For most offshore teams the evidence points to a dedicated workspace on two or three anchor days a week, with home days kept for individual work, because home working succeeds only under conditions that an employer abroad cannot verify: a private room, a stable connection, work that is individual and measurable, and a person who has already learned the job. Where those conditions held, as at Ctrip, home working beat the office. Where they failed together, as in India and at a large Asian IT services firm, it lost by double digits. A dedicated workspace guarantees the conditions; hybrid takes the retention gain without the daily commute.

This guide covers one layer of a larger operating standard. For how the workspace decision fits with hiring, onboarding, devices, performance and compliance, see the employer’s guide to professional remote work.

Two experiments, one dataset, one pattern

Two randomised experiments and one very large company dataset account for most of what is genuinely known about home versus office productivity. They appear to contradict each other until you look at who was in each.

Ctrip, 2015: home working won by 13%. Stanford’s Nicholas Bloom and colleagues ran a nine-month randomised experiment at Ctrip, a 16,000-employee Chinese travel agency. Call-centre employees who volunteered to work from home were assigned by lottery to home or office. Home working raised performance by 13%: about 9 percentage points from working more minutes per shift (fewer breaks and sick days) and 4 from handling more calls per minute, which the authors attributed to a quieter environment. Home workers reported higher satisfaction and left less often, though their promotion rate, conditional on performance, fell (Bloom, Liang, Roberts and Ying, Quarterly Journal of Economics, February 2015).

The detail that matters for an employer is the entry condition. Of the 503 employees who volunteered, only 249 qualified, because Ctrip required at least six months’ tenure, broadband internet and a private room at home to work in. The 13% belongs to that filtered group.

India, 2023: home working lost by 18%. David Atkin, Antoinette Schoar and Sumit Shinde ran a randomised trial in India’s data-entry sector, allocating newly recruited workers to home or office regardless of their home setup. Productivity at home was 18% lower. Two-thirds of the gap appeared on the first day of work; the rest came from workers in the office learning the job faster. The study also found a selection effect that should worry anyone who sets workspace policy by asking people what they prefer: workers who preferred home were 27% less productive at home than in the office, against 13% for workers who preferred the office, with the difference concentrated among those with children or other care responsibilities at home (Atkin, Schoar and Shinde, NBER Working Paper 31515, July 2023, revised December 2023).

Asian IT services, 2021: skilled work lost 8% to 19%, and hours went up. Michael Gibbs, Friederike Mengel and Christoph Siemroth studied more than 10,000 professionals at a large Asian IT services company before and during the pandemic move to home working, using the company’s own personnel and activity data. This was forced home working under lockdown, with schools closed, so it is an observation rather than an experiment, and the finding that employees with children fared worst should be read in that light. Output did not rise while hours did, so productivity fell by 8% to 19%. Time in meetings and coordination went up, uninterrupted working hours shrank, employees communicated with fewer people inside and outside the firm, and they received less coaching and fewer one-to-one meetings with their managers (Gibbs, Mengel and Siemroth, Becker Friedman Institute Working Paper 2021-56).

Read together, the three stop contradicting each other. Home working succeeds for experienced people doing individual, measurable work from a room of their own on a reliable connection. It fails for people who are still learning, for work that depends on coordination and coaching, and for homes that were never built for it. The Ctrip group had every advantage. The Indian group was allocated without regard to any of them. The IT professionals kept their skills and lost the coordination.

What the winning home offices had in common

Ctrip did not show that homes are productive. It showed that a specific kind of home is. Its three entry conditions map almost exactly onto what we ask of anyone working from home on our own Indonesian team.

• A private room. Ctrip required one. Our baseline asks for a quiet, neutral background and a webcam at eye level facing the light, because clients judge the professionalism of an offshore team through the camera frame on every call.

• A connection that does not depend on the household. Ctrip required broadband. Our baseline is at least 20 Mbps with a mobile hotspot as backup. In Jakarta the backup is viable: the province’s median mobile download speed was 41.42 Mbps in the first half of 2025, the highest in Indonesia (Ookla data reported by Katadata). Whether the fixed line is shared with the rest of the household is a separate question, and it is the one most often left unasked.

• Someone who already knows the job. Ctrip required six months’ tenure. The Indian trial showed why: home working slows learning. For an offshore hire, the first 90 days are when context, correction and observation are worth most, which is why we treat remote onboarding in the first 30 days as a workspace question as much as an HR one.

• The equipment. An external monitor and a headset with a microphone are on our baseline because they remove the two most common quality failures on client calls, and the laptop itself should be a managed device. That is covered in our guide to remote work device security and MDM.

An employer can write these conditions into a policy. What an employer cannot do, across a border, is verify them. That is the gap a dedicated workspace closes.

Where home working quietly loses

None of the losses in the research announces itself in a week’s output. They compound.

Learning slows. Two-thirds of the Indian gap was present on day one, but the remaining third was the office group getting better faster. Over a first year, the office-trained hire becomes a different employee.

Coordination replaces work. The IT services professionals did not slack off. They worked more hours and produced the same output, because meetings and coordination expanded into time that used to be uninterrupted. For a team working across time zones with a client, that pattern is easy to create and hard to see, since the hours look committed.

Coaching disappears. The same study found fewer one-to-ones and less coaching from supervisors. That informal correction is what turns a good hire into an independent one, and it is the first thing a remote manager drops when the calendar fills.

Retention gains are real, but they are hybrid gains. The 33% fall in resignations in Bloom’s 2024 Trip.com trial came from two days a week at home and three in the office, among university-educated professionals. Our guide to hybrid versus remote work and productivity sets out why we run our Indonesian teams that way.

The Jakarta specifics

The general research needs one local adjustment in each direction.

The commute is a genuine cost, so a five-day office is the wrong answer too. In TomTom’s Traffic Index for 2025, a 10 km drive in Jakarta averaged 26 minutes 19 seconds, rising to 38 minutes 43 seconds in the evening rush hour at an average 15.5 km/h, and a rush-hour commuter lost 125 hours to congestion over the year, the equivalent of three working weeks. Jakarta’s congestion level was 59.8%. Every one of those hours argues for home days.

The home is harder to guarantee than in the markets most of the research comes from. This is our operating observation from running Indonesian teams, and we state it as such: a good hire’s home setup fails for ordinary reasons. The fixed line is shared with several other people, the power drops, there is no room that can be closed, or a year of working alone quietly loosens someone’s attachment to an employer they have never physically visited. None of these shows up as a performance problem until it shows up as a resignation.

Put the two together and the shape of the right answer is already visible: a professional workspace the team actually comes into on anchor days, and home days for the deep individual work that the research says home is good at.

Home office, dedicated workspace, or hybrid: which fits?

A decision rule for offshore teams

Use Ctrip’s three entry conditions plus two of our own as a checklist per person, and decide per person.

1. Has the person done this job for at least six months? New hires learn faster in a room with colleagues. Weight the first 90 days towards the workspace.

2. Can you verify a private room and a stable connection? Asking is not verifying. If the honest answer is no, the workspace is the only arrangement whose conditions you actually know.

3. Is the work individual and measurable, or does it depend on coordination and coaching? Support, finance and operations roles with clear output can carry more home days. Roles that depend on daily collaboration should not.

4. Does the work touch client or customer data? If it does, a home network of unknown devices is a compliance question before it is a productivity one, and an access-controlled office with a business line is a materially different risk position.

5. Is the person new to your company, even if experienced? Context about your clients, your systems and your standards is learned fastest in the room, whatever the CV says.

Most teams end up in the same place: two or three anchor days in a professional workspace, home days for deep individual work, and a home-office baseline that applies on the home days too.

Where MixWork Spaces fits

We run our own Indonesian teams this way, which is why the product exists. MixWork Spaces: Dedicated Workspaces in Jakarta gives your team a permanently assigned desk in our own managed office (our own premises, not a resold coworking pass), with bookable meeting rooms for client calls and interviews, business-class amenities, 24/7 physical security, a business internet line that does not slow down when the household comes online, attendance and leave recorded on site and reported to you, and a MixWork HR manager working in the same building. That last item is the one clients notice: problems get raised while they are still small enough to fix, because someone from our HR team is in the room.

It costs from USD 199 per workspace per month, and that fee carries the desk, the rooms, the amenities, the security, the connectivity, the attendance management and the on-site supervision. Salaries and employer statutory contributions are separate, and the whole picture is on our pricing page. You choose the pattern: some clients run fully office-based teams, most use two or three anchor days a week, and some keep desks available for the people whose home setup cannot meet the baseline.

We treat the workspace as a retention investment because the ordinary, cumulative things that make a good offshore hire leave (an unreliable setup, no colleagues, no visible career context, no sense the employer is real) are all addressed by a real office at once. Twelve-month retention across our placements runs above 90%, and every habit in this series compounds with tenure and resets to zero with churn.

If you are deciding how your Indonesian team should work and want the employment, HR and workspace layers handled by one accountable partner, talk to us.

Note: The studies cited are two randomised field experiments (China, India) and one company dataset (an unnamed Asian IT services firm); their findings are indicative for other markets and are attributed with their publication details above. MixWork figures are our own operating data and pricing as at September 2026. This article is general guidance, not legal advice.