The strongest evidence is external and verifiable: published wage data for your occupation and metropolitan area, a written competing offer, and a dated record of scope you have taken on. Personal need, length of service and effort are all true and none of them are arguments about the market. Your manager has to justify the case to someone who has never met you, so it has to survive being repeated by a third party.
Two kinds of evidence, and you need both
Your manager is going to repeat your case to somebody who has never met you, probably from memory, in a meeting where several people are competing for one budget. That single fact should decide everything about how you build it. The test is not whether the argument is true. The test is whether it survives being repeated by a third party.
Market evidence is the first half: the published median and spread for your occupation in your metro, plus advertised ranges locally including your own employer’s. That says what the role is worth to anybody, independent of your view of yourself. It is checkable by the person your manager is talking to.
Delivery evidence is the second half: what you actually did, with numbers wherever numbers exist. That says why you should sit above the middle of the range rather than at it. Market evidence alone invites the answer that the band is the band. Delivery evidence alone invites a conversation about how everybody works hard. Together they are much harder to decline.
Getting numbers onto your delivery
Most work has a measurable side even where the job itself is not formally measured. Volume handled, time saved, errors avoided, revenue influenced, people trained, processes you now own that somebody else used to run. None of that requires a dashboard, only a note kept at the time.
Where genuinely nothing is countable, scope still is. What you are responsible for now that you were not responsible for a year ago is a fact with a date attached. Growth in scope is the clearest justification for a pay change available to anybody, and it is checkable without a single metric. Write it as a before and after rather than as a description.
Keep the record as you go
A running note updated monthly takes a few minutes and produces something you cannot reconstruct at review time. People systematically forget their own first two quarters, and the things that felt significant in March are invisible by November. The note is not for reflection; it is raw material.
It also protects against manager turnover, which is one of the most common reasons a genuinely strong year goes unrewarded. The person who watched you do the work has moved to another team and the record is the only thing left. A new manager cannot advocate from memories they do not have. They can advocate from a page you hand them.
Match yourself to the occupation, not the title
Published wage data is organized by occupational code and those codes are defined by duties rather than by branding. Titles are invented by employers and follow no standard, so looking up your title averages across genuinely different levels. The figure you get back describes nobody in particular.
Describe your week instead and read the published occupational definitions until one of them matches. If your role genuinely spans two codes, look up both and say so explicitly. Framing it as roughly two thirds one occupation and one third another, with the median for each, is a stronger argument than picking whichever pays more. It also shows your reasoning, which is what makes it hard to dismiss.
The one page to hand over
Three or four delivery points with numbers, the published median and seventy-fifth percentile for your occupation and metro, and one specific request. That is the entire document and it should fit on a single side. Anything longer gets summarized by somebody else, and their summary is what actually reaches the decision.
Write it for your manager to use rather than for yourself to feel prepared. Those are different documents. The version that helps you feel ready is comprehensive; the version that wins is short enough to be repeated accurately in a meeting. Cut anything that would not survive being said aloud from memory.
What to leave out
Leave out colleagues’ salaries even when you know them, because it changes the subject to how you know and creates a problem for somebody else. Leave out your personal expenses, which are real and are not an argument about the market. Leave out how long it has been since your last increase, on its own, because time served is not evidence of value.
Leave out any suggestion of what you would do if the answer is no, unless you genuinely mean it and are prepared to follow through. Each of those items moves the conversation off ground where you are strong and onto ground where the employer has no obligation and no comparable. They feel like additional arguments. They function as replacements for the good ones.
Ask for a specific number
Ask for a figure rather than for more, and avoid offering a range whose bottom you would accept. Name the number and tie it directly to the evidence: the seventy-fifth percentile for the occupation in the metro, justified by the scope you have documented. That is a request somebody can carry into a room and defend.
A vague request gets converted into whatever the budget was going to produce anyway. It gives your manager nothing to argue for and nothing to lose by conceding less. Specificity is not aggression; it is the thing that makes advocacy possible. The people who ask for a number get a number.
A worked one-pager
Three lines of delivery: what you took on, what it produced, and what it replaced. Two lines of market: the published median and seventy-fifth percentile for your occupation in your metro, with the source and reference period named. One line of request. Six lines in total and every one of them checkable.
The request line does most of the work and it should be one sentence. The May 2025 wage survey puts the median for this occupation in this metro at one figure and the seventy-fifth at another, and given the scope above you are asking for the higher one. Named source, stated period, specific geography, justified placement. Nothing in that sentence requires anybody to agree with you about anything except arithmetic.
What to do with the answer
If the answer is yes, get it in writing with an effective date attached. If it is partial, ask what would close the remainder and by when, and ask for that in writing too. If it is no, ask whether the constraint is the budget, the band, or the case.
That last question is the most useful one in the entire conversation and almost nobody asks it. Those three constraints are completely different problems with completely different responses: a date, a level change, and better evidence respectively. Leaving the room without knowing which one you are facing wastes everything you prepared. Ask it before the meeting ends.
Why checkability is the whole point
Numbers survive being retold and adjectives do not. A case built on figures arrives in the decision meeting intact, in roughly the form you wrote it. A case built on how hard you have worked arrives as a general impression sitting alongside everybody else’s general impression.
That is the reason the market half matters even when your delivery is genuinely exceptional. Your manager needs something a stranger can verify, because a stranger is who they have to persuade. Give them the verifiable half and the persuasive half together. The combination is what actually moves a number.
Common questions
What is the strongest single piece of evidence?
A written competing offer, because it is checkable and immediate. It is also the one with a cost, since it raises the question of whether you are leaving regardless.
Should I use national or local wage figures?
Local. Published metropolitan figures for your occupation are the fairest comparison and the hardest to wave away.
Does length of service count?
Only through what it produced. Time alone is not a market argument; the scope you picked up along the way is.
Should I name a number?
Yes, with its basis. A direction invites a no; a figure with a source behind it invites a specific answer.
What about personal circumstances?
Leave them out. They are true and they are not evidence about what the role is worth, and they move the conversation onto sympathy.
What evidence do I need for a raise?
Market evidence — published medians and advertised ranges — and delivery evidence with numbers. Either alone is easy to decline.
What if my work is not measurable?
Use scope: what you are responsible for now that you were not a year ago. Growth in scope is checkable without a metric.
What should I ask if the answer is no?
Whether the constraint is the budget, the band, or the case. Those are three different problems with three different responses.