How to negotiate your salary: the one-page method

The most common mistake is to walk in with a number drawn from what you earn now plus a feeling. Your current salary is not the market; it is one data point with your own history baked in. Before anything else, find out what this role, at this level, in this city and industry, actually pays: published ranges, salary surveys, recruiters who place the role, and colleagues or peers who will tell you honestly. Aim for a range you could defend to a skeptic, with sources.

Then find out what you can about the other side's constraints: the band the role sits in, whether the budget is fixed or flexible, how recently they raised anyone, how hard the role has been to fill. An offer that took four months to make is a different negotiation from one that took two weeks. Everything below is stronger when the number and the constraints are known and weaker when they are guessed.

Write the sources on the page. If you are challenged in the room, "the two surveys I looked at put this role between X and Y" is a sentence a manager can take to their own boss; "I feel I deserve more" is not.

Do the same for the whole package, not only the base. Bonus structure, equity, pension contribution and leave vary enormously between employers, and a role that pays less in base can pay more in total.

Fisher and Ury's BATNA, the best alternative to a negotiated agreement, is the single most important line on the page. If this negotiation fails, what will you actually do? Take the other offer, stay where you are, keep looking, go freelance? The quality of that alternative is your real power, more than any argument you make, and knowing it is what lets you stay calm when the first answer is no.

Be honest about it. "I will walk" is only a BATNA if you would, and the other side can usually tell. A weak alternative is not a reason to avoid negotiating; it is a reason to negotiate on more than the base number, and to spend the weeks before the conversation improving the alternative, for example by running a second process.

Estimate theirs too. If they say no and you leave, what does it cost them: a restarted search, a team short-handed, a project slipping? Their alternative is often worse than they let on, and it is the reason a fixed budget moves.

Kahneman's anchoring effect is the reason preparation beats improvisation: the first number in a conversation pulls every later number toward it, even when everyone knows it is arbitrary. So decide three numbers in advance and write them down: the anchor you will open with, ambitious but defensible from your research; the target you would be pleased to sign; and the floor below which you will take your alternative instead. The floor comes from the BATNA, not from hope.

When you know the market, say your number first. The old advice to make them go first assumes you know less than they do; if your research is good, your anchor sets the range, and a well-sourced high anchor is rarely punished. When you genuinely do not know the range, ask for theirs, and treat the first number you hear as their anchor, not the answer.

Give a precise figure rather than a round one. A specific number signals that it came from somewhere, and it is harder to slice in half.

Rehearse saying the anchor aloud until it comes out flat, without a laugh or an apology attached. The number is only as strong as the way it is said.

A position is a number. An interest is the reason behind it, and Fisher and Ury's method is to negotiate the interests. Your rent going up is your problem; the value you bring to the team is their interest, and it is the only thing that justifies a number to the person who has to approve it. So prepare the case in their terms: the results you delivered, the ones you are about to, what it would cost to replace you, what the role is worth to the business.

Write three concrete items, with numbers where you have them. Then translate: not "I need more to cover my costs" but "the market for someone who has done X and Y is at this level, and here is what I plan to do next." A manager can carry that sentence into the room where the decision is actually made.

Chris Voss's accusation audit takes the sting out of the awkwardness. List what the other side may be thinking of you in this conversation, and name the strongest ones first: "You may feel this is coming at a bad time for the budget" or "I know you have already gone through a process to make this offer." Said aloud, the objection loses its charge, and the person hears that you understand their position before you have asked for anything.

Then say what you want plainly. The audit is not an apology; it is a way to clear the objections off the table so that the number can be discussed on its merits.

People say yes when they feel understood, and a hiring manager or a boss with a fixed band is a person under constraints. Before your number, describe their situation accurately: "It sounds like the band for this level is tight and you have already stretched it once." Then stop and listen. When they correct you, you have learned something about the constraint; when they say "that's right", they have stopped defending, and the conversation can move to how.

This costs a minute and changes the tone of everything after it. It is also the step that people skip because it feels like giving ground. It is not; you can understand a constraint completely and still ask for the number.

When the first answer is a number below your target, do not argue and do not accept. Ask a calibrated question that hands the problem back: "How can we get closer to X?" "What would need to happen for this to work?" "How is the band set, and what would move it?" These cannot be answered with a no, they keep you out of the position of the complainer, and the answers reveal the levers: a review date, a signing bonus, a title change, a different band.

Prepare three and write them down; you will not invent them under pressure. Avoid "why", which sounds like an accusation, and avoid questions that are really demands with a question mark.

Then be quiet. After a calibrated question, the silence belongs to the other side; filling it with a concession is the most common way to lose a raise you had already won.

A loss hurts about twice as much as an equal gain pleases, and a budget holder feels a raise as a loss. So describe the same money as the thing it protects: continuity on the project, the cost of a search, the knowledge that leaves with a person. "Bringing this to market rate keeps the team whole through the launch" is the same request as "I would like more money", heard by a different part of the brain.

Do this honestly. The framing works because it is true; the cost of losing you is real, and naming it in their terms is the fair version of the argument.

If the base cannot move, the negotiation is not over; it has become a trade. Fisher and Ury's method of inventing options for mutual gain applies here: things that cost the company little and are worth a lot to you, or the reverse. A signing bonus from a different budget, a review in six months with agreed criteria, an extra week of leave, remote days, a title, training, an earlier start on the higher band. Make a list before the meeting, ranked by what each is worth to you.

Every concession you make should be exchanged for something, and said as an exchange: "If the base has to stay at X, could we agree a review at six months against these targets?" A concession given for nothing teaches the other side that there is more to take.

Expect a counter, and plan how you will move: in small steps that get smaller, each one in exchange for something, never straight to the midpoint. Splitting the difference is the reflex that gives away the most, because the person who anchored higher wins it. If the counter is below your floor, say so calmly, thank them, and take your alternative; if it is between the floor and the target, trade for the rest.

When you agree, repeat the terms back and ask for them in writing before you resign from anything or stop the other process. A verbal yes from a manager is not an offer until it has survived their own approval chain. Then the page: the market range with sources, your BATNA and theirs, the three numbers, the case in their interests, the accusations to audit, the calibrated questions, the loss you are protecting them from, the trade list, and your floor. Filled in the night before, it takes twenty minutes.