Microsoft Offered 60,000 People a Way Out, Not a Layoff: the 3 Reasons I Rebuilt Our January Hiring Plan

My first reaction on Monday evening was the lazy one. Sixty thousand people, a quarter of a workforce, the word Microsoft in the headline: that reads like the biggest talent event of the quarter, and the reflex in our business is to open a spreadsheet and start sourcing. I spent Tuesday reading the actual terms instead, and ended up rewriting our January plan rather than accelerating it. The reason is one word in the announcement that most summaries skipped over: voluntary.

What Microsoft Actually Put on the Table

On 26 October 2026, Microsoft opened what it calls a Voluntary Exit Offer. TechCrunch and CNBC both put the eligible pool at roughly 60,000 people, which is about a quarter of the company. GeekWire, reporting on the internal memo from chief people officer Amy Coleman, adds the detail that decides everything else: the offer carries no age or tenure requirement and performance is not a factor, and the ceiling is internal level 70, which keeps corporate vice presidents out but sweeps in a great deal of senior individual-contributor and manager-level engineering.

Geography is narrower than the headline suggests. The programme runs in the United States, Canada, India and the United Kingdom. It is also not company-wide in any real sense: GeekWire reports Coleman telling staff that some organisations are running the programme and others are not. Bloomberg names Xbox, Windows, Copilot and Finance among the participating divisions, and reports data centre, sales and marketing staff as excluded. That exclusion list is the single most useful line in the coverage for anyone hiring, and I will come back to it.

On money, the public position is that the package is equivalent to standard Microsoft severance: continued stock vesting and healthcare for a fixed period, plus career services. TheStreet and SiliconANGLE both describe it that way. Secondary reporting of the severance grid, which Microsoft has not published, puts one week of pay per six months of service at the lower levels, two weeks per six months in the middle band, a cap in the high thirties of weeks for most eligible levels, and a materially larger package at the top of the eligible range. Treat those figures as reported rather than confirmed. The direction is what matters: this is months of runway, not a fortnight.

Then the calendar. Deadline and GeekWire both report an early-December decision deadline and a 31 December leaving date for participants. A Microsoft spokesperson told reporters the company is not planning a further buyout after this one. For context, this follows a first voluntary retirement programme earlier in 2026 and a round of roughly 4,800 involuntary cuts in July, and the framing across coverage ties it to the efficiency pressure created by the company’s AI and data centre spending.

Why Voluntary Breaks the Layoff Playbook

Every piece of advice written about hiring from tech layoffs over the last four years rests on an assumption nobody says out loud: the candidate is under pressure. They were cut without warning, severance is thin or already spent, and the clock on their visa, their mortgage or their self-esteem is running. That pressure is what makes a fast process and a merely competitive package work.

Remove the pressure and the whole thing inverts. A person who opts into a voluntary exit in early December has done arithmetic before they clicked accept. They know what their severance is worth, they know their stock keeps vesting for a defined window, they know their healthcare runs for a fixed period, and they chose the date they stop working. They are, in the most literal sense, a passive candidate with a funded sabbatical. Three consequences follow, and all three are uncomfortable.

First, speed stops being a differentiator. The pitch that wins against a layoff cohort is “we can have you signed in ten days”. Against this cohort that pitch reads as desperation, or worse, as a signal that you have a hole to plug rather than a role to fill. Nobody who just negotiated themselves six months of runway is impressed by urgency.

Second, your package gets compared properly. These candidates have time to run three or four processes to final stage and put the offers side by side. That is fatal for the Dubai pitch most employers still default to, which leans on the headline number and the absence of personal income tax and stops there. It is also an opportunity, because the tax-adjusted comparison genuinely does hold up, but only if you do the arithmetic for the candidate in writing rather than implying it.

Third, self-selection cuts both ways. People who volunteer to leave a company of that size in a programme with no performance filter are not a random sample. Some are the people you most want: senior engineers who had already decided the next two years of their roadmap were not interesting, who have been quietly preparing to move. Some are the opposite: people who were already disengaged and have now been handed a paid exit. The screening problem for this cohort is not competence. It is motivation, and it is a harder thing to interview for.

The Three Things I Changed in Our Own Plan

We had three engineering roles pencilled in for a January start. None of the role descriptions survived Tuesday. Here is what moved.

1. We stopped writing for stack and started writing for scale of ownership. The original description for our backend role listed frameworks. That filter is useless against this cohort, because somebody who spent nine years on a Windows component or an Xbox service may not have touched the framework in your job post and will still out-engineer every candidate you have seen this year. The rewrite asks for a different thing: a system you did not build, that you were responsible for, that had users who complained. Then the interview goes after that story in detail. Our interview question sets needed the same surgery.

2. We wrote the money down, in both currencies, with the tax line visible. Not a range. A single number, split into base, allowances and any variable, converted at a stated rate, with the UAE position on personal income tax noted as a line item rather than as a sales pitch, and with the one-off costs we cover for the move listed separately. A candidate comparing four offers over a Christmas break will reward the one they can actually read. If you want the arithmetic before you write the offer, the salary calculator and team cost calculator get you to a defensible number faster than a spreadsheet will.

3. We moved the visa conversation to the first call. This is the part Dubai employers consistently under-sell. A candidate in the United States on a work visa, or in the United Kingdom weighing a move, is carrying immigration anxiety that has nothing to do with your role. The UAE offers long-term residency routes for qualifying professionals, and the processing timeline is measured in weeks rather than visa lotteries. Saying that out loud on the first call, with the specific route you intend to use and who pays for what, removes more friction than another five per cent on base. Our UAE hiring guide covers the ground we now walk candidates through on call one.

The Motivation Screen This Cohort Actually Needs

Since competence is not the bottleneck here, the screen has to go somewhere else. Four questions have done the work for us, and none of them are clever:

What This News Is Not

Two misreadings are already circulating, and both will cost somebody a quarter.

The first is treating 60,000 as a number of available people. It is the number of people eligible to consider an offer, in a programme that is opt-in, in four countries, on a subset of teams. Voluntary programmes historically convert at a fraction of the eligible pool. Nobody outside Microsoft knows what that fraction will be here, and anybody publishing a confident figure for it this week is guessing. Plan for a good cohort, not a flood.

The second is reading this as a tech downturn signal and slowing your own hiring in response. The framing across coverage is the opposite: this is a cost reallocation to fund AI and data centre commitments, by a company that is spending more, not less. Fortune and Axios both place it in that frame. The headcount is moving from one kind of work to another, and the same reallocation is happening inside plenty of UAE businesses right now. If anything, it is an argument for getting your own engineering plan explicit before the people you want are deciding between four offers.

FAQ: Microsoft’s Voluntary Exit Offer and Dubai Hiring