Guide
How much of a raise should you ask for?
The short answer
Ask for what the role is worth, not for a percentage of what you're currently paid.
Almost every answer to this question is a percentage — ask for 10%, ask for 15%, ask for 20% if you're changing jobs. The problem with all of them is the thing they multiply. If you're underpaid today, a percentage of your current salary carries that error forward and makes it permanent. You've anchored the conversation on the number you're trying to correct.
Price the work instead. Two comparable roles advertised at $95,000 is an argument. "I'd like 12%" is a preference, and your manager has no way to defend a preference to their own boss.
One number, not a range. A range is an invitation to the bottom of it.
Why percentages are the wrong anchor
Your salary is not a measure of your value. It's a measure of what your employer estimated your value to be, at some point in the past, under whatever conditions applied then — the budget that year, what you'd accepted before, how well you happened to negotiate on a single afternoon.
That figure then gets carried forward by percentage increases forever, which means an error made once compounds quietly for the rest of your career.
So when you take that number and multiply it by 1.1, you are asking to be slightly less wrong. The percentage framing guarantees that the gap between what you're paid and what the work is worth can only ever narrow slowly, and only if someone else decides to let it.
Your salary tracks your perceived value, not your actual value.
There is one honest exception. If you already know you're paid correctly for your role and you're asking for a cost-of-living or performance increase inside a fixed band, then a percentage is the right unit, because you're not disputing the underlying number. That's a real situation and this page isn't about it.
What to anchor on instead
Price the job, and price it externally. Three sources, in rough order of how much weight they carry:
What comparable roles are advertised at. The strongest evidence you can get, because it's public, current, and it's your own employer's competition setting it. Live postings beat salary surveys.
Published ranges. Many jurisdictions now require salary bands in job ads, and many employers publish their own. Where they exist, they're hard to argue with.
What your organization already pays for equivalent scope. The most persuasive of the three and the hardest to get. If someone in a neighboring team owns roughly what you're proposing to own, their band is the internal precedent, and internal equity is something most organizations genuinely care about.
Notice that all three price the work, not you. That's deliberate. It keeps the conversation off your performance — which is contestable, personal, and where these discussions go to die — and onto scope, which is a matter of record.
Where to find the evidence
You need three or four data points, not a spreadsheet.
- Live job postings for the role you're proposing, at organizations you'd plausibly lose you to. Screenshot them, because postings disappear.
- Any published band your own employer uses, including in its own job ads.
- Industry surveys, professional bodies, and leveling frameworks where your field has them.
- What you know about internal equivalents, handled carefully — you can reference the scope without naming a colleague's salary.
Then state the conclusion, not the research. "Roles owning this scope are advertised between $92,000 and $101,000; I'm proposing $96,000" is one sentence and it's enough. You are not obliged to show your working unless asked, and volunteering all of it reads as justification rather than confidence.
What if the gap is large?
If pricing the work honestly produces a number far above what you're paid, that's not a problem with the number. It's the finding.
But a 30% jump is hard for a manager to get approved against your current job title, because the system they're arguing inside is built around incremental moves within a band. So don't ask for it against your current job. Propose a different one.
That's the whole logic of the method: a large increase is indefensible as a raise and entirely defensible as a new role with a new title and genuinely different responsibilities. The title isn't vanity — it's the thing that makes the salary approvable by finance and portable to your next employer.
If the gap is large, you have your answer about which conversation to have.
Say one number
When you get there, say it plainly, once, and stop.
Not a range, because you will get the bottom of it. Not "somewhere around," because that's a range wearing a coat. Not a question. And don't immediately fill the silence — the pause after a number feels much longer to the person who said it than to the person who heard it.
Name it fifth, too, after the opportunity, the problem, the responsibility and the evidence. By then the number is the conclusion of an argument rather than an ask.
What the number is actually worth
Before you decide the number is too bold, it's worth knowing what you're negotiating over.
A raise is not a one-off payment. It lifts the base every future percentage increase is calculated from, it raises what goes into your retirement accounts, and it resets what your next employer offers. A modest increase, compounded across the years you have left to work, becomes a figure most people find genuinely startling — and the difference between asking for the right number and a slightly safer one is usually most of it.
Questions
Is 10% a reasonable amount to ask for?
It's a reasonable amount to receive and a poor way to decide what to request, because it takes your current salary as correct. Work out what the role you're proposing pays elsewhere and use that. If the answer happens to be around 10% more, you now have evidence for it rather than a convention.
Should I give a range to seem flexible?
No. You'll be offered the bottom of it, and you'll have signaled that the top was decorative. If you want to show flexibility, be flexible about timing, scope or the path to the number — not about the number.
What if they ask what I'm currently paid?
Answer if you're required to, and in many places they're not allowed to ask. Either way, redirect to the work: what matters is what the role you're proposing is worth, and your current salary is a fact about a different job. Some of the strongest cases exist precisely because those two figures have drifted apart.
Should I ask high so there's room to come down?
Not far. An unevidenced number damages the thing doing the work, which is that every figure you've said so far has been defensible. Ask for what the evidence supports, at the upper end of what it supports.
What if I'm told the number is above my band?
That's useful information, and it's an argument for a different band rather than a smaller number. Ask what scope the next band up is defined by, and what would have to be true for the role to sit in it. You've just been told exactly what to propose. ---
Ready to do it properly?
The free course helps you work out the number, the evidence behind it, and the role it belongs to — applied to your own job, one question at a time. About fifteen minutes, nothing to buy, and nothing you write ever leaves your browser.