Competing Offers at Google, Meta and Big Tech: Proof, Matching and Timing

You have one offer and a second process that is almost done. Or you have two offers and a recruiter who just asked the question every engineer dreads: do you have that in writing?

Competing offers are the strongest leverage in a tech negotiation, and also the easiest to misplay. Companies differ in what they ask for, what they will match, and how fast they move. This guide covers how proof requests work, who matches whom, how to compare packages that vest differently, and the order to play your offers in.

The short version: Google often asks for written proof, Meta usually doesn’t unless your numbers look unusual, and nobody matches dollar for dollar. Compare offers on four-year value, not year one, and start your slowest process first so the offers land together.

Who asks for proof

No big tech company publishes a policy on this. What is known comes from press coverage and from consistent candidate reports.

Company	What candidates report
Google	Often asks for written proof, such as the letter or a screenshot of it. The Information reported in 2021 that this became common practice, and candidates still report it. Without proof, comp may not try to match
Meta	Usually works from a verbal breakdown; asks for documents mainly when numbers are outside the band Meta knows for that company and level
Apple	Mixed: verbal is often accepted, and some recruiters ask to see the letter
Microsoft	Reported to ask for proof, and to respond mainly to offers from FAANG-tier competitors
Amazon	Named in the 2021 reporting as also asking for proof
Nvidia	Mixed: some candidates asked, others negotiated without it
Anthropic, OpenAI	Reported not to negotiate the number much at all, so the question is less about proof and more about level

Two practical points follow. First, assume the question will come at Google, and have the compensation section of your competing offer ready to share, with personal details removed. Second, many companies won’t put an offer in writing until you verbally accept, and recruiters generally understand that. If all you have is a verbal offer, say exactly that and give the components you were quoted.

What the recruiter actually needs is narrow: the company, the level, and the components (base, equity and its vesting schedule, bonus target, sign-on). Everything else can stay private.

Nobody matches dollar for dollar

The most common mistake with a competing offer is expecting a match. What usually happens is a partial move, and the shape of that move depends on the company.

Google is the clearest example. Candidates repeatedly report Google matching roughly the first two years of a Meta package and pointing to future refreshers to cover the rest. Others describe Google countering below the competing number and relying on its brand. And Google is reported to move fastest, often within a day or two, against Meta and other top-tier competitors, while dragging on offers from companies it considers less of a threat.

Microsoft is reported to negotiate mainly against offers from FAANG and similar companies, with large stock increases needing heavy approval.

Amazon builds its offers around a back-loaded vest (5/15/40/40 over four years) offset by two years of sign-on cash, and recruiters have wide discretion on that sign-on. When you bring Amazon a competing offer, the sign-on is where you should expect movement.

AI labs that standardize pay by level, such as Anthropic, are reported not to negotiate the number at all. A competing offer still matters there, as calibration evidence for a higher level and support for a sign-on, but it won’t move the band.

So set your expectation as “a meaningful move toward the competing number”, and decide in advance which offer you’d take if each company comes back with its likely partial match.

Compare four-year value, not year one

Big tech vesting schedules are different enough that the same grant can look very different in year one.

Schedule	Who uses it	What year one hides
Front-loaded	Google and some DoorDash grants (40/30/20/10)	Year one overstates the package; later years drop unless refreshers fill the gap
Even 25% a year	Meta and many others	Year one is representative
Back-loaded 5/15/40/40	Amazon	Year one understates the equity; sign-on cash fills years one and two

When you put two offers side by side, compute both year-one total comp and the four-year total, then the average. A front-loaded offer can win year one by a wide margin and lose on the four-year average, and recruiters at the company with the less favorable schedule know it. Their counter will usually be framed in whichever year makes them look best.

Private equity needs a discount on top of that. Levels.fyi reports that Meta may discount startup equity by as much as 75%, and in general private stock is valued well below liquid public RSUs. A startup offer is real leverage, especially on base and sign-on, but don’t present its equity at face value.

The order you play your offers in

Most competing-offer leverage is won or lost in timing, before any negotiation starts.

1. Start the slowest process first. Google’s hiring committee and team matching can take weeks or months, while Meta and Amazon move in days. Start the slow ones early so the offers can overlap. (Our guide to big tech offer deadlines has the timelines.)
2. Set your own decision date and tell every company. A clear date gives every recruiter a reason to move, and it makes your timeline a fact rather than a request.
3. Don’t share expected numbers early. Stay specific about your timeline and vague about your number until you have offers to compare.
4. Raise the competing offer as high as you can before you share it. Negotiate the first offer to its best version, then bring that version to the second company.
5. Bring the competing offer to the company you’d rather join. Use the strongest number to improve the offer you actually want, not to win an auction you don’t intend to finish.

Without a competing offer

You can still negotiate, but the levers change. Meta is reported to move very little without another offer, though a sign-on is often still on the table, and Meta may consider unvested equity you’d leave behind. Google sign-ons can be negotiated even when the first offer has none. At companies that standardize pay by level, the level is the negotiation.

The common thread: when you don’t have a competing number, anchor on something documented. Unvested stock and a bonus you would forfeit by leaving are both real, verifiable costs, and they support a sign-on ask at almost any company.

The one rule: never bluff

Comp teams know what other companies pay at each level, because they see those numbers from candidates every week. A claimed offer that doesn’t fit the known band invites a request for the letter, and a bluff that gets called costs you more leverage than having no competing offer at all. Secondhand reports describe offers withdrawn after exactly that.

The proof request is a test, not a formality. “Do you have a competing offer in writing?” is one of the most common recruiter pushbacks, and the way you answer it decides whether your leverage survives the conversation, whether or not you have a letter to send. SalaryScript is the 125-page playbook for that exact back-and-forth: a counter-move for every recruiter pushback, including this one, word for word. Fold here and the gap doesn’t stay in this offer: it becomes the base every raise and refresher is measured from. Get the Bundle · $129 → (instant download, 14-day results-based guarantee).

The bottom line

• Expect Google to ask for proof, and have a redacted compensation section ready.
• Expect a partial match, not a dollar-for-dollar one, and decide in advance what you’d accept.
• Compare offers on the four-year total and average, not year one, and discount private equity.
• Start slow processes first and set your own decision date.
• Never bluff. Use what you actually have, and anchor on documented costs when you have nothing else.

Holding offers from specific companies? The company playbooks cover each one’s levels, equity and levers: Google, Meta, Amazon, Waymo, Anduril and DoorDash.