How to Ask for a Raise. (And Why the Annual Review Is the Worst Time.)

A raise is not granted to you. It's negotiated, and the single worst moment to negotiate it is the one everyone uses: the annual review. By the time your manager sits across from you in spring, the money is already gone. The budget got locked the previous fall, divided among the team, and signed off before anyone scheduled a single review meeting. You're not asking for a raise in that room. You're asking for a redistribution of money that's already been promised to other people.

Most workers believe the review is when the company decides what you're worth. They prepare for it like an exam, hoping a strong year speaks for itself. It doesn't. The real driver isn't your performance summary. It's the timing of the budget cycle and whether you made a specific business case before that budget closed. The people who get the biggest raises aren't the best performers. They're the ones who asked at the right moment, with the right number, for the right reason.

Why is the annual review the worst time to ask?

Because the budget is already locked. The conversation you think is a negotiation is actually a notification.

Compensation budgets get finalized in the fall and early winter, long before review season. By October 2025, just over 60% of organizations had already presented their 2026 salary-increase budgets to leadership for approval or secured it outright — months before the spring raise conversation most employees treat as the decision point. Mercer's own planning guidance flags a November checkpoint "to finalize your plans", with merit budgets projected at 3.2% and total increase budgets at 3.5%. That 3.2-3.5% envelope is the ceiling. When your manager sits down with you in March, they're not deciding what you deserve. They're dividing a pie that was baked in November.

So the review isn't where you win. It's where you find out whether you won months ago. If you want to actually move the number, you have to be in the conversation before the pie gets cut — which means asking off-cycle, with a case strong enough to make them carve out budget that wasn't planned for you.

Does asking actually work, or does it just annoy people?

It works, and it works at a rate that should embarrass anyone still waiting to be noticed.

In a survey of over 2,000 full-time workers, 82% of those who asked for a pay raise in the past year received one, versus 66% who got a raise whether or not they asked. The Federal Reserve's own data tells the same story from a government source: of workers who asked for a raise in 2023, 66% received one — yet only 13% of workers asked at all. Sit with those two numbers. Two-thirds of people who ask succeed. Almost nobody asks.

That gap is the entire opportunity. The reason most people don't get raises isn't that they were denied. It's that they never opened their mouth. They waited, assumed good work would be recognized, and let the budget cycle decide for them. The 13% who ask aren't braver in some abstract way. They've just figured out that the default — silence — pays the worst.

Behavior	Share who do it	Outcome
Asked for a raise	13% of workers	66-82% got one
Waited to be offered	the other 87%	got the 3.2-3.5% budget rate, or nothing
Stayed put through inflation	most workers, 2021-22	real earnings down 3.6%

That last row is the quiet tax. Not asking doesn't keep you flat. It moves you backward.

What number should you actually ask for?

A specific one, anchored to the market and the business — not a percentage of your current salary. Asking for "a 10% bump" tells them you're negotiating against your own paycheck. Asking for a market-rate number tells them you're negotiating against what the role is worth.

Start from the data. Find what the role pays from sources that aren't optimistic self-reports — published pay ranges in current job postings, a tool like Levels.fyi for tech, and one or two people actually doing the job. One catch: most postings don't disclose pay at all. Across Praxy's live index of 53,802 active job postings (June 2026), only about 9% list a salary — and among those that do, the median advertised ceiling runs roughly $122,000 in the US (n≈1,980 US postings with disclosed pay). So the published range exists, but you have to actively dig for the listings that show it rather than assume the first number you find is the market. Market rate is a range, not a single number, so your job is to find where you sit in that range and ask for a point slightly above the midpoint, justified by what you specifically deliver. Then frame the gap, not the percentage:

Weak (anchored to your own salary, easy to shrink): "I've been here two years and worked really hard, so I was hoping for about a 10% raise this year."

Strong (anchored to the market and the business): "Based on the market for this role — postings in our area are landing between X and Y for the same scope — and the fact that I now own [specific revenue-driving or cost-saving thing], I'd like to bring my base to [target number]. I want to make sure my comp reflects the work I'm actually doing now."

The second version hands your manager the exact case they'll use to get it approved upstairs. They don't fight your boss with "she works hard." They fight with "she's below market for the scope she owns and we'll lose her to replace her at the same number." You're not asking them to like you more. You're writing their internal memo for them. That's the whole mechanic of getting paid before the negotiation even starts: the decision is made by the case, not the conversation.

Why does a business-case ask beat job-hopping right now?

For years it didn't. For most of 2022 and 2023, the fastest way to a raise was to leave: workers who changed jobs saw median wage growth roughly 2 percentage points higher than those who stayed, per the Atlanta Fed Wage Growth Tracker. The switcher premium was real, and it was large.

That edge has collapsed. In the latest data, wage growth for job stayers (4.4%) actually edged out job switchers (4.2%) — the first time stayers outpaced switchers since September 2010. The market that made quitting the obvious move has flipped. When the switcher premium is gone, the math changes: an internal raise that costs you no interviews, no relocation, no first-90-days risk, and no loss of accrued equity or tenure can now beat the external move outright. And leaving without a clean reason carries its own penalty — job-hopping without a story is the trap — so the headline is simple: the leverage moved back inside the building, and most people haven't noticed.

This is also why the off-cycle business case matters more than ever. You no longer need a competing offer to justify a raise, and chasing one is now often the slower path — the market built the case for you. Negotiate hard with the evidence you already have instead of waiting on an external offer that may never pay off.

How do you handle "the budget is locked" or a flat no?

You expect it, because you already know it's often true — you read the section above. A no on the headline number is rarely a no on everything. It's the start of the second conversation, not the end of the first.

When they say there's no budget, you have two moves. The first is to pin the future, so this isn't a dead end:

"I understand the budget for this cycle is set. I'd like to agree now on a specific number and a date — say, a review at the start of next quarter tied to [defined deliverable] — so when the budget opens, I'm already in the plan, not at the back of the line."

The second is to widen the ask beyond base, because base is the one line most tied to the locked budget:

"If base can't move this cycle, can we look at a one-time bonus, an equity refresh, a title change, or a four-month review with a defined bump? I want to find the piece that does have room."

If base is locked, ask for	Why it's often easier to move
One-time / spot bonus	Comes from a different budget line than merit
Equity refresh	Not always inside the annual salary band
Title change	Resets your market value for the next ask
Defined review at 3-4 months	Costs them nothing today, commits the future
Off-cycle merit adjustment	Some companies hold a reserve outside the main budget

A "no" on base is frequently a "yes" somewhere else — but only if you ask for somewhere else, calmly, in the same breath, instead of deflating.

When should you not ask — and why does waiting still cost you?

Here's the part the cheerleading posts skip. Asking has a cost and a wrong time, and pretending otherwise gets people hurt.

Don't ask in the week your team missed a major target, the month after a layoff round, or while your manager is visibly fighting for headcount they're about to lose. A business-case ask depends on the business having a case to say yes to. Read the room. If the company is contracting, the honest move may be to bank your evidence and wait for a stabler quarter — or to start looking, because the best time to job search is while you're still employed. And if you're a strong performer in a genuinely broke budget, a raise may not be available at any timing, full stop. That's information about the company, not about you.

But understand what waiting actually costs, because "I'll ask next year" is not neutral. Over the 12 months ending June 2022, real average hourly earnings fell 3.6% — nominal pay rose 5.2% while CPI inflation ran 9.1%. That wasn't a one-month blip either; real earnings were down 3.0% over the 12 months ending May 2022 on the same pattern. A standard 3-5% "raise" in that environment was a pay cut you said thank you for. This is the trap nobody frames honestly: the annual raise frequently doesn't beat inflation, so doing nothing isn't holding steady — it's a quiet, compounding loss. Asking off-cycle for a real market number is how you stop subsidizing the gap with your own purchasing power.

What to do now

1. Find the budget cycle, then ask before it closes. If your company runs a calendar-year budget, the money locks in the fall. Put your ask in late summer or early fall, before the 60%+ of orgs that finalize budgets by October close theirs — not in the spring review.
2. Set one target number from two non-self-reported sources. A published pay range plus one person in the role beats any Glassdoor average. Place your ask slightly above the midpoint for your scope.
3. Write the business case, not the personal one. One sentence on what you now own that you didn't before, one on what it's worth, one on where market rate sits. That's the memo your manager forwards upstairs.
4. Pre-load the fallback before the meeting. Decide your base number, your acceptable bonus/equity/title alternative, and your "agree on a date now" line — so a "budget's locked" answer moves the conversation forward instead of ending it.
5. Ask once, with the number and the reason, then stop talking. The silence after a specific ask is doing work. Let it.

Want to pin down the exact number to ask for, find your real market rate, and rehearse the ask — including the comeback when they say the budget's locked? That's what I'm built for. Message me on WhatsApp and we'll build your business case, set the number, and roleplay the conversation until it stops feeling scary.