Everything you lose in a salary negotiation, you lose in the first ninety seconds after the number is said. Most of it is lost by talking.
You want the job, they know you want the job, and the person making the offer is often the person you will report to on Monday. That makes the usual hard-bargaining advice actively harmful. The difficulty is not extracting the maximum — it is improving the offer without spending the goodwill you will need for the next two years.
Thank them, say you are pleased, and ask for the detail — base, bonus, equity, review cycle. This is not stalling; you cannot evaluate an offer you have only heard the first component of, and the pause costs nothing.
A counter without a rationale is a wish. A counter with one — market data, the scope discussed in interview, an existing offer — is a position. Say the number, say why, and then stop.
This is the single most valuable move and the hardest. Most candidates negotiate against themselves in the silence after their own counter, softening it before the other person has said anything at all.
If base is genuinely capped, the constraint is usually one budget rather than all of them. Signing bonus, an earlier review date, extra leave and title often come from somewhere else entirely, and are frequently easier to say yes to.
'What's the band for this level?' or 'What's making this difficult?' invites the other person to explain their limits, and people who explain their limits usually reveal where the room is.
The number you said in the first recruiter call becomes the ceiling. Deflect early salary questions to a range and only commit once you have seen theirs.
Saying 'I was hoping for 115, but I'm flexible' concedes before anyone has pushed. Say the number and let them respond to it.
It is checkable, it is remembered, and it occasionally gets accepted at face value in a way you cannot then walk back. Real leverage stated plainly beats invented leverage.
The hiring manager wants you to accept — they have spent weeks on this and reopening the search is expensive. That is shared interest, not opposition, and the tone should reflect it.
Almost always ask, and do it once, with a reason. Most offers have some room and the downside of a single well-reasoned counter is close to zero — offers are very rarely withdrawn for it. What does damage is repeated rounds of haggling after a final position has been stated.
Enough to be worth the conversation, close enough to be justifiable — commonly ten to twenty percent, with a reason. The reason matters more than the size: a number backed by market data or by scope discussed in interview is a position, and one without is a wish.
Ask what else is flexible. Base is often genuinely capped by a band while signing bonus, review date, leave or title come from different budgets and different approvals. 'Final' usually means final on that one component.
Out loud, against someone who uses silence. Almost everyone knows the theory and loses the money in the four seconds after their own counter, when they start talking again. That is a delivery problem and reading about it does not fix it.
Everything above is knowable by reading. The gap is between knowing the move and making it while someone is pushing back at you, and that only closes out loud. These are the negotiation scenarios — live voice, a counterpart who does not simply agree, and a scored debrief afterwards.
They've made an offer below your number and asked what you think. Everything you lose here, you lose in the next ninety seconds.
Procurement has your renewal and a competitor's quote. They opened with twenty-five percent off and went quiet.
No competing offer, no invitation, and a manager who will happily agree to revisit this later. Getting a yes is not the hard part — getting a date is.
Four objections cover almost every deal, and almost every rep answers all four by talking. The part that moves a deal is understanding what the objection is standing in front of.
The conversation managers avoid for months. Avoiding it is the expensive option, and the usual softening techniques are what make it land badly.
Solving the problem too early is what keeps people angry. The order matters more than the fix.
Salary, contract and scope. A real voice on the other end and a scored debrief against four published axes. The first call is free.