How to ask for a raise (and what to do when the answer is no)
Asking for a raise is a different exercise from negotiating an offer, and treating them the same is why so many internal requests fail. In an offer negotiation you have leverage, a deadline and an alternative. Internally you usually have none of those, and the request runs through a budget process you cannot see.
That changes the strategy completely. The goal is not to win a negotiation in a single meeting; it is to give one person the material to argue your case to people you will never meet.
Your manager is not the decision maker
This is the thing most people get wrong. In most organisations of any size, your manager cannot approve a meaningful increase alone. They can advocate for it, in a cycle, against a fixed pool, competing with other managers doing the same for their own people.
So the meeting is not a negotiation. It is a briefing. You are equipping your manager with evidence, framing and numbers they can carry into a room you are not in. Everything below follows from that: make it easy to repeat, easy to justify, and hard to argue with.
Timing matters more than delivery
Compensation decisions are made on a calendar, and the calendar is usually invisible from below. By the time increases are announced, the pool was allocated weeks or months earlier. Asking after that is asking for an exception, which is a much higher bar than asking for a share.
Find out when your organisation sets budgets — your manager will usually tell you if you ask directly — and open the conversation one cycle ahead. Six to eight weeks before the decisions are made is roughly right. The second-best timing is immediately after a visible win, when the evidence is fresh and unarguable.
Build the case before you ask for the meeting
Write down what has changed since your salary was last set. Not what you have been busy with — what is different because of you. New scope you own, results with numbers attached, things that used to require someone more senior and now do not, people you support, problems that stopped happening.
The test for each line is whether your manager could repeat it to their own boss without you in the room. Delivered the migration is not repeatable. Led the ledger migration with zero downtime, which removed the reconciliation errors that were costing two days a month is repeatable, because it carries its own justification.
Two or three of those are enough. A long list reads as padding and dilutes the strong items.
Why market rate is weaker than people expect
Salary data feels like the strongest possible argument and usually lands as the weakest. Ranges are wide, sources disagree, titles are not comparable across companies, and every manager has heard a number pulled from a site that does not match their reality.
Market data works as a sense check, not as the argument. Use it to make sure your ask is plausible, then lead with what you have delivered. Where market evidence is genuinely strong — you have a written offer, or the company is visibly hiring for your role at a higher band — it becomes powerful, but that is a different conversation with different risks.
The number
Name one. A range invites the bottom of the range, and a vague request for more puts the work of deciding what you deserve onto someone with competing demands on a fixed pool.
Anchor it to something. A named percentage tied to a named change in scope is easier to defend upward than a number with no reasoning. And be realistic about what internal processes can absorb: large jumps usually require a title change, because bands are attached to levels. If what you actually want is a promotion, ask about the promotion, since asking for promotion-sized money without the title tends to fail on process grounds alone.
The meeting
Ask for a dedicated conversation rather than raising it at the end of a one-to-one about something else. Say what it is for when you book it — you would like to discuss your scope and compensation — so nobody is ambushed and your manager can prepare.
Open with the ask and the evidence, not with a preamble. Then stop and let them respond. The most common failure in the room is talking past the point: candidates deliver a strong case, sense silence, and start negotiating against themselves before anyone has said no.
Do not use need. Rent, a mortgage, a new child and the cost of living are real and they are not arguments, because they are not connected to the value of the work. They also cannot be repeated upward, which by now you know is the actual test.
Do not threaten to leave unless you mean it
An implied threat converts a discussion about your contribution into a discussion about your retention risk, and once that frame is set it is hard to remove. Some managers respond to it; many quietly start planning for your departure.
If you genuinely have another offer, that is a different situation and it is worth being straightforward about, while accepting the risk: telling an employer you are prepared to leave changes how they see you even if they match. Only do it if you would actually go.
When the answer is no
It often will be, at least initially, and this is where the outcome is actually decided. A refusal is not the end of the exercise unless you treat it as one.
Ask two questions. What would need to be true for this to be a yes, and when can we revisit it? Both are reasonable, both are hard to refuse, and both convert a closed door into a defined path. Then get the answer in writing — a short follow-up email restating what you agreed, which is not aggressive, it is how commitments survive a reorganisation or a change of manager.
Vague answers are themselves the answer. Keep doing what you are doing and we will see is not a plan, and if a second cycle passes with the same response, you have learned that the increase you want is not available here. That is useful information, and it is worth acting on rather than repeating the request a third time.
The follow-up that most people skip
Whatever the outcome, send a short note the same day summarising what was discussed and any agreed next step. It creates the record, it makes your manager's advocacy easier because the material is now written down, and it means the conversation survives the six weeks before decisions get made.
If a raise was agreed, get the amount and the effective date in writing before you consider it done. Verbal agreements about money have a way of becoming approximate.
The short version
Ask a cycle early. Bring two or three results your manager can repeat upward without you. Name one number and anchor it to changed scope. Leave need and market printouts out of it. If the answer is no, get the conditions and the date, in writing — and if the same no arrives twice with nothing specific attached, believe it.
Promotion and pay rise are different conversations
It is worth being clear with yourself about which one you are having, because the arguments are not interchangeable. A pay rise says you are being paid below the value of the work you already do. A promotion says you are already operating at the next level and the title should catch up. Organisations handle them through different processes, on different timelines, with different approvers.
The signal that you want a promotion rather than a raise is that your evidence is about scope rather than output: you are making decisions that used to be escalated, you are the person others consult, you are accountable for outcomes rather than tasks. If that is your case, ask for the level. Asking for promotion-sized money without the level often fails purely on process, because bands are attached to titles and nobody can approve an exception quietly.
What to do while you wait
There is usually a gap of weeks between the conversation and any decision, and it is worth using rather than enduring. Keep a running note of what you deliver in that window, because it becomes the evidence for the next conversation whether this one succeeds or not.
Most people cannot recall what they achieved six months ago, which is why these conversations so often rest on the last three weeks. A file you add to monthly — one line, what changed, with a number where there is one — removes that problem permanently and takes about two minutes a month. It also makes your next resume dramatically easier to write.
If the honest answer is that you are underpaid and it will not be fixed
Sometimes the case is good, the delivery is right, the timing is right, and the answer is still no with nothing specific attached. At that point the useful conclusion is not that you argued badly; it is that the ceiling here is lower than what you want, and no amount of further asking moves a budget that does not exist.
The market rate for your work is generally set when you change jobs, which is an uncomfortable fact but a well-observed one — internal increases tend to be incremental and external moves reprice you. That does not mean leave immediately. It means stop treating the next internal conversation as the thing that will resolve it, and start looking with the same seriousness you brought to the raise request.
You can do that quietly and without drama while continuing to do the job well. Knowing what you are worth elsewhere also changes how the next internal conversation feels, whether or not you ever mention it.
Related guides
- They gave you 48 hours to decide. Now what?
How to tell real deadlines from pressure, how to ask for more time, and what a rushed deadline says about the employer.
- How to negotiate a job offer (scripts, numbers and when to stop)
A practical salary negotiation guide: researching a range, answering the expectations question, and scripts for countering an offer.
- How to answer 'what are your salary expectations?'
Why you should rarely answer first, how to deflect politely, and when a range beats a single number.