, such as replacing the terms ‘collaborative space’ and ‘shared space’—which had been used initially—with terms used internally at Company A, such as ‘high tables/bench seating’ and ‘communication spaces’. Each component was measured using a single item. A sample item is ‘When working in the office, I frequently use the communication space’.
3-3-2. Knowledge sharingThe Japanese knowledge-sharing scale developed by Morinaga et al. (2023) was used. However, since this study required distinguishing between IGKS and EGKS, we modified the scale from Morinaga et al. (2023) and created 3 new items each for IGKS and EGKS—the 2 aspects of knowledge sharing—to measure them separately. Sample items for IGKS are ‘When I learn something new, I share it with my colleagues in the same group’ and ‘I share information with my colleagues in the same group’; sample items for EGKS are ‘When I learn something new, I share it with my colleagues in other groups’ and ‘I share information with my colleagues in other groups’. To ensure construct equivalence across the 2 types of knowledge sharing, we employed identical item wordings, varying only the referent of the knowledge-sharing target—either members of the same group or members of other groups—thus allowing for a controlled comparison between IGKS and EGKS. The results of the reliability analysis confirmed sufficient internal consistency for both the IGKS and EGKS scales, with Cronbach’s alpha coefficients exceeding 0.70 at both measurement points (Table 1).
| Mean | SD | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Age | 45.2 | 10.19 | |||||||||||||||
| 2 | Sex (1 = men 0 = others) | 0.6 | 0.48 | .248*** | ||||||||||||||
| 3 | Job (1 = manager 0 = others) | 0.4 | 0.5 | .385*** | .465*** | |||||||||||||
| 4 | t1 Use frequency: Concentration desks | 2.0 | 1.18 | .158* | .072 | .008 | ||||||||||||
| 5 | t1 Use frequency: High tables | 2.4 | 1.46 | −.118 | .022 | .112 | .065 | |||||||||||
| 6 | t1 Use frequency: Worktables | 4.3 | 1.26 | −.100 | −.116 | −.073 | −.195** | −.247*** | ||||||||||
| 7 | t1 Use frequency: Communication spaces | 3.3 | 1.27 | −.118 | .135* | .072 | .112 | .187** | .200** | |||||||||
| 8 | t1 IGKS | 4.3 | 0.68 | −.104 | .006 | .114 | −.001 | .180** | .100 | .118 | (0.788) | |||||||
| 9 | t1 EGKS | 3.9 | 0.87 | .038 | .111 | .209** | .073 | .204** | .007 | .254*** | .497*** | (0.841) | ||||||
| 10 | t2 Use frequency: Concentration desks | 2.1 | 1.22 | .036 | .046 | −.022 | .617*** | .059 | −.144* | .082 | −.065 | .024 | ||||||
| 11 | t2 Use Frequency: High Tables | 2.5 | 1.45 | −.182** | .061 | .065 | .125 | .673*** | −.253*** | .250*** | .173** | .212** | .151* | |||||
| 12 | t2 Use frequency: Worktables | 4.4 | 1.10 | −.081 | −.070 | .078 | −.120 | −.264*** | .449*** | .111 | .073 | .086 | −.159* | −.403*** | ||||
| 13 | t2 Use frequency: Communication spaces | 3.5 | 1.31 | −.120 | .118 | .035 | .117 | .223*** | .115 | .554*** | .156* | .259*** | .200** | .296*** | .080 | |||
| 14 | t2 IGKS | 4.3 | 0.67 | −.087 | −.064 | .089 | −.046 | .185 | .096 | .150* | .620*** | .313*** | .028 | .198** | −.042 | .145* | (0.776) | |
| 15 | t2 EGKS | 3.9 | 0.91 | −.024 | .140* | .256*** | .058 | .245*** | .081 | .413*** | .324*** | .562*** | .093 | .291*** | −.003 | .335*** | .425*** | (0.840) |
***p < .001, **p < .01 and *p < .05. The numbers in parentheses are coefficient alphas. EGKS, inter-group knowledge sharing; IGKS, intra-group knowledge sharing; t1, before the intervention: t2, after the intervention.
Source: Author’s own work
Prior to hypothesis testing, descriptive analyses were conducted on the frequency of utilisation for each of the 4 workspace types: concentration desks, high tables/bench seats, worktables and communication spaces. Among these, worktables exhibited the highest mean utilisation score (M = 4.30), followed by communication spaces (M = 3.30). In contrast, the utilisation of high tables (M = 2.40) and concentration desks (M = 2.00) was substantially lower.
Subsequently, we computed the average levels of knowledge sharing. At time 1 (t1), the mean score for IGKS was 4.27 (SD = 0.68), whereas EGKS averaged 3.86 (SD = 0.87). At time 2 (t2), the means were 4.28 (SD = 0.67) for IGKS and 3.87 (SD = 0.91) for EGKS. These results suggest that both types of knowledge sharing were perceived to occur at relatively high frequencies, although EGKS consistently scored lower than IGKS. Furthermore, the standard deviation for EGKS was larger than that for IGKS at both time points, indicating greater individual variability in EGKS. Table 1 presents the correlations between the measured variables. These results indicate that there is a correlation between knowledge sharing and male employees, as well as between knowledge sharing and holding a managerial position.
Finally, reliability analyses demonstrated adequate internal consistency for both IGKS and EGKS scales, with Cronbach’s alpha coefficients exceeding 0.70 across both measurement occasions (Table 1). To examine the potential for bias due to dropouts, we compared 2 groups: continuing respondents (n = 224) and dropouts (n = 100). We performed chi-square tests for gender (male) and job title (managerial or higher), and t-tests for the frequency of use of each workspace at time t1, as well as the extent of knowledge sharing within and outside the group. The results revealed that, compared to dropouts, continuing respondents had a higher proportion of managerial staff, a higher rate of work desk usage at time t1 and higher levels of both IGKS and EGKS. However, no significant differences were observed regarding gender, the use of concentration desks, high tables/bench seating, or communication spaces. These results suggest that the analysis in this study may be biased toward employees who were already proactive in knowledge sharing.
4-2. Differences in EGKS by communication space utilisationTo test Hypothesis 1, we divided the participants into 2 groups based on their communication space utilisation: those who answered 4 or 5 (102 participants) and those who answered 3 or below (122 participants). We then calculated the average values for each group and performed a t-test to compare the differences between the t1 average values of EGKS (Fig. 1).
The group with high communication space utilisation had higher values than the group with low utilisation, indicating a statistically significant difference (t = −2.435, df = 222, p = .016, d = −0.326) and supporting Hypothesis 1. However, no statistically significant difference was observed in the degree of IGKS (t = −1.187, df = 222, p = .236). Furthermore, when the analysis was conducted on a sample of 324 respondents—including the 100 who answered only t1—the results remained unchanged.
4-3. Effect of the introduction of mobile group address-based ABWPrior to testing Hypothesis 2, we conducted a paired t-test to examine the effect of the introduction of mobile group address-based ABW on employees’ office utilisation patterns and investigated possible differences in the utilisation patterns of the 4 components that make up the office.
The utilisation of worktables and communication spaces increased at statistically significant levels (workspaces: t = −2.091, df = 223, p = .038, d = −0.147; communication spaces: t = −2.082, df = 223, p = .038, d = −0.131). Evidently, the new initiative encouraged 2 different types of behaviour: one where employees tended to stay within their Group address, and another where they actively engaged in communication with many people. Given the small effect sizes (Cohen’s d) observed in the 2 analyses, whilst the impact of the intervention was not particularly large, there was a change before and after the intervention.
To verify Hypothesis 2a, we calculated the average level of knowledge sharing between groups at time t1 (before the introduction of the mobile group-address ABW) and time t2 (after the introduction). The average value after introduction was 0.016 higher than that before introduction. However, a paired t-test showed no statistically significant difference (t = −0.280, df = 223, p = .779). Therefore, the introduction of mobile group-address ABW did not increase the degree of EGKS across Company A, and Hypothesis 2a was not supported.
To examine the effects of communication space usage (unmatched: high/low) at time point t1, and the introduction of mobile Group address (matched: before/after) on EGKS with regard to Hypothesis 2b, we conducted a 2-way analysis of variance (mixed design) controlling for gender and job title (as covariates). The results showed a simple main effect of communication space utilisation on EGKS (F (1,220) = 16.562, p < .001, partial η2 = 0.070) and an interaction effect between the utilisation of communication spaces and the intervention effect (F(1,222) = 7.324, p < .01, partial η2 = 0.032), which were both statistically significant.
We finally conducted a simple main effect test to interpret the interaction. In the group with high communication space utilisation, the degree of EGKS increased at a statistically significant level between pre-intervention and post-intervention (p = .045). Contrastingly, in the group with low utilisation of the communication space, the extent of EGKS decreased between pre-intervention and post-intervention, although this decrease was not statistically significant. We also analysed the degree of knowledge sharing within groups, but no statistically significant effects were observed for any of the factors (Fig. 2).
These results indicate that the introduction of a mobile Group address has a statistically significant effect on the degree of knowledge sharing between groups that already make extensive use of communication spaces, but no such effect was observed in groups that did not. Therefore, Hypothesis 2b was supported.
To conclude, this study conducted a disaggregated analysis of knowledge sharing, distinguishing between IGKS and EGKS. With respect to IGKS, the findings revealed no statistically significant effects associated with either communication space utilisation frequency or the implementation of mobile Group address. The consistently high levels of IGKS observed at Company A can plausibly be attributed to the group address configuration, wherein employees are physically co-located with their team members within an open-plan office. Moreover, the data suggest that this intra-group communication pattern remains stable and is not disrupted by the introduction of a mobile Group address. These findings indicate that, when appropriately designed and implemented, mobile Group address interventions may maintain high levels of IGKS while concurrently enhancing EGKS.
The results of this study supported Hypotheses 1 and 2b but did not support Hypothesis 2a. The analysis of Hypothesis 1 revealed that, in group-address ABW spaces, the extent of knowledge sharing behaviour between groups differed depending on whether employees frequently utilised communication spaces. Employees who frequently used them also frequently shared knowledge with employees from different groups, while those who did not use them frequently shared knowledge less frequently. These results can be interpreted from 2 perspectives. First, given that the extent of knowledge sharing varies depending on the frequency of communication space usage, increasing this frequency could promote knowledge sharing. However, based on the findings of this study, it is not possible to definitively state the reasons for variations in the frequency of communication space usage among employees. We can infer possible causes from several perspectives. First, the use of communication spaces may be influenced by the need to interact with others, such as the interdependence of work tasks. Cheng et al. (2023) demonstrated that task interdependence moderates the relationship between flexible work arrangements and knowledge sharing. Furthermore, Irving, Ayoko, and Ashkanasy (2020) noted that some employees intentionally avoid interacting with others. Given these points, office utilisation patterns may vary due to various factors—such as differences in job characteristics, office usage habits, and individual differences—or combinations thereof.
Analyses related to Hypotheses 2a and 2b revealed that the introduction of a mobile Group address had a partial effect on knowledge sharing between employee groups. In other words, employees who already frequently used communication spaces showed an increase in knowledge sharing between groups, but no such change was observed among employees who did not. As noted in prior studies, employees are often proactive in utilising the office (Hoendervanger et al. 2016; Irving, Ayoko, and Ashkanasy 2020; Roskams et al. 2021). However, employees do not necessarily use the office as intended by its designers. Among those who normally do not use communication spaces, the introduction of a mobile Group address may not necessarily lead to increased communication or knowledge sharing, even if such employees were forced to change departments adjacent to each other in the office. While the introduction of a mobile Group address can be expected to have a certain effect, it is important to simultaneously address employees’ attitudes toward office utilisation, as discussed in the analysis of Hypothesis 1. As such, the extent of benefits of mobile group addresses varies depending on differences in how employees utilise office spaces. Based on these findings, even when introducing mobile Group addresses, it is advisable to proceed with an awareness of the limitations of their effectiveness. Although this study was unable to identify the specific factors driving differences in the frequency of communication space usage, these differences in usage patterns can be an effective starting point for determining the scope of application.
5-1. Theoretical implicationsPrevious studies on ABW-type offices have mainly focused on Western companies. However, this study revealed that the effectiveness of ABW varies depending on employees’ office usage styles. This study is the first to focus on knowledge sharing in group address-based ABW, which is often implemented in Japanese companies. Furthermore, similar trends observed in Western countries were confirmed in group address-based ABW introduced in Japanese companies, demonstrating the importance of considering ABW operational processes across countries and companies.
5-2. Managerial implicationsThis study partially recognises the significance of introducing group address-based ABW and mobile Group address. Although group address-based ABW is an adaptive utilisation style based on the historical background of Japanese companies, it can also be introduced in Western ABW-type offices.
The findings of this study indicate that Company A, which has implemented a group address-based ABW system, exhibits consistently high levels of IGKS. This pattern appears to be robust and invariant across variations in individual office space utilisation styles. For organisations aiming to enhance knowledge sharing both within and across workgroups in ABW environments, the adoption of a Group address configuration may serve as an effective structural intervention to reinforce intra-group knowledge exchange while providing a foundation for broader collaborative practices.
Additionally, introducing mobile Group addresses has been shown to have a certain effect on employees who already utilise communication spaces. Introducing mobile Group addresses as an additional policy after a certain period of implementing group address-based ABW may have a partial effect on promoting knowledge sharing within the organisation. However, the survey results also indicate that the impact is limited. Notably, introducing a one-size-fits-all policy regarding the use of office space does not necessarily have a significant impact when viewed from the perspective of the organisation as a whole.
5-3. Implications for administrative policiesThe findings of this study also hold unique implications for the formulation and implementation of office usage rules and policies within organisations. Based on the findings, effective office utilisation can be achieved by considering spatial design and policy design in conjunction. While office space design must inevitably be uniform for all users, there is room to incorporate a certain degree of flexibility into the design of office usage policies, such as by limiting the target audience. As such, the perspective that office usage policies can be utilised to compensate for the limitations of uniform spatial design is unique to this study.
Second, organisations considering the introduction of new office usage policies should take the following 2 points into account to promote effective implementation. First, it is necessary to identify the policy’s target audience and conduct effectiveness evaluations accordingly. Organisational policies are expected to have 2 effects: suppressing specific behaviours and promoting specific behaviours (Adler and Borys 1996). Regarding the latter, office utilisation policies should be accompanied by initiatives to communicate the policy’s objectives and significance to employees in order to encourage their voluntary participation. As Irving, Ayoko, and Ashkanasy (2020) pointed out, future research should focus on the differences in the effects of office design and new office utilisation policies on employees and examine the combined influence of the factors determining these effects. In this study, the introduction of new organisational policies in the office was limited to increasing the use of communication spaces among some employees. Some previous studies have attempted to explain differences in employee reactions to the same policy as stemming from differences in the underlying reasons for policy implementation (HR attribution) (Dawson, Parker, and Okimoto 2024; Nishii, Lepak, and Schneider 2008). Based on the previous studies, when office usage policies are introduced, employees may attribute different reasons for the policy’s implementation, and these differences may lead to variations in employees’ office usage behaviours following the policy’s introduction, and consequently, to differences in their motivation to interact with employees from other groups. Such attributions regarding policies and measures are believed to be influenced by the actions of various stakeholders, including senior management, the leading department and managers (Bowen and Ostroff 2004). These stakeholders must collaborate and communicate a consistent message regarding the new office usage policy. However, such attribution processes were not directly investigated in this study and should be examined in future research.
5-4. LimitationsThis case study focuses on initiatives implemented throughout the headquarters building of Company A in Japan. While it was possible to survey a relatively large number of employees, it was not possible to set up a control group. Another limitation of this study was the short intervention period. Considering that measurements were limited to comparisons between 2 time points—before and after the intervention—we were unable to verify the long-term effects of the intervention. The fact that we were unable to capitalise on the strength of quasi-experimental designs—namely, the ability to construct theories that account for temporal changes in intervention effects (Grant and Wall 2009)—is a significant challenge. To evaluate the effectiveness of the intervention more accurately, future research must be based on a more rigorous design, such as a randomised controlled trials that includes measurements at 3 time points: before and after the intervention, as well as several months later. Furthermore, caution is required when generalising the findings obtained from the case study of a specific Japanese company. This study indicates a certain degree of bias among the employees who responded to both surveys. Caution is warranted, as the survey results may reflect a tendency toward managers and employees who are proactive about knowledge sharing. Furthermore, future research should survey companies in other countries that have implemented group address-based ABW, as well as different companies, to verify the external validity of the results.
The first author, Morinaga, designed the questionnaire survey, analyzed the data, and undertook the writing. The second author, Nihei, led the coordination with the research site and assisted in designing the questionnaire survey. The approximate contribution ratio was 8:2.
Competing interestsThere are no conflicts of interest concerning this research.
Use of artificial intelligence (AI)DeepL was used for part of the English translation. Based on DeepL’s translation results, the first author carefully reviewed and revised the text while verifying its correspondence with the original, and the final version was completed after proofreading by a translation agency.
FundingThis research was supported by the Management Science Research Grant from the Japan Management Association and JSPS KAKENHI Grant Number 25K05243.
