Know your number: researching what you're actually worth.

The most common negotiation failure isn't poor tactics. It's arriving without a number — or with a number built from vibes, fear, and what your last employer happened to pay you. If the figure in your head comes from anywhere other than evidence, you will negotiate defensively, and defensive negotiators leave money on the table every single time.

Step 1: Find the real range (not the internet average)

Generic salary sites give you a national average for a job title, which is nearly useless — it flattens seniority, industry, company size, and geography into one meaningless middle. Go to sources with resolution:

• Disclosed pay ranges in job postings. Pay transparency laws have made this a goldmine. Search current postings for your exact role, your region, your company size. These are real ranges real employers are currently paying — not averages.
• Levels-style data. For tech and increasingly beyond, crowdsourced comp data broken out by level and company. Granular, current, honest.
• Recruiters. Ask directly: "What's the realistic band for this role right now?" A good recruiter will tell you; it's their job to know, and it costs them nothing.
• Peers, asked well. Not "what do you make" — ask "what range would you expect for this role at this level?" People answer the second question freely and the first almost never.
• Industry salary surveys. Associations in your field often publish them. Almost nobody reads them.

You're looking for a band, not a point: the low, the mid, and the high for your role, level, industry, and geography.

Step 2: Price yourself inside the band

The band tells you what the role pays. Your evidence tells you where in the band you sit. Be honest and specific:

• Above mid if you have documented, quantified outcomes; scarce skills for that market; a track record of doing the exact thing they need; and visible third-party proof.
• At mid if you're solidly qualified but your evidence is thin or generic.
• Below mid only if you're genuinely stretching into a level you haven't operated at.

Notice what determines your position: not how good you are — how demonstrable your goodness is. Two equally capable people, one documented and one not, will land in different parts of the same band. That gap is not fairness. It's evidence.

Step 3: Set three numbers

1. Your ask. Top of the defensible range, based on your evidence. Slightly ambitious, fully justifiable. This is what you say out loud.
2. Your target. What you'd genuinely be happy with. This is where you expect to land after negotiation — and you'll only land here if you asked higher.
3. Your walk-away. The number below which you decline. Decide it before the conversation, in a calm room. A walk-away invented under pressure is not a walk-away; it's a wish.

Step 4: Build the justification

Your number needs a sentence behind it that isn't about your needs. "I need X because of my mortgage" is not an argument; it's a disclosure of weakness. "Based on the market range of A–B for this role, and the documented results I bring — [two receipts] — I'm targeting X" is an argument. One is a plea, one is a case. They get answered very differently.

The uncomfortable part

Say your number out loud, alone, right now. If your voice wavers, you don't believe it — and they will hear that in the room. The cure is not confidence practice. It's evidence: when you can name three documented results that justify the figure, the number stops feeling like a request and starts feeling like a fact. That's the whole difference between negotiating with adjectives and negotiating with receipts.