Return-to-Office Mandates: What Actually Happens

Whether to require attendance is a legitimate business decision. How it is done determines whether it costs you a quarter of adjustment or your most mobile people.

This is not an argument against mandates. It is a description of what tends to follow one.

Who leaves first

Not a random sample.

The people who leave soonest after a mandate are those with the most options — senior individual contributors with scarce skills, and anyone who took the job on an explicit remote basis. They are also, disproportionately, the people whose departure is most expensive.

The people who stay are those with fewer alternatives. That is not a criticism of them; it is an observation about what a mandate selects for.

The second effect is quieter. People who relocated on the strength of a remote arrangement, or arranged childcare around it, cannot simply comply. For them a mandate is not an inconvenience but a demand to reverse a life decision, and the response is usually a job search rather than a negotiation.

The consent problem

The damage is rarely the requirement itself. It is the reversal.

An organisation that hired people as remote, encouraged relocation, then required attendance has changed the terms of a bargain unilaterally. Even people who can comply notice what that means about future commitments.

This is why mandates frequently produce more attrition than their attendance requirement would suggest. People are not leaving over two days a week. They are leaving because they learned what a written arrangement is worth.

If you are going to do it

State the reason, specifically. Not "collaboration". What activity is failing, how you know, and how attendance addresses it. A reason that survives scrutiny is worth a great deal; a slogan invites people to supply their own explanation, usually a worse one.

Give real notice. Three months minimum for anything affecting where someone lives. Less than that is a resignation request with extra steps.

Grandfather explicitly remote hires. People hired on a stated remote basis are a different case from people whose arrangement drifted. Treating them identically is the fastest way to lose them and to teach everyone else that written terms are provisional.

Decide at team level. A company-wide mandate applied to teams with different work produces obvious absurdities — people commuting to sit on video calls with colleagues elsewhere — and every absurdity erodes the rationale.

Make the office worth the journey. If people arrive to a hot-desk floor and a day of video calls, the mandate demonstrates its own pointlessness weekly. Coordinate teams onto the same days. Ensure the people someone needs are actually there.

Say what happens to non-compliance before it arises. An unenforced mandate is worse than none: it rewards the people who ignore it and punishes those who complied at cost.

Measure something

Most mandates are introduced with a stated goal and never assessed against it.

Decide in advance what should improve, over what period, and check. If collaboration was the reason, what measure of collaboration moved? If it did not, that is information — and reversing a policy that did not work is more credible than maintaining one that visibly did not.

The version that works

The mandates that cause least damage share three features: a specific reason tied to specific activities, generous notice, and consistency between what people were told when hired and what is now required.

The ones that cause most damage are announced abruptly, justified in generalities, applied uniformly across dissimilar work, and enforced unevenly.

The difference is not the policy. It is whether people can see the reasoning and whether the organisation kept its word.