My offer letter was dated March 3. The email announcing the restructuring that dissolved my team went out on May 19. Seventy-seven days.
I had done the homework. By then I'd been following tech layoffs the way other people follow a sport — the trackers, the screenshots, the LinkedIn posts that start "I'm one of the ones affected." Before I signed, I went and read the company's own announcement from the previous autumn, the one where they cut eight percent and called it sharpening focus. It read fine to me. Regretful. Specific about severance. Signed by a founder. Adult.
It was not fine, and I couldn't tell, because I was reading it for tone. Tone is the part that gets written last and edited most. I've since read that memo against about sixty others I've saved in a folder since 2022, and the thing it did not contain turns out to be the only thing that would have helped me.
So this is a comparison, not a lecture. Three kinds of restructuring announcement, side by side, judged on named criteria, with a verdict at the end about which one you can actually plan around when you're deciding whether to apply, whether to take the offer, or whether to keep the recruiter warm and stay where you are.
One disclosure up front, and only once: there are no affiliate links in this piece and nothing here to buy. I'm not selling you a course on reading corporate communications. The three memos below are composites. I've reworded them so I'm not putting invented sentences in a specific real executive's mouth, but the structures are real, the patterns repeat, and if you've read more than a dozen of these you will recognize all three within two paragraphs.
What the three memos actually are
Sort enough announcements and they stop looking like sixty documents and start looking like three, wearing different clothes.
The Cost Memo says the environment changed and we grew too fast. The AI-Era Rewrite says the company we built no longer matches the company the market now rewards. The Shape Memo says the org chart itself is the problem — too many layers, decisions moving too slowly — and describes surgery on the structure rather than on the budget.
They overlap. A good memo from a well-run company often does two at once. But the dominant mode is usually obvious, and it tells you different things about what happens next.
Memo A: The Cost Memo
The Cost Memo is the oldest form and the most honest-feeling. Here's the shape of one, compressed:
Today we are reducing our team by 12%, roughly 340 roles. We hired against a demand curve that did not hold. That call was mine. Everyone affected will receive 14 weeks of base pay plus two weeks per year of service, healthcare through the end of the year, and accelerated vesting on the next tranche. Immigration counsel is available starting today.
What this memo is best at: it is checkable. There's a percentage and a headcount. There's a severance formula you could put in a spreadsheet. There's an admission of cause that doesn't route through the market, the macro environment, or the abstract noun "headwinds." When somebody writes "that call was mine," they've made a small hostage of themselves, and people who make hostages of themselves tend to be telling you something closer to true.
For a job seeker, the Cost Memo is legible. If the company overhired against a demand curve, and the demand curve is knowable — enterprise seats, ad spend, GPU contracts, whatever the business actually sells — you can go look at the demand curve yourself and form a view about whether the cut was sized correctly.
Here's the honest negative, and it's the one that cost me: the Cost Memo never tells you whether the cutting is finished. It is structurally incapable of it, because no general counsel on earth will let a CEO write "this is the last one" unless they can guarantee it, and nobody can guarantee it. So the memo ends on a forward-looking paragraph about the road ahead, and you read that paragraph as closure because you want closure.
In my folder, of the memos that read as pure Cost Memos, a clear majority were followed by another reduction inside twelve months. Not because leadership lied. Because a cut sized to a spreadsheet is sized to that quarter's spreadsheet, and there's another spreadsheet coming. The memo that hired me was a Cost Memo. It was accurate, it was decent, and it was the first of three.
Memo B: The AI-Era Rewrite
This is the form that has taken over. Compressed:
We are making the hardest decision in our history and parting with roughly 20% of the company. This is not a cost exercise. The structure that carried us through the last chapter is not the structure that wins the next one. Our customers are adopting AI-native workflows faster than any product shift we have measured, and we intend to build for that reality rather than defend the one we already have. To those leaving: this is not a judgment of your work.
Read it cold and you'll notice it does something the Cost Memo doesn't. It offers a reason that isn't arithmetic. It reframes a subtraction as a redirection. It also, almost always, includes the sentence about how this doesn't reflect on the performance of the people leaving — which is genuinely worth saying, and is also the single most-repeated sentence in my entire folder.
The real strength, and I want to be fair here: the AI-Era Rewrite is often the most truthful memo about direction. It tells you which functions the company has decided are now cheaper, and that information is worth money to you. When a company says it's rebuilding around AI-native workflows, go look at which orgs took the deepest cuts. It's usually support, content, QA, junior-heavy delivery pods, and the middle of the sales organization. That is a live signal about which roles are being repriced across the industry, delivered by someone with a financial incentive to be right about it. Take it seriously even when you're angry about it.
The honest negative is that almost nothing in it can be falsified. "The structure that carried us through the last chapter" is not a claim. You cannot check it in six months and find out whether it was true. Compare that to a severance formula, which either arrives in your account or doesn't. When a memo's central assertion is unfalsifiable, the memo has stopped being information and started being positioning — usually positioning aimed at analysts and the press, with employees reading over their shoulder.
And the second negative, which matters more if you're weighing an offer: the AI-Era Rewrite is the easiest form to put over an ordinary budget cut. The words cost nothing. A company that missed its number and a company genuinely rebuilding its product surface can file identical prose. Sometimes the only way to tell them apart is to notice that one of them also shipped something.
Memo C: The Shape Memo
The rarest of the three, and the strangest to read. Compressed:
We are removing two management layers between the people building the product and the people accountable for it. Teams will move from an average of 6 reporting steps to 4. Product, design, and engineering leads will own their roadmaps end to end, including hiring. This affects roughly 260 roles, the majority of them in management and program coordination. We are also stopping work on three initiatives — [named] — so the remaining teams inherit fewer commitments, not more.
That last clause is why I've come to weight this form heavily. It contains subtraction: a named list of work that stops. Almost no memo in my folder contains subtraction. Announcing what you will no longer do is embarrassing in a way that announcing headcount is not, because it means admitting that something you told the market about last year is dead.
The Shape Memo describes a mechanism. Fewer approval steps, fewer coordination roles, fewer commitments per team. Those are things a person joining in four months can verify from the inside within two weeks, and things a person interviewing can probe: how many people between you and the person who decides what you build, and what got cancelled this year.
The honest negative is sharp. "Flatter, with fewer layers and more ownership" is also exactly how you describe deleting every manager who would have hired you, sponsored you for promotion, and argued for your headcount in the room you're not in. If you came up outside the FAANG pipeline, you know what a manager who advocates for you is worth, because you've mostly worked without one. A flat org is a wonderful place to be if you already have internal reputation. It's a hard place to be new.
The second negative: in practice, "greater ownership" often means the surviving senior engineers absorb the program management, the incident comms, and the stakeholder updates that the deleted layer used to carry. The work doesn't disappear because the coordinators did. Ask, in the interview, who runs the release now.
The four criteria I run every memo through
This is the part I'd have wanted seventy-seven days before May 19.
1. The denominator. Not "hundreds of roles." A percentage, a headcount, and of what — global, or one org? A memo that gives you 12% of 2,800 has handed you a number you can hold against the next announcement. A memo that says "a small number of positions" has decided you don't get to do arithmetic. Vagueness in the denominator is the cheapest possible tell, and it is almost never accidental.
2. Subtraction. Does the memo name work that stops? This is my heaviest criterion and the one nearly everything fails. If headcount goes down and the commitment list doesn't, you are reading a document about the same roadmap with fewer people on it, no matter how the paragraph about focus is phrased. If you're evaluating an offer at a company that recently restructured, this is the question to ask the hiring manager directly: what did we stop doing? If they can't name something, the survivors are carrying it.
3. Severance specifics. Weeks of pay, healthcare end date, equity treatment, immigration support, and whether the notice period is paid. Specificity here is a proxy for whether the company planned this or panicked. It is also the one part of the memo that has to be true, because it's a promise to people who will publish it if it's broken. In the US, check whether a WARN notice was filed with the state — those are public, searchable, and often list role titles and counts the memo rounded off.
4. Falsifiability. Find the forward-looking claim and ask whether you could check it in two quarters. "We expect the new structure to reduce time from decision to ship" is checkable. "We are positioning for the era ahead" is not. Memos heavy on unfalsifiable claims aren't necessarily dishonest; they're aimed at a different audience than you.
Here's how the three forms score:
| Criterion | Cost Memo | AI-Era Rewrite | Shape Memo |
|---|---|---|---|
| Names a denominator | Usually | Sometimes — often percentage only | Usually |
| Names work that stops | Rarely | Rarely, despite the framing | Sometimes, and it's the point |
| Severance specifics | Strongest of the three | Mixed; often deferred to HR | Mixed |
| Falsifiable claims | Moderate | Weakest | Strongest |
| Predicts another cut soon | Often | Often | Less often |
| Best signal it gives you | The size and honesty of the hole | Which roles the market is repricing | Whether the company you'd join actually changed |
So the verdict, since I said I'd give one plainly: if you are deciding where to send applications or which offer to take, the Shape Memo with named subtraction is the only one of the three that reliably describes a company that will be recognizable in six months. The Cost Memo is the most honest document and the least predictive one — it tells you what happened, not what's next. The AI-Era Rewrite is the weakest evidence about that specific company and the strongest evidence about the industry; read it for the map, not the territory.
And if a memo does none of the four — no denominator, no subtraction, severance handled offline, nothing falsifiable — that isn't neutral. That's a company that has decided its employees are an audience rather than a constituency, and you should price the offer accordingly.
What the memo will never tell you, and where it's actually written
The memo is the marketing. The disclosure is the accounting.
If the company is public, the restructuring shows up in the next quarterly filing as a charge with an actual number attached, sometimes broken into severance versus contract termination versus impairment. A restructuring that's mostly severance is a headcount event. A restructuring carrying large contract-termination or write-down charges usually means something got killed — which is the subtraction the memo declined to name. That document is free and it was written by people with legal liability for its accuracy.
Then there's the job board. Six weeks after a restructuring, go look at what the company is hiring for. A company that cut 20% and is posting forty roles in the same function it cut is doing wage arbitrage or geographic relocation, not restructuring. A company that cut 20% and is posting eight roles, all in one new area, has told you where it actually went. Nobody announces this. It's sitting on the careers page.
And if you're interviewing there: ask the hiring manager how many people were on this specific team a year ago. Not the company — the team. I've had that question answered honestly enough times to believe it's worth the small awkwardness. Twice it changed my answer.
Who this is for, and who it isn't
This is for you if you're weighing an offer from a company that restructured in the last year and everyone in your life is telling you to take it because the market is what it is. They're not wrong about the market. They're also not the one who'll be on the other side of the next email. Run the four criteria, ask about subtraction in the interview, look at the filing.
It's for you if you're building an application list and trying to decide whether a company that made the news is worth your hours. It probably is — post-cut companies hire, and they hire people who ask sharp questions about what changed.
It is not for you if you're inside a company right now trying to divine whether you're on the list. I've been there, refreshing my calendar for meetings I wasn't invited to, and reading the memo forensically didn't help me at all. Nothing in a public announcement forecasts an individual name. Spend that energy on your resume and on three conversations with people who left last quarter, who will tell you more in twenty minutes than any memo will.
And it isn't for you if what you want is reassurance that a well-written announcement means a stable employer. It doesn't. The best-written memo in my folder came out of a company that cut twice more inside a year. Prose quality measures the communications team.
Seventy-seven days
I went back to the memo that hired me — the one I'd read for tone and found decent — and I read it against the four criteria.
Denominator: present, eight percent of a named headcount. Severance: specific, generous, honored in full. Falsifiability: thin, but not empty. Subtraction: nothing. Not one line about work that would stop. Every initiative from the prior year's roadmap was still standing in that memo, being carried by ninety-two percent of the people.
That was the paragraph I skipped, because it wasn't there to be read. Absence doesn't announce itself; you have to go looking for it, which means knowing in advance what you're looking for. Seventy-seven days wasn't the company blindsiding me. It was how long it took a smaller group of people to fail to do the same amount of work, and the arithmetic on that was published in the autumn, in the space where a sentence should have been.
Read the memo for what it declines to subtract, and it will tell you how long you have.
