Microsoft required Puget Sound employees to work on-site three days a week starting in February, covering anyone who lives within fifty miles of a Microsoft office. UPS mandated five days for all thirty-six hundred of its corporate workers beginning in March, making it the largest public company in Georgia to eliminate remote work. Amazon had already moved to five days for its corporate staff. The list of major employers issuing formal in-office requirements has grown through 2026, and the direction of the trend is not ambiguous.
What is more complicated is what happens after the policy is announced.
The gap between policy and practice
Cisco data published earlier this year found that forty-nine percent of employees globally are now mandated to return five days a week. Another forty-five percent are under hybrid arrangements with one to four days required on-site. Those are mandate figures, not attendance figures. The gap between what a policy requires and what a manager actually enforces is where most of the action is.
IBM's situation illustrates the complexity. The company's broader return plan has been described by insiders as operating as a phased ultimatum across corporate roles, with US employees expected to co-locate or resign. The enforcement, they say, hits older and more expensive staff particularly hard. IBM did not confirm that characterization.
Research from JLL found that the majority of employees who feel positive about their workplace also feel positive about attendance policies, while forty-eight percent of employees who are unhappy with their workplace feel negatively about mandates. The policy is not the variable most companies can easily change. The workplace experience is.
What the data says about outcomes
The empirical case for in-office work is harder to make cleanly than the mandates imply. Proponents point to collaboration, mentorship, and the kind of ambient organizational knowledge that does not travel well over video. Skeptics note that productivity measures during the remote-work period did not collapse in most knowledge-work categories and that the workers most likely to leave over a mandate are the ones with the most options.
Microsoft's own framing acknowledged the tension. Amy Coleman, the company's chief people officer, said the change was "not about reducing headcount" but about "working together in a way that enables us to meet our customers' needs." UPS's statement pointed to the relationship between in-person culture and customer service. Both framings avoid the productivity argument and make a softer claim about connection and culture.
What companies are learning is that the mandate is the beginning of the conversation, not the end. The attention budget question does not resolve itself by moving people to a building. It requires intentional design of how the hours in that building are spent. Organizations that issue a mandate without redesigning the meeting culture and workspace experience are solving the presence problem while leaving the productivity problem untouched.
The return-to-office debate has not resolved into a settled answer. What 2026 has produced is a clearer picture of what enforcement actually costs, in attrition and morale and management time, and a growing recognition that the three-day and five-day mandates are experiments, not conclusions.
Topics leadershipworkplacebusinessmanagement
