Remote work has become a permanent feature of labour markets. It can reduce commuting costs, expand flexibility, and help firms recruit workers from a wider talent pool (Barrero et al. 2023, Aksoy et al. 2025a). But it also changes how employees learn from each other, receive feedback, and build attachment to the firm. This creates a practical question for firms: how can they preserve the benefits of remote work while restoring some of the workplace interactions that support performance and retention?

In a recent paper (Aksoy et al. 2026), we study this question using a nine-month randomised controlled trial at TEMPO BPO, one of Türkiye’s largest business-process outsourcing firms. The experiment took place within a fully remote inbound customer-service team. Employees managed incoming calls under fixed schedules, using the same technology, protocols, and performance metrics. The intervention was intentionally minimal. Among 248 eligible employees, 124 were assigned to work from the office together one designated day per month, while the remaining 124 continued working fully remotely. No other aspects of the work environment changed, including compensation, schedules, workload, call-routing protocols, production technology, incentives, or formal training. The intervention solely introduced recurring opportunities for in-person interaction among employees and team leaders. Assignment was randomised based on employees’ pre-existing company IDs. Compliance was about 95%.

Monthly office days raised productivity

Calls handled per hour is the firm’s standard measure of employee productivity. Prior to the intervention, the treatment and the control group were balanced, with both averaging about 10.8 calls per hour. Following the introduction of monthly office days, the treatment group gradually began to outperform the control group. Notably, the effect did not appear immediately – it accumulated through repeated visits and persisted after the intervention ended and treated employees returned to fully remote work.

In the five months after the experiment ended, employees in the treatment group handled 12.4 calls per hour, compared with 11.5 in the control group – a raw productivity gap of 7.8%. Employee fixed-effects estimates indicate that this improvement largely reflected gains in individual performance rather than changes in workforce composition. There is no evidence of a speed-quality trade-off. Customer ratings and internal audit scores did not fall after the intervention began.

Figure 1 Calls per hour before, during and after the randomised controlled trial  

Notes: Weekly mean calls per hour for the control (‘no office days’) and treatment (‘one office day per month’) groups. Both groups averaged 10.8 calls per hour before the intervention; post-intervention means are 11.5 and 12.4, respectively. Shaded areas are 95% confidence intervals.

The gradual build-up of productivity gains is important for interpretation. A simple Hawthorne effect or novelty response would be expected to produce an immediate spike that fades over time. Instead, the opposite pattern was observed – gains accumulated progressively month by month and persisted after the programme ended. This suggests that the intervention worked through more enduring mechanisms, such as learning, stronger team cohesion, and more effective feedback. The results are robust when treated employees are compared to 185 employees in the same province who did not volunteer to participate in the experiment and who remained remote throughout, ruling out disappointment effects among control workers.

The same firm realised sizeable productivity gains a few years earlier when it shifted from fully onsite work in a traditional office setting to fully remote work. The firm undertook that earlier shift in reaction to the COVID-19 pandemic. While the earlier shift brought productivity gains without a loss of service quality, it also brought higher turnover and turnover costs (Aksoy et al. 2025b).

In-person contact strengthened communication

The experiment included randomised seating assignments during office visits, providing direct evidence on how proximity shapes communication. Employees randomly assigned to sit next to each other were 11 percentage points more likely to communicate in the following week, relative to a baseline of 9% for all within-group pairs. Employees who shared a coffee break or lunch showed even greater increases in subsequent interactions. These patterns are consistent with earlier work showing that brief physical proximity can create lasting communication ties (Battiston et al. 2021, Emanuel et al. 2023).

At the end-line survey, treated employees reported spending an additional 36 minutes per week communicating with their closest colleagues. They were also more likely to report receiving regular feedback from their team leaders and to describe team communication as effective. Taken together, these findings point to monthly office days restoring informal learning and managerial support, which were more difficult to sustain in a fully remote work environment.

Monthly office days reduced attrition

The intervention also improved retention. By the end of the observation period, cumulative attrition reached 21.0% in the control group and 13.7% in the treatment group, representing a roughly one-third reduction in attrition. The gap between the two groups continued to widen after the final office visit, indicating that the policy had a lasting effect on retention, rather than merely postponing exits.

Figure 2 Monthly office days reduced cumulative attrition 

Notes: Cumulative attrition for the treatment group (‘one office day per month’, red line) and the control group (fully remote ‘no office days’, black line). The gap widens after the first office visit and persists after office visits end. Shaded areas are 95% exact binomial confidence intervals.

The composition of employee turnover also differed across groups. In the control group, employees who left the company were, on average, more productive than those who stayed. In contrast, the opposite pattern emerged in the treatment group: employees who stayed outperformed those who departed. This suggests that monthly office days may have helped managers identify and retain higher-performing employees – an important advantage in settings where recruiting and onboarding new workers is time-consuming and costly.

Implications

The programme was inexpensive to implement. The firm covered the direct costs of monthly office visits, including transport, meals, refreshments, and office space. These costs were more than offset by gains from higher productivity and lower turnover, as fewer employees needed to be recruited and trained. Overall, these gains exceeded programme costs by a factor of five. These gains were broad-based: we find no statistically significant heterogeneity by gender, age, marital or parental status, or baseline productivity.

The results are most directly relevant to remote teams where communication, feedback, and peer learning matter but a full return to the office is neither necessary nor desirable. While Bloom et al. (2024) show that hybrid working can improve retention without hurting productivity, our findings extend that evidence by showing that even one coordinated day per month can generate meaningful improvements in a fully remote setting.

Two features of the design are especially noteworthy for managers and policymakers. First, the office visits were coordinated, with employees coming into the office on fixed office days. By contrast, staggered individual office days – an approach some firms have adopted to manage space constraints – may not generate the same peer learning and communication benefits. Second, the gains accrued while employees remained fully remote for the rest of the month. This suggests that the value of occasional co-location comes less from monitoring or discipline than from the quality of interaction.

The broader message is straightforward: remote work does not have to mean isolated work. A small amount of well-designed, coordinated in-person contact can improve communication, raise productivity, and reduce attrition, while preserving most of the flexibility that makes remote work attractive to both firms and employees.

References

Aksoy, C G, J M Barrero, N Bloom, S J Davis, M Dolls, and P Zarate (2025a), “The global persistence of work from home”, Proceedings of the National Academy of Sciences 122(27).

Aksoy, C G, N Bloom, S J Davis, V Marino, and C Özgüzel (2025b), “Fully remote work expands recruitment and boosts productivity”, VoxEU.org, 1 June.

Aksoy, C G, N Bloom, S J Davis, V Marino, and C Özgüzel (2026), “The value of one office day a month”, CEPR Discussion Paper 21597. 

Barrero, J M, N Bloom, and S J Davis (2023), “The evolution of work from home”, Journal of Economic Perspectives (Fall): 23–50.

Battiston, D, J Blanes i Vidal, and T Kirchmaier (2021), “Face-to-face communication in organisations”, Review of Economic Studies 88(2): 574–609.

Bloom, N, R Han, and J Liang (2024), “Hybrid working from home improves retention without damaging productivity”, Nature 630(8018): 920–25.

Emanuel, N, E Harrington, and A Pallais (2023), “The power of proximity to coworkers: Training for tomorrow or productivity today?”, NBER Working Paper 31880.