salary-negotiator

Salary Negotiator

The 10 minutes of a compensation conversation move more money per minute than anything else in a career — yet most people walk in unprepared, accept the first number, and leave 10-30% on the table. This skill does the preparation: market anchoring, walk-away math, total-comp evaluation (equity! bonuses! benefits!), and the actual scripts to say.

Overview

scripts/salary_negotiator.py runs the numbers that matter before any negotiation:

• Market anchor builder — turn salary-band data (levels.fyi / Glassdoor / recruiter info) into a defensible anchor: target, acceptable, and stretch numbers with rationale you can say out loud
• Walk-away floor calculator — personal financial floor (monthly costs × runway, debt, savings goals) → the number below which the offer is a no regardless of charm
• Total compensation evaluator — normalize offers with different shapes: base + bonus % + RSUs + options + 401k match + sign-on into an annualized expected value with risk weighting (startup options are lottery tickets; price them like one)
• Raise conversation planner — build the case from impact bullets, market deltas, and commitment signaling; includes the "no budget" counter-moves
• BATNA analyzer — best-alternative analysis: what you do if this fails, and how strong that makes your position
• Script generator — annotated negotiation scripts: acknowledging-then-countering, anchoring, silence technique, and the graceful multi-round dance

Everything runs offline with explicit assumptions; the user brings market data (the tool tells them exactly which sources to check).

When to Use

• A job offer arrived (or is coming) and the user wants to negotiate it
• Preparing an annual review / raise ask
• "Am I underpaid?" — evaluating current comp against market data
• Comparing two offers with different structures (big-base vs equity-heavy)
• Deciding whether to counter, and by how much

Don't use for: contract/freelance rate-setting (that's freelance-rate-calculator territory), executive-level comp with accelerators/cliffs beyond RSU basics, or legal review of equity agreements (recommend an employment attorney for non-competes and option terms).

How It Works

1. Inputs: current comp (or none), offer details, market data points (min/median/max for the role+level+location), personal floor inputs (monthly costs, runway needs).
2. Anchor math: target = market median × experience factor (if you have leverage evidence, above-median; if career-switcher, at-median). Stretch = market p75-p90. Floor = personal financial floor, never below.
3. Total comp EV: base + (bonus % × base × probability-of-payout) + RSU annualized value × vesting-risk factor + options × (strike-adjusted probability-weighted value) + benefits delta (401k match is cash; insurance differences priced roughly).
4. Scripts: pick the scenario (new offer, counter round 2, raise ask) and the tone (collaborative/firm); generate the script with fill-in numbers from the computed anchor.
5. Practice Q&A: likely pushbacks ("budget is fixed", "this is our standard band") and the counters.

Quick Start

# Build the negotiation strategy for an offer
python3 scripts/salary_negotiator.py offer \
  --role "Senior Backend Engineer" --location "Remote US" \
  --offer-base 145000 --offer-bonus 10 --offer-signon 15000 \
  --offer-rsu-annual 40000 --market-min 130000 --market-med 160000 --market-max 210000

# What's my walk-away number?
python3 scripts/salary_negotiator.py floor --monthly-costs 4200 --runway-months 6 \
  --other-income 0 --benefit-gap 500

# Compare two differently-shaped offers
python3 scripts/salary_negotiator.py compare \
  --a "BigCo: base 150k, bonus 15%, RSU 60k/yr, 401k match 4%" \
  --b "Startup: base 130k, options 0.5%, strike 1.50, valuation 60M"

# Script for the raise conversation
python3 scripts/salary_negotiator.py raise --current 120000 --market-med 145000 \
  --impact "led payments migration, -38% infra cost" --ask 138000

# Full walkthrough
python3 scripts/salary_negotiator.py demo

Steps (Agent Workflow)

1. Gather: role, level, location, offer components, market data (point the user to levels.fyi / Glassdoor / LinkedIn / recruiters — insist on ≥3 sources), and their financial floor.
2. Run floor first — knowing the walk-away changes everything downstream.
3. Run offer (or compare if multiple offers) to get anchor numbers and total-comp EV.
4. Generate the script (offer includes it; raise for internal conversations).
5. Rehearse the pushback Q&A with the user — have them say the counter number OUT LOUD.
6. Remind: get it in writing; never resign on a verbal; silence after their number is a tool, not rudeness.

Output Shape

NEGOTIATION STRATEGY — Senior Backend Engineer (Remote US)
═════════════════════════════════════════════════════════
MARKET ANCHOR
  Market: $130k — $160k — $210k (min/med/max)
  Your target: $172,000  (median +7%: strong skills evidence, competitive offer)
  Your stretch: $189,000  (p75; justifiable with the RSU haircut argument)
  Your floor:  $151,000  (personal financial floor — see floor calc)

OFFER EVALUATION
  Base $145,000 + bonus 10% ($14,500 × 85% payout = $12,325)
  + sign-on $15,000 /3yr = $5,000 + RSU $40,000 × 90% retention = $36,000
  TOTAL COMP EV: $198,325/yr   vs your target $172k → offer is ABOVE target
  Base alone: $145k is 9% below market median — negotiate BASE, not RSUs.

SCRIPT (collaborative-firm, round 1)
  "Thank you — I'm genuinely excited about this role and the team.
   [pause] Based on my research for this scope in [location], market
   medians run around $160k base, and given [specific qualification],
   I was targeting $172k. Is there flexibility to get closer to that?"
  ▸ If "band is fixed": pivot to sign-on/RSU/equity refresh/PTO — see Q&A sheet.

Common Pitfalls

1. Negotiating total comp when cash is the constraint. RSUs and bonuses are risk-weighted; base is (nearly) certain. If the base is 15% below market, a bigger option pool does not fix it — evaluate EV, then negotiate the component that's mispriced.
2. Anchoring with a range. "I'm looking for 150-170" → they hear 150. Always one number, always your target, always with rationale.
3. Not having a floor. Without a walk-away number, "no" sounds like an argument; with one, it's a fact. Compute it before anything else.
4. Accepting on the call. Enthusiasm is fine; commitment is premature. "I'd like 48 hours to review the full package" is standard and loses nothing.
5. Countering the day the offer arrives without reading the shape. A weak base with a giant sign-on may beat a strong base after year 1 — do the multi-year math (compare does 1/2/3-year views).
6. Negotiating against yourself. State the number. Then stop talking. The next person to speak loses money — make sure it's them.
7. Ignoring the "no budget" counters. There are always other levers: sign-on, equity refresh, start-date bonus, PTO, remote stipend, title, review-in-6-months-with-criteria. The Q&A sheet enumerates them.

Verification Checklist

• Market data from ≥3 sources, role+level+location matched
• Personal floor computed BEFORE strategy
• Total comp evaluated as EV, with risk weighting on variable components
• One target number chosen (never a range)
• Script rehearsed out loud, including the number
• Pushback Q&A reviewed
• All agreements to be confirmed in writing

One-Shot Recipes

"They offered 115k, recruiter says band tops at 125k, I wanted 130k"

python3 scripts/salary_negotiator.py offer --role "PM" --offer-base 115000 \
  --market-min 105000 --market-med 125000 --market-max 140000
# → anchor at 128-130 with rationale script; levers if base is truly capped

"Raise review in 2 weeks; I make 95k, market is 115k, I shipped X and Y"

python3 scripts/salary_negotiator.py raise --current 95000 --market-med 115000 \
  --impact "shipped X (+12% activation), led Y migration" --ask 108000
# → script + the 'budget lands next cycle' counters (criteria + date in writing)

References

• references/comp-structures.md — equity types, vesting, bonus shapes, benefit valuation
• references/scripts.md — full script library: offers, counters, raises, rescission-safe exits