The data behind a good ask
of employees who negotiate receive a counter (PayScale, 2025)
career-earnings impact from negotiating first salary (Linda Babcock)
typical lift on first counter in tech/professional roles
Salary negotiation is one of the few moments where 60 minutes of preparation can produce a six-figure career outcome. The 7 prompts below are the preparation: market data, the case for your number, the exact script, the counter playbook, the close. AI does the structuring; you supply the specifics and deliver the conversation.
Why do most negotiations end below where they should?
Walk into a salary conversation without market data and you anchor on the company’s opening number instead of an independent reference point. Walk in with market data but no rehearsal for the counter-to-your-counter, and you fold the first time the recruiter pushes back. Walk in negotiating only the base salary, and you leave five or six other live levers (sign-on bonus, equity, PTO, start date, title, early review timeline) on the table. The prompts below address all three failure modes. They are designed to be run the night before the call, not in the call itself; AI used live during a salary conversation is rarely a good idea.
One ground rule: every number in the output of these prompts is something you have to be able to defend if asked. Run the prompts, mark which claims are anchored in real data versus which are AI-suggested, and only walk in with the ones you can back up. Recruiters can spot a fabricated benchmark within two questions.
The candidate who lost the most money in my career as a recruiter was the one who said “I'm flexible, just make me a fair offer.” Flexible always reads as anchor low.
Senior tech recruiter, 2025, paraphrased
⚠️ Do not script yourself into a corner
A rehearsed AI script that you read verbatim will sound rehearsed. The output of these prompts is a structure to deliver in your own voice, not a script to recite. Recruiters with a decade of experience can tell when a candidate is reading prepared lines, and it shifts their stance. Our editing pass for any prepared text: How to Edit AI Out of Your Writing.
What are the seven negotiation prompts?
1. Build the market data set
Help me build a market data set for this role:
– 50th / 75th / 90th percentile base salary (cite sources: Levels.fyi, BLS, Glassdoor, etc)
– Typical sign-on bonus range
– Typical equity grant range (RSU / options / both)
– Median total comp for this band at this company tier
– Geographic adjustment if the role is location-tied
For each number, name the source and the year. Mark any number I should verify before citing.
If a number is your estimate rather than from a source, say so. Do NOT invent percentile data.
Why this works: The “do not invent percentile data” constraint is the single most important rule in any salary-negotiation prompt. AI will confidently quote “the 75th percentile for this role is $X” without a source if you let it. Asking it to cite or flag the source teaches you which numbers are real and which were extrapolated. The marked verify-list is your homework before the call.
2. Build the case for your number
Here is the role I am negotiating for: [paste job description].
Here is the market data set: [paste from prompt 1].
Here is what I want to ask for: [your target number, base + bonus + equity].
Help me build the case for that number. Output:
– The 3 strongest evidence points from my background for this target (cite resume lines)
– The 1 weak spot the recruiter will likely flag, and the candid response to it
– The “competing offer” framing IF I have one, the alternative framing if I don't
– The 2-sentence summary I should be able to deliver if asked “why are you worth this?”
Constraint: do not exaggerate my experience. The 2-sentence summary should be defensible in interview-style follow-up.
Why this works: The “weak spot the recruiter will flag” output is the most-skipped preparation step. Every candidate has one; pretending you do not is what makes the negotiation feel forced. The “candid response to it” output is what you practice; the recruiter who hears a real acknowledgment of the gap moves on faster than the one who senses you are hiding it. The 2-sentence summary is your anchor when the conversation drifts.
3. Draft the initial ask
My target: [paste].
Draft three versions of my counter:
– Version A: the polite anchor (asks for the top of my range, frames it as based on market data)
– Version B: the candid counter (names the specific elements I want adjusted, in priority order)
– Version C: the silence variant (a 2-sentence response that asks them to explain how they arrived at their number)
For each, tell me which kind of recruiter responds best to that version. Mark which one is the highest-expected-value play for the offer they made.
Hard constraint: each version is 4 sentences max. Real negotiators are concise.
Why this works: Version C (the silence variant) is the one most candidates never consider. Asking the recruiter to walk through how they arrived at the number is the highest-information move in the early conversation; their explanation often reveals where the flexibility lives. The 4-sentence ceiling is the discipline; long counters dilute the ask.
4. Prepare for the counter to your counter
List the 5 most likely responses I will get back, and for each:
– What the response usually means (yes-but / firm-no / partial / stall / probe)
– The next move that earns the most ground without burning the relationship
– The specific sentence I should say to deliver that move
– The trap I might walk into (over-explaining, accepting too quickly, going silent)
For each response, mark whether it should be answered in real time or whether I should say “let me think about it and come back tomorrow.”
Why this works: The “answer now or sleep on it” output is the most useful single move in any negotiation. Most candidates over-commit live because the recruiter is on the phone and the conversational pressure tilts toward immediate yes. The 24-hour pause is almost always available and almost always improves the final number. The mapped 5-response table is the muscle memory you need so the pause comes naturally.
5. Negotiate the non-base components
Help me build the multi-lever ask. For each of these, give me:
– A specific dollar value or quantity (calibrated to the role tier)
– The 1-sentence framing of why this matters to me
– The likelihood the company can move on this lever
The levers:
– Sign-on bonus
– Equity grant (RSU value, options strike, vesting cliff)
– Year-1 cash bonus target
– Annual PTO / WFH flexibility
– Start date
– Title / level
– Early review / promotion timeline
Then rank the 7 levers by my likely value-per-dollar return.
Why this works: “Base is firm” is rarely the end of the conversation. The 7-lever map is the play that turns one declined ask into seven smaller asks where 2-3 typically land. The “value-per-dollar return” ranking is the prioritization that most candidates skip; signing bonus often beats base for a year-one comp boost, equity often beats both for multi-year, and start-date flexibility costs the company nothing but is worth thousands to you.
6. Write the email follow-up
The call covered: [paste your bullet notes].
Draft an email that:
– Thanks the recruiter for the call, briefly
– Restates the elements we discussed AND the elements still open
– Names a specific next step with a date
– Does NOT re-negotiate (that happens on the next call, not by email)
– 5-7 sentences total
– Professional, warm, not chipper, no exclamation marks
End with the one question I should ask in the email to keep the conversation moving toward close.
Why this works: The “do not re-negotiate by email” constraint is the most-violated rule in salary negotiation. Email negotiations almost always result in worse outcomes than phone or video. The email is for confirming what was said and keeping the loop tight; the next round of asks happens on the next call. The single question at the end is the bridge that moves the timeline.
7. The close (or walk-away)
My minimum acceptable: [paste].
My target: [paste].
Help me think through the close. Output:
– The 3 conditions that would make this a clear YES
– The 3 conditions that would make this a clear NO
– The 5 conditions where I should NEGOTIATE further
– For the YES case, the exact acceptance sentence that closes graciously
– For the NO case, the exact decline sentence that keeps the door open for future roles
– For the NEGOTIATE case, the next single ask I should make
End with: which outcome is the current offer pointing toward, and what is my next move?
Why this works: The yes/no/negotiate threshold is the decision most candidates make under pressure rather than ahead of time. Writing the threshold out before the conversation removes the in-the-moment cognitive load. The “decline that keeps the door open for future roles” sentence is the part that matters most over a 10-year career; the company you turn down today is the company you may apply to in three years.
What is the worst thing you can do with AI in salary negotiation?
- Read AI-generated lines verbatim on the call. Recruiters can spot a script within two sentences. The output of these prompts is structure; the words you say should be yours.
- Cite fabricated percentile data. “The 75th percentile is $X” without a real source is the line that, when probed, exposes the whole prep as guesswork. Mark every AI-generated number and verify before citing.
- Use AI to draft an angry email. If you are upset enough to want AI to soften the message, sleep on it. Decisions made under negotiation stress, then sent via AI-polished email, are the regrets that show up later.
- Forget the recruiter is on your side. The recruiter wants you to accept the role. They are not the enemy. Prompts that frame negotiation as adversarial produce worse outcomes than prompts that treat it as a structured partnership conversation.
- Skip the multi-lever ask (prompt 5). Negotiating only base when 6 other levers are available is the single biggest underclaim in salary-negotiation data.
What if you negotiate two or three offers a year?
If you are in a stage of career where negotiations come up regularly (active job-hunters, contractors, executives), the seven prompts above are a candidate for the ladder. Save each as a Claude skill. Bundle them as a plugin called negotiation-prep that you invoke the night before any negotiation call. The prompt set produces a 1-page prep brief you read on the way to the call. The setup cost is 60 minutes; the financial return across 3 negotiations is comfortably five figures for most professionals.
📊 The Prompt-to-Workflow Ladder
Tier 1: the prompts (this post). Tier 2: the skill (one per prep step). Tier 3: the plugin (negotiation-prep bundle). Tier 4: the workflow (prep brief auto-generates when you receive an offer email). When to climb →
What are the common questions about AI in salary negotiation?
Should I tell the recruiter I used AI to prepare?
Unprompted, no. If directly asked, yes (framed as: AI helped me structure my thinking and gather market data; the conversation and the asks are mine). Most recruiters do not care; they care that you are prepared, and AI-assisted preparation is the same category as Glassdoor research, just better-organized.
How much should I ask for on a counter?
In tech and professional roles, 7-15% on base is typical and usually well-received. Above 15% needs a stronger case (competing offer, deep specialization, scope expansion at this level). Below 5% under-claims and may signal you are uncertain in the data. Prompt 2 is what tells you which range to anchor in.
What if I do not have a competing offer?
Use the market data as your anchor instead. “Based on Levels.fyi and BLS data for this role at this tier, the 75th-percentile total comp is $X, and my background supports the upper end of that band” is a structurally equivalent and more defensible frame than a competing offer. Never fabricate a competing offer; the recruiting community is small and reference checks compound.
Is it OK to negotiate when the role is junior or entry-level?
Yes. Junior negotiations have smaller dollar movement (3-8% is typical) but compound enormously over a career. Linda Babcock's research on first-salary anchoring is the standard reference. Negotiating $5K more on your first job, if it carries forward into raises and future-job offers, can compound to over $1M over a 30-year career.
Where do these prompts come from?
They are part of the career section in the larger AI Prompt Library. The Library has over 500 prompts across 33+ categories with three difficulty levels, including the full career stack: cover letters, resumes, interview prep, salary negotiation, onboarding the first 90 days, and the harder cases like raise asks inside your current company.
Sources to read next?
- Levels.fyi (compensation benchmarks)
- Bureau of Labor Statistics Occupational Employment Statistics
- Harvard Business Review on negotiation
- PayScale role research
- Salary negotiation (Grokipedia)
✏️ Before the call
Polished AI-generated lines sound scripted within two sentences. Read through your prep, then deliver in your own voice. Editing pass for any written negotiation messages: How to Edit AI Out of Your Writing → The full 29-pattern catalog from Wikipedia's “Signs of AI writing” guide is documented there for any written follow-up email or LinkedIn message tied to the negotiation.
The AI Prompt Library · $39
over 500 tested prompts including the full career stack
The seven salary-negotiation prompts above are a free preview. The full Library has the career stack: cover letters, resumes, interview prep, salary negotiation, the raise-ask conversation inside your current company, and the harder cases like declining offers, accepting partial counters, and renegotiating after the start date.
1-on-1 Deep Work Session with James · $175
Walk into your negotiation with a custom prep brief
A focused 2-hour session. We run all 7 prompts on your actual offer (with the actual numbers, the actual recruiter, the actual market data), pressure-test your answers to the likely counters, and walk you out with a 1-page printable brief you reference during the call.
Get Smarter About AI Every Morning
Free daily newsletter. Built for people who want to use AI well, not chase every model.
Free forever. Unsubscribe anytime.
You may also like
- Best AI Prompts for Cover Letters — the front of the job-search funnel
- Best AI Prompts for Resume Writing — what gets you in the door
- Best AI Prompts for Job Interviews — the room you walk into
- How to Edit AI Out of Your Writing — strip AI patterns before you publish
- Prompt to Workflow: The AI Ladder — when to climb from prompt to workflow
Two ways to go further
The AI Prompt Library
1,000+ ready-to-use prompts for Claude, ChatGPT, and Gemini. Stop staring at a blank box.
Get it for $39 →2-Hour Live AI Crash Course
A private, beginner-friendly session across Claude, ChatGPT, Gemini, and the wider landscape.
Book for $125 →