Picture the most expensive non-meeting on your calendar. It is Tuesday, the designated anchor day. Your team commuted in, badged through the turnstile, found desks, put on headphones, and spent the day on video calls with people in other buildings. At five o’clock, they commuted home. Attendance was 100 percent. Collaboration was zero. And somewhere, a dashboard turned green.

That is the hybrid work policy many companies have built: a fixed number of days, enforced by a badge reader, with no serious answer to the question that matters most — what are those days supposed to produce? The numbers on where people work have stopped moving. What has not settled is whether the office time you are paying for, and asking your people to pay for in commute hours, is producing anything worth the cost.

This week’s topic is hybrid work with a purpose — not the tired fight over whether remote work is good or bad, because the data has mostly settled that, but the harder leadership question underneath it: once people are in the office, what makes the day worth the trip?

The Numbers Have Stopped Moving

Start with the stable facts, because many leadership conversations still treat hybrid work as a temporary arrangement waiting to snap back.

Gallup’s September 2025 report found that hybrid work among remote-capable U.S. employees had leveled off since late 2022. Hybrid stood at 51 percent, down from 55 percent over the previous two quarters, with fully on-site and fully remote work each picking up two points. Gallup’s headline was simple: the retreat from hybrid had barely happened. Hybrid workers were spending 46 percent of their week in the office, roughly 2.3 days, up from 42 percent in 2022. That shift happened in 2023, with no change since.

Stanford economist Nick Bloom, who has tracked this longer than anyone, puts remote work at about 28 percent of all U.S. workdays and confirms it three ways: surveys, building badge swipes, and cell phone location data. It has held there for more than two years. Kastle Systems, which tracks badge entries across more than 2,600 buildings, shows office occupancy hovering just above 50 percent of pre-pandemic levels, with midweek attendance consistently stronger than Fridays.

Employee preference is just as settled. In 2026, Gallup reported that six in ten remote-capable employees want hybrid work, about a third want fully remote work, and fewer than one in ten want to be in the office every day. If you are waiting for that to change, you will be waiting a long time. This is the baseline. Plan around it.

What a Mandate Costs You

Every few months, a large company announces a return-to-office mandate, the press covers it, and leaders elsewhere wonder whether they are falling behind. So it is worth being clear about what the research says happens after those announcements.

Researchers at the University of Pittsburgh tracked the employment histories of more than three million technology and finance workers on LinkedIn to measure what return-to-office mandates did at S&P 500 firms. After a mandate, turnover ran abnormally high. Senior employees, women, and more skilled workers left at higher rates than everyone else; the turnover increase for senior managers was close to 19 percent, and for women it was 20 percent, compared with 7 percent for men. Hiring got harder too: vacancies took 23 percent longer to fill, and the hire rate fell 17 percent. The same research team had found earlier in 2024 that mandates did not improve firm profitability. Their summary of the effect was brain drain, and they noted that it hit even the largest firms in the world.

Employers already know this. In a Resume Builder survey, about 80 percent of companies said they had lost talent because of return-to-office rules, and a quarter still planned to increase in-person requirements. One in five workers said they simply do not follow their company’s attendance rules. Meanwhile, Flex Index data shows that 72 percent of companies with structured hybrid policies now mandate specific days rather than letting teams choose.

Put those together and the picture is clear: mandated days that a fifth of people ignore, that push out senior and skilled staff first, that make vacancies take longer to fill, and that do not show up in the profit line. That is not a policy. It is a tax on your best people.

The Study That Changed Managers’ Minds

The counterargument has always been productivity, and here the best evidence comes from the largest randomized trial on hybrid work to date. Bloom and colleagues at Stanford and the Shenzhen Finance Institute worked with Trip.com, a large travel company, to run a six-month experiment in 2021 and 2022 with 1,612 engineers, marketers, and finance staff, including 395 managers. Half were assigned to work from home on Wednesdays and Fridays. Half stayed in the office five days.

The results, published in Nature in 2024, found no difference in performance reviews over the following two years and no difference in promotion rates, overall or for any major subgroup. What did change was quitting. The hybrid group had a 4.8 percent attrition rate, compared with 7.2 percent for the office-only group — a drop of about a third. Among women, it was 4.2 percent versus 9.2 percent. Among people with commutes over 90 minutes, it was 2.9 percent versus 6 percent.

The part worth remembering is what happened to the 395 managers. Before the trial, they expected hybrid work to hurt productivity. After it, their views moved sharply positive. They had been wrong about their own teams, and the data corrected them. If you have a gut feeling that your people are less productive at home, the best randomized evidence on hybrid work says your gut may be miscalibrated.

Give the Office a Job

So mandates cost talent, and hybrid does not cost output. That does not mean the office is worthless. It means the office has to earn its keep. That is a design problem for managers, not a compliance problem for HR.

Ranya Nehmeh, an HR strategist who has held senior roles at the European Central Bank and the OPEC Fund and co-authored a book defending the office, argued in August 2026 that hybrid work works only when companies give people a clear reason to come together. Her point is that a three-day policy does not create collaboration, mentoring, or innovation on its own. Employees can all be in the same building and still work separately from their teams. The useful question is not where people should work but what the office is for.

Her list of what people do better in person is short and specific: brainstorming, problem-solving, mentoring, relationship-building, and learning by watching how experienced colleagues make decisions and handle conflict. That last one matters most for people early in their careers, who learn by proximity in ways that are hard to schedule over video. Some ideas come out of the hallway conversation after the meeting, not the meeting itself.

None of that requires everyone to be in every day. It requires building office days around the work that benefits from proximity. If your anchor day is a day of solo work and video calls with other offices, you have the cost of the commute and none of the benefit. Cancel the anchor day or redesign it. Half measures are the worst of both.

Let the Team Write the Rules

The second design lever is who decides the schedule. Gallup’s 2025 data split hybrid workers almost evenly three ways: 34 percent set their own schedule, 35 percent had their manager or team decide, and 31 percent had it set by senior leadership. The interesting question is which group did best.

Workers who set their own schedules, or whose teams set them, were more likely to see the arrangement as fair. But people with fully self-determined schedules also reported more trouble with burnout, fatigue, and work-life balance. The sweet spot was the middle: teams deciding together. Gallup’s own conclusion was that hybrid models work best when teams, not individuals or the executive floor, decide the rules, because shared norms make people more productive and less anxious.

That tracks with a Cisco report from mid-2025, which found that top talent ranked flexibility and clear communication as their top priorities and that hybrid arrangements work best when employees are involved in scheduling decisions. It also tracks with something Gallup flagged as a growing complication: the share of on-site employees whose teams are spread across different locations doubled from 13 percent in 2023 to 27 percent in 2025. If part of your team is somewhere else anyway, a blanket attendance rule does not put everyone in the same room. Only deliberate coordination does.

There is a management skill hiding in here. A study from Durham University and Athens University of Economics found that managers who lead with empathy — creating structure, showing consideration, and communicating their vision clearly — get better performance and productivity from remote and hybrid workers. Nehmeh said something similar on the record: policy matters less than manager quality. A good hybrid policy run by a poor manager fails. A mediocre policy run by a good one mostly works.

A Playbook for the Next Quarter

If you run a team or a function, here is what this adds up to in practice.

First, audit your anchor days before you defend them. Pick two recent in-office days and ask what happened on them that could not have happened remotely. If the answer is a team lunch and eight hours of headphones, you have a commute with no return on it. Managing hybrid teams starts with admitting that.

Second, replace the day count with a purpose list. Decide, with your team, which activities belong in person: onboarding new hires, quarterly planning, the messy first hour of a new project, hard feedback conversations, and the mentoring your junior people are not getting on video. Then schedule office time around those activities, and let the day count fall out of the calendar rather than the other way around.

Third, let the team set the norms and write them down. Gallup’s data says team-decided schedules beat both individual free-for-alls and top-down mandates on fairness and burnout. Put the agreement in a document people can point to, revisit it once a quarter, and fix what is not working.

Fourth, treat attendance data as a symptom, not a target. Kastle’s data shows office use remains strongest midweek and weakest on Fridays. If your Thursday is empty, the answer is not a stricter rule. It is a better reason to come in on Thursday or an honest decision to drop it.

Fifth, watch who is leaving. The Pittsburgh research says the people most likely to walk after a mandate are the senior, skilled, and women on your team. If your regretted attrition is skewing that way, your return-to-office policy is a likely cause, and no amount of engagement-survey wording will hide it. Employee-retention strategies that ignore this are working against themselves.

Finally, spend the effort on managers. The Trip.com managers changed their minds when they saw their own data. The Durham study says empathetic structure beats surveillance. Hybrid work best practices are not a policy document; they are a set of habits your first-line managers either have or do not have, and that is where the training budget belongs.

Conclusion

The where-do-people-work argument is over. The evidence now points in one direction: hybrid work has become the operating model for a large share of remote-capable employees, and forcing a broad reversal carries real costs. Mandates that fight the baseline push out senior and skilled staff, make vacancies take nearly a quarter longer to fill, and do not show up in profits. The strongest randomized evidence on hybrid work found no cost to performance or promotions and a meaningful reduction in quitting.

That leaves leaders with a harder, better question: what is the office worth when people get there? The answer will not come from a badge policy. It will come from managers who design in-person time around work that benefits from proximity, teams that set clear norms, and leaders who measure whether office days produce value. Counting days was the easy version of the job. Making those days matter is the real test.