Overview

A company wants to set salaries for a new role. They have no idea what's market competitive. They ask AI: "What should we pay a Senior Product Manager?"

AI produces: "Market rate is $150K-200K depending on location, experience, and company size."

That sounds reasonable. They set their range at $160K-190K. Then they can't hire anyone because they're 20% below market for their location and company size. Or they overpay a junior person with minimal experience.

The problem: AI's benchmark is generic. Market data is contextual. You need to know your specific market: location (SF is different from Austin), company size (50 person startup is different from public company), stage (Series A vs Series C), industry. AI can help you research and synthesize data. But it can't replace real market data and local intelligence.

This lesson teaches you how to use AI to help with compensation analysis. You'll learn how to benchmark against actual market data (not AI guesses). You'll learn how to build salary ranges that are competitive but sustainable. You'll learn how to present compensation data to leadership. You'll learn how to communicate comp to candidates so they understand positioning.

Why This Matters for HR Professionals

Compensation is one of your biggest expenses. It's typically 50-70% of company expenses depending on industry. And it's where you most need to be strategic.

Pay too little and you can't hire or retain people. You get a reputation as low-pay and can't attract talent. Pay too much and you run out of budget. Pay inconsistently and you have equity issues (someone discovers they make 20% less than a peer) and morale problems.

Good compensation strategy:
- Attracts the talent you need
- Retains good people
- Doesn't blow your budget
- Feels fair internally (consistent within your company) and externally (competitive with market)

AI helps you analyze, synthesize data, and think through recommendations. But you need good data sources, not just AI generation.

The Data You Need for Benchmarking

Before you ask AI anything, you need real market data. This is non-negotiable.

Sources of salary data:
- Salary surveys: Levels.fyi (tech-focused), Blind, PayScale, Comparably, Mercer, Radford (enterprise)
- Job postings: What are competitors posting for the same role? LinkedIn, job boards, your own network
- Recruiter insights: What are they telling you companies are paying? What are they pitching candidates?
- Your own historical data: What have you paid for this role before? Did you retain people or lose them?
- Industry reports: Some industry associations publish comp data
- Public company disclosures: If you're competing with public companies, check their SEC filings for executive comp

Collect this data by: role, location, company size, company stage (early, growth, mature, public).

Why location matters:
A Senior PM in San Francisco might be $180K-220K. Same role in Austin: $140K-170K. Same role in Midwest: $120K-150K. Pay Austin rates in SF and you can't hire. Pay SF rates in Austin and you're massively overpaying.

Why company size matters:
A startup Senior PM (50 people): $130K-160K + equity that might be worth something. Established company (1000+ people) Senior PM: $170K-210K. They're the same role, but market compensation is very different.

How to gather this data:

Spend 2-3 hours per role and location:
- Check 2-3 salary survey sites (average the data, don't cherry-pick)
- Look at 10-15 job postings from competitors (note which companies, what they're offering)
- Talk to your recruiter: "What are companies actually paying for this role right now?" (reality often differs from postings)
- Check your internal data: "We paid $X for this role 2 years ago. What happened to those people? Are they still here?" (if they left, you underpaid)

Compile into a simple spreadsheet: role, location, company size, salary range, source. This is your benchmark data.

How to use AI (correctly):

Don't ask AI "what should we pay a Senior PM?"

Instead: "I'm benchmarking Senior Product Manager in San Francisco at a Series B (40 person) company. Here's the salary data I collected from surveys and job postings: [paste 5-10 data points]. What's the market range based on this data? What patterns do you see?"

AI synthesizes: "Based on the data you provided, the market range is $145K-185K, with most offers around $165K. Your data shows slightly higher on the lower end and slightly lower on the higher end, which suggests a range of $150K-175K might be right for your context."

This is useful. AI isn't generating the benchmark; it's synthesizing real data you provided.

Building Your Salary Range

Once you have market data, you build your own salary range based on:
- Your location and specific market
- Your company stage and size
- Your budget constraints
- Internal equity (what do you pay similar roles?)
- Your comp philosophy (are you market leader? At market? Conservative?)

How to structure a range:

A typical salary range has:
- Minimum (floor): Below market or at lower end. You won't hire many people at this price, but it's your absolute floor.
- Midpoint (target): At market rate or slightly above. Where most people land.
- Maximum (ceiling): Above market. What you pay strong performers or experienced people.

Typical spread: 20-30% from minimum to maximum. (Some companies go 50%; depends on range design.)

Example:
- Market data: $145K-185K (median $165K)
- Your company budget: $145K per role max
- Your range: $130K (floor) - $155K (ceiling)
- Midpoint: $142K (where you'd typically start)

This is at the lower end of market. If you want to hire strong candidates, you might need to increase your budget.

How to use AI to build a range:

"Market data shows $145K-185K for this role in our market. Our budget is $150K per hire. Build a salary range that's competitive but fits our budget. Also show what we'd need to spend to be competitive at market median."

AI might suggest:
- Conservative range: $130K-155K (under market, harder to hire)
- Market range: $145K-180K (at market, requires $155K average)
- Competitive range: $160K-190K (above market, requires $175K average)

You decide which fits your strategy and budget.

Benchmarking Against Your Own Historical Data

You've probably hired this role before. What did you pay? Did you retain them?

If you paid a Senior PM $120K three years ago and they're still here, you probably underpaid (market has gone up). If you paid $180K for someone who left in a year, maybe you overpaid, or they had other reasons for leaving, or you hired someone overqualified.

Use your own data to understand: Am I being consistent? Am I competitive? Do I have equity issues?

Example analysis:

You've hired 5 Product Managers over the past 3 years:
- Hire 1 (3 years ago): $110K, still here, promoted
- Hire 2 (2 years ago): $120K, left after 1 year
- Hire 3 (18 months ago): $130K, still here, solid performer
- Hire 4 (1 year ago): $125K, still here, struggling (performance)
- Hire 5 (6 months ago): $140K, still here, strong start

Pattern: People who got $120K+ stayed (or in Hire 2's case, left for other reasons). People at $110K+ from 3 years ago would be way underpaid by today's market. This suggests your historical pricing was low.

How to use AI:

"Here's our internal comp history for [role]: [list with dates, salary, tenure, performance]. Market data is currently [data]. What does this suggest? Are we consistent? Do we have equity issues? What should we pay going forward?"

AI helps you see patterns you might miss.

Building the Internal Equity Check

Beyond market data, you need internal equity. Roles that are similar should be paid similarly. Roles that are different (more senior, different skills) should be paid more consistently.

Simple internal equity check:

Make a list of all roles and their pay bands:
- Junior PM: $80K-110K
- Senior PM: $130K-160K
- Director of Product: $160K-200K

Are the progressions consistent? Does each level have a clear relationship to the others?

If not, you have equity issues. Someone finds out a peer is paid 15% more for similar work and morale tanks.

How to use AI:

"Here's our internal comp structure: [list roles and current ranges]. Does this feel internally equitable? Are the level progressions clear? Are there obvious gaps or inconsistencies?"

AI can spot inconsistencies you might miss.

The Comp Committee Case

When you're presenting comp recommendations to leadership, you need a strong case. They'll ask: "Why this range? Why this budget?"

The case should include:
- Market benchmark (here's what others pay, with sources)
- Internal equity check (here's what we pay similar roles)
- Your recommendation (here's what we should pay this role)
- Justification (here's why this makes sense: competitive + sustainable + equitable)
- Budget impact (here's the cost if we hire [X people] at this range)

Structure:
Market: Senior PM in our market (SF, Series B): $145K-185K median $165K
Internal: Our current PMs range $130K-160K, avg $145K
Recommendation: $150K-175K with $160K starting salary
Justification: Competitive with market (just above median), aligned to internal structure (PMs average $145K, we're starting new hire at $160K recognizing they're external hire), sustainable at current burn rate
Budget: 2 Senior PMs at $160K = $320K annual cost

How to use AI:

"Prepare a comp recommendation memo for [role]. Include: market benchmark (sources: [list]), internal equity check (current ranges: [list]), recommended range (reason: [why]), and budget impact. Keep it to 1 page, professional tone."

AI produces something structured and persuasive. You provide the substance; AI organizes it.

The Offer Letter: Communicating Compensation

When you make an offer, how you present compensation matters. Good communication helps the candidate understand the offer and feel it's competitive.

Bad offer letter comp section:
"We're offering $150,000 salary."

Good offer letter comp section:
"We're pleased to offer a compensation package of $150,000 annual salary. This salary is positioned at the market median for a Senior Product Manager in the San Francisco Bay Area at a growth-stage company ($145K-185K per market research). Your total compensation package includes:
- Salary: $150,000
- Health insurance: Medical, dental, vision (company covers 85%)
- 401(k) matching: 4% (up to company contribution maximum)
- Equity: [shares/options] representing [x]% of the company
- Professional development: $2,500 annually
- PTO: 20 days paid time off + holidays
- Other benefits: [list]

Your estimated total compensation value is approximately $190,000 annually, including salary, benefits, and equity value at [valuation]."

The good version shows you know the market, positions the offer competitively, and shows total comp (which makes the offer look better).

How to use AI:

"Draft the compensation section of an offer letter for a [role, level]. Salary: $X. Include: salary statement with market positioning, benefits summary, total comp value. Tone: professional, warm, compelling. Make the offer look good without being dishonest."

AI drafts comp section. You customize with actual details (company size, benefits, equity, etc.).

Adjusting Comp Over Time

As markets change, you need to adjust. If you don't, you'll gradually become uncompetitive.

When to adjust:
- Annually: Update benchmarks, review if ranges are still competitive
- Market shift: If your market heats up (everyone's hiring), salaries go up
- Retention issue: If you're losing people to competitors, you might be underpaying
- Fairness issue: If you hire someone at $160K and an existing person doing the same job is at $140K, you have a problem

Adjustments can be:
- Range adjustment: Increase the range top and bottom
- Individual adjustment: Bring someone who's below band up to band
- New hire increase: New hires at market, existing people gradually adjusted

How to use AI:

"We set salary ranges 12 months ago: [old ranges]. Current market data: [new data]. Should we adjust our ranges? By how much? How do we handle people who are now below band?"

AI helps you think through the adjustment.

Try This Now: Three Exercises

Exercise 1: Gather Market Data

Pick a role you hire for frequently. Spend 20-30 minutes researching:
- Check 2-3 salary survey sites (Levels.fyi, Blind, PayScale)
- Look at 10-15 job postings from competitors (note salary ranges)
- Talk to your recruiter: "What are companies actually paying for this role?"

Compile into a spreadsheet: source, range, company size, location. What range emerges?

Exercise 2: Internal Equity Review

List all roles in your team/company. What are the current salary ranges?

Ask: Are the progressions logical? Does each level pay more than the level below? Are the gaps consistent (20% between levels, for example)?

Ask AI: "Review our internal comp structure: [list]. Do the level progressions make sense? Are there obvious equity issues?"

Exercise 3: Comp Recommendation

Pick a role you're hiring for. Gather market data (Exercise 1). Build a salary range. Draft a comp committee memo including: market, internal equity check, recommendation, justification, budget impact.

Ask AI: "Draft a comp recommendation for [role] in [market]. Market data: [paste]. Current internal structure: [paste]. Budget: [X]. Recommended range: [your recommendation]. Please structure as: market, internal equity, recommendation, justification, budget impact. One page, professional tone."

Practical Application - "What to Do Monday Morning"

• 
  Establish your comp philosophy: Are you market-rate leader? At market? Conservative? Get leadership alignment.
• 
  Benchmark key roles annually: Update your market data each year. Markets move fast.
• 
  Build ranges by role, location, level: Not one-off offers. Structure reduces inequity and bias.
• 
  Check internal equity quarterly: Make sure similar roles are paid similarly. Flag inequities.
• 
  Build a comp committee case: Use AI to organize. Present to leadership with confidence.
• 
  Communicate clearly to candidates: Show the market positioning. Make the total comp case.
• 
  Adjust as markets change: Don't let comp drift. Review annually, adjust as needed.
• 
  Document your methodology: Why you're paying what. This is important for equity and legal defense.

Key Takeaways

• AI synthesizes data, doesn't create benchmarks: You need real market data from surveys, postings, recruiters.
  - Build ranges, not one-off offers: Structure reduces inequity and bias.
  - Internal equity matters: Similar roles should pay similarly. Inequity destroys morale.
  - Document your methodology: Why this range? Why this offer? If someone challenges it, you have a defensible answer.
  - Communicate the market: Help candidates understand positioning. Good communication makes offers more compelling.
  - Adjust for market shifts: Don't let comp drift. Update annually.

FAQ

Q: How often should I update comp benchmarks?
A: Annually minimum. Markets move quickly, especially for in-demand roles. If you have high turnover or are losing people to competitors, benchmark more frequently.

Q: Should salary bands be public within my company?
A: Some companies do this for transparency (helps employees understand where they are). Others keep it confidential. Either way, be consistent in how you communicate and don't pick winners/losers based on who asked.

Q: What if someone is above market (paid more than the range top)?
A: You don't cut their salary (that destroys morale and trust). But you don't increase it as much as you would for someone at-market. You keep them stable while the range grows, gradually bringing them back into range.

Q: How do I handle internal inequities (two people in same role paid differently)?
A: First, understand why (different start dates, different levels of experience, different negotiation). If it's unfair, adjust the lower-paid person's salary (don't cut the other one). Do this quietly; don't broadcast.

Q: Should I tell people the market range?
A: When recruiting and making offers, yes. It helps them understand your offer and feel it's competitive. Internally, it's your choice, but consistent transparency is better than mystery.

What's Next

Lesson 6.2 is about benefits communication: how to help employees understand complex benefits, which saves you support tickets and helps them make good choices about healthcare, retirement, and other benefits.