Return-to-office edicts often lean on a fear few executives name plainly: if people are not seen, their careers will stall. The best evidence cuts both ways. Hybrid schedules reduce resignations without harming performance or promotion rates. Yet a classic experiment on fully remote work found a promotion penalty even when results were equal. The issue is not location. It is whether a company can see and judge the work without relying on proximity.
What the strongest experiments actually say
A randomized controlled trial inside Trip.com assigned 1,612 university-trained employees to two days at home per week or five in the office. Resignations fell by a third in the hybrid group, from 7.2 to 4.8 percent, and there was no difference in performance reviews or promotions across two years of follow-up (Bloom, Han and Liang, 2024). Coverage of the study notes that the retention gain was concentrated among non-managers, a detail senior leaders should not miss when thinking about where attrition risk is highest (Stanford Report, 2024). Trip.com estimates the reduction in turnover saved millions of dollars in recruiting and training costs, which flowed straight to profit, according to the study’s managerial write-up (Bloom, Han and Liang, 2024, HBR).
Contrast that with an earlier randomized trial at Ctrip, the same firm in an earlier era, where call-center staff worked fully from home or stayed in the office for nine months. Working at home raised performance by 13 percent, mostly from more minutes worked per shift and some from a quieter environment, and it halved attrition. Yet remote workers were less likely to be promoted conditional on the same performance, a career penalty tied to visibility rather than value (Bloom, Liang, Roberts and Ying, 2015). When Ctrip later let everyone choose their arrangement, more than half switched, and the measured productivity gains rose to about 22 percent, combining selection and learning. The lesson is simple. Where and how people work are design variables, and the design determines whether the benefits survive contact with real promotion practices.
The real variable is legibility, not location
The Trip.com authors offer three conditions for a hybrid model to work in practice: rigorous performance management, coordinated team schedules, and visible support from top leadership (Bloom, Han and Liang, 2024, HBR). Each one increases the legibility of work. If a manager can see the targets, the cadence of review, and the times teams overlap in person, they are less tempted to use attendance as a proxy for contribution.
Asynchronous practices are the operating system for that legibility. GitLab, which runs one of the largest all-remote organizations, states the rule openly in its public handbook. Async work is impossible without strong documentation, because otherwise a single follow-up question can block a colleague for hours or days. The company treats its living handbook, which would run to more than 2,000 printed pages, as the first venue for decisions, processes, and norms, so anyone can self-serve answers rather than wait for someone to come online (GitLab handbook, 2026). This is not a randomized trial. It is a working blueprint from a practitioner that explains the mechanism the experiments require.
Matt Mullenweg, the chief executive of Automattic, frames the same idea in a taxonomy of distributed work. At what he calls Level 4, organizations measure output rather than hours and rely on asynchronous communication as the default. That is a practitioner’s view, not a statistical claim, but it captures the cultural move hybrid firms must make if they want retention gains without career penalties (Mullenweg, 2020).
A practical playbook to close the visibility gap
Start with outcomes. Define unit-level output metrics and a review cadence that survives time zones and mixed schedules. The Trip.com authors emphasize rigorous performance management for a reason. It removes the alibi for managing by attendance and provides a record that travels across locations and weeks (Bloom, Han and Liang, 2024, HBR).
Standardize artifacts. Require decision logs for major choices, design briefs for projects, checklists for recurring processes, and searchable notes from key meetings. GitLab’s handbook-first approach is extreme by design, but the principle is general. A shared repository lets work stand on its own, closes loops faster, and gives promotion committees something better than anecdote to judge (GitLab handbook, 2026).
Coordinate in-office time around mentorship and creative collision, not status updates. The hybrid experiment used a fixed pattern of two at home and three in the office. The authors argue that team-level coordination matters as much as the ratio. If your Tuesdays in the office are spent reading slides that could have been a page, you are squandering the only scarce commodity you have together (Bloom, Han and Liang, 2024, HBR; Bloom, Han and Liang, 2024).
Audit fairness. The 2024 study’s retention benefits were strongest among non-managers, according to Stanford’s coverage. That is a signal to monitor promotion and exit rates by subgroup. If a unit’s remote staff consistently lag in advancement after you control for results, you have an evaluation problem, not a location problem (Stanford Report, 2024). This is an interpretive step, but it follows directly from the career patterns the experiments revealed.
Test your own model. The Trip.com researchers recommend A/B testing management practices, not just policies. Trial fixed-team office days against free choice. Vary the performance metrics and the review cadence. Close the loop by publishing the results internally. Leaders cannot copy a single firm’s allocation of days and expect the same effect. They can copy the discipline of testing and the insistence on measuring output instead of presence (Bloom, Han and Liang, 2024, HBR).
The case for hybrid is now stated in the language boards hear. Fewer resignations without a productivity hit, and a credible path to profit improvements from lower churn (Bloom, Han and Liang, 2024; Bloom, Han and Liang, 2024, HBR). The career risk comes from old habits that judge proximity. Build an artifact layer that makes contribution visible, align schedules to where it adds the most value, and enforce performance management that puts outcomes in writing. Do that, and out of sight will no longer mean out of career.
Sources
- Bloom, Han and Liang (2024). Hybrid working from home improves retention without damaging performance, Nature. A randomized trial at Trip.com found two days at home cut resignations by a third with no loss to performance or promotions. https://www.nature.com/articles/s41586-024-07500-2
- Bloom, Han and Liang (2024). One Company A/B Tested Hybrid Work. Here's What They Found, Harvard Business Review. Reports the turnover savings in the millions and spells out three conditions for hybrid to work, urging firms to A/B test their practices. https://hbr.org/2024/10/one-company-a-b-tested-hybrid-work-heres-what-they-found
- Bloom, Liang, Roberts and Ying (2015). Does Working from Home Work? Evidence from a Chinese Experiment, Quarterly Journal of Economics. A randomized trial where full-time WFH raised performance 13 percent and halved attrition but reduced promotion odds conditional on performance; choice later doubled gains to 22 percent. https://academic.oup.com/qje/article-abstract/130/1/165/2337855
- GitLab (2026). Handbook-first & asynchronous work, GitLab Handbook. Case evidence describing a public handbook exceeding 2,000 pages and the claim that async work is impossible without strong documentation. https://handbook.gitlab.com/handbook/company/culture/all-remote/asynchronous/
- Stanford Report (2024). Hybrid work is a win-win-win for companies and workers, Stanford News. Secondary coverage noting retention gains concentrated among non-managers and reiterating no performance or promotion penalty. https://news.stanford.edu/stories/2024/06/hybrid-work-is-a-win-win-win-for-companies-workers