Most advice about asking for a raise is advice about the meeting: what to say, where to put your hands, how to survive the silence after you name a number. That matters, but it is the last five per cent of the work. By the time you are sitting in the room, the answer has largely been shaped by things you did, or failed to do, over the previous six months.
This guide covers that earlier part: how to tell which kind of pay gap you actually have, how to build evidence your manager can use, who really signs off on pay, and how to choose a number you can defend. When you are ready to plan the conversation itself, the salary negotiation script generator will draft the wording for you.
First, Work Out Which Pay Gap You Have
"Underpaid" describes three different problems. They have different fixes, different odds of success, and they are argued in different ways. Naming yours before you ask is the single highest-value thing you can do.
The market gap
You earn less than someone with your title, experience and location would earn at another employer. This is the most familiar case and the easiest to research, though the published figures are noisier than they look — salary sites mix job levels and regions freely, so treat any single number with suspicion and look for a range that several sources agree on. Recruiters who place people in your field will usually tell you honestly what they are seeing, because it costs them nothing.
It is a strong case, because the employer's alternative is to hire your replacement at the market rate anyway, then spend months getting them up to speed.
The band gap
Your employer pays the market rate, but you sit near the bottom of your own grade while colleagues at the same level sit near the top. Most medium and large organisations run formal salary bands, and many will tell you where you fall within yours if you ask plainly. This is usually the easiest raise to get approved, because it needs no new headcount approval and no change of title — only a movement within a range that already exists on paper.
The scope gap
Your salary matches your job title, but the job has quietly outgrown the title. You absorbed a departing colleague's responsibilities, you now supervise people, or you own a process nobody else understands. Strictly this is a regrading request rather than a raise request, and it is the most winnable of the three, because the work is already being done and the organisation is currently getting it at a discount.
Pay compression is a common version of this: new hires arrive on more than people who have been there for years, because the external market moved and internal pay did not. Managers know it happens and are usually uncomfortable about it. Use it, but frame it as your own market value rather than as a complaint about what a named colleague earns — the second version makes the conversation about someone else's business.
Build the Evidence Before You Build the Ask
Managers rarely refuse a raise because they think you are bad at your job. They refuse because they cannot construct a case for the person above them, who has a budget and several other people asking for the same thing. Your real task is to hand your manager the argument, already assembled.
Keep a running record. One document, updated for ten minutes at the end of each month, is enough — and it is far more accurate than trying to reconstruct a year from memory the week before you ask. For each entry, capture four things.
- What changed, in numbers where you have them. Time saved, errors prevented, revenue brought in, a backlog cleared. Where there is no number, describe the before and after concretely.
- What it would have cost to not do it. A problem you caught early is invisible unless you say what it would have become.
- Who noticed. Save the forwarded thank-you, the client compliment, the note from another department. Third-party praise carries weight that self-assessment never will.
- What is new since your pay was last set. This is the one that decides the outcome.
That last point deserves emphasis. The question your manager must answer upward is not "is this person good?" — they already agreed you were good when they set your current salary. It is "what has changed since we agreed that number?" Everything in your file should answer that question. Work you were already doing two years ago is not evidence; it is the reason you are paid what you are paid now.
Three or four specific, substantial items beat a list of twenty routine ones. If you find the page hard to fill, that is genuinely useful information — it usually means you should spend two more quarters building the case rather than asking now. A structured career development plan is a reasonable way to do that deliberately, and the SMART goal builder will turn a vague intention into something with a date attached.
Find Out Who Actually Decides
Your manager is usually your advocate rather than the decision maker. Knowing what constraints they are working under tells you when to ask and what to ask for, and it stops you reading a budget problem as a judgement on your work.
Three things are worth establishing quietly, well ahead of any conversation. When is the salary budget actually set — because once it is locked, even an enthusiastic manager can do very little until the next cycle. Whether raises come from a fixed pool shared across the team, which makes your request implicitly competitive with your colleagues'. And whether there is a formal review process that all increases must pass through, or whether managers have discretion within limits.
None of this requires anything underhand. Most of it comes up naturally in conversation with longer-serving colleagues or with HR. It is simply part of understanding how decisions really get made where you work.
Choosing a Number You Can Defend
Go in with a specific figure derived from something outside your own feelings about it. "A bit more" invites the smallest increase the company can get away with, and an unexplained number invites a negotiation about the number rather than about your work.
As rough orientation: movement within an existing band tends to land somewhere around five to ten per cent, while a regrading that reflects genuinely expanded scope can justify considerably more. A well-evidenced market gap sits between the two. Whatever you land on, be ready to say where it came from in one sentence — the research, the band position, or the new responsibilities.
Decide your own floor before you walk in, too: the number below which you would start looking elsewhere. You never say it out loud, but knowing it stops you accepting something in the moment that you resent three weeks later.
Open the Conversation Early
The formal meeting should not be the first your manager hears of this. Springing it forces an immediate reaction, and an immediate reaction from someone who has not checked the budget is almost always a defensive one.
Flag it a few weeks ahead, briefly and without drama: that you would like to talk about your compensation, and could you find some time. That short warning gives your manager space to look up your position in the band, sound out their own boss, and arrive having already done some of the work on your behalf. It also signals that you are treating this as a professional process rather than a grievance.
For the meeting itself — how to open, how to state the number, and how to respond if you are turned down — the negotiation script generator builds a script around your own figures and achievements.
If the Answer Is No, Make the No Specific
An unqualified "not right now" is the least useful outcome, because it tells you nothing about whether to wait or to leave. Before the meeting ends, convert it into something dated and measurable: what specifically would need to be true for this to be approved, and when can we look at it again?
Then put it in writing. A short, friendly email summarising what was agreed — the criteria, the review date — is not an act of aggression; it is the thing that stops the commitment evaporating when the quarter gets busy or your manager changes.
How the conversation ends is itself information. A manager who gives you clear, achievable criteria and a date is worth working with. A manager who will not commit to anything, or who sets conditions that keep moving, has told you where you stand without having to say it. For most people, changing employer remains the fastest route to a significant pay increase, and it is worth being honest with yourself about that rather than waiting through a third review cycle. If the answer was no because your last rating was poor, deal with that first — our guide to handling a bad performance review covers how to recover before you reopen the pay conversation.
The Mistakes That Sink an Otherwise Good Case
- Arguing from your own costs. Your rent, your loans and your plans are real, but they are not a reason for the business to pay more. The case is the value you deliver.
- Naming a colleague's salary. It shifts the discussion to how you know, and it makes you a problem to be managed rather than an asset to be retained.
- Apologising your way in. Opening with "I know this is a difficult time" hands over the reason to say no before you have made the request.
- Bluffing about another offer. If you are not genuinely prepared to take it, do not mention it. This one gets called more often than people expect, and there is no recovery.
- Asking once and never again. One refusal in one budget cycle is not a verdict on your career. The people who end up well paid are usually the ones who ask, act on the answer, and ask again.
None of this makes the conversation comfortable. It does make it winnable — because you arrive having already done the part of the work that actually decides the outcome, and the meeting becomes a formality rather than a performance.