I see the same mistake in every job search cycle. Someone lands an interview, preps for behavioral questions, polishes their resume. Then the recruiter asks about salary expectations - and they freeze. They never looked up what the role pays. They throw out a number based on what they used to make, what a friend told them, or worse, what they think the company can "afford."
Your starting salary at a new company is the single largest lever over your lifetime earnings. Every raise, every bonus percentage, every future offer benchmark traces back to that first number. A $10,000 gap in year one compounds into a six-figure difference over a decade. Internal adjustments rarely correct a low anchor. You do not get a do-over.
The good news: in 2026, salary data is not guesswork. Public datasets, self-reported platforms, and equity-tracking tools give you a granular picture of what your role pays - by city, by company size, by experience level. You do not need inside access. You need method.
Why you need to know your salary range before the interview
Without market data, you are negotiating blind. The recruiter knows the budget, the midpoint, the ceiling. You are guessing. That is not an interview - it is information asymmetry.
When you walk in with data, the conversation changes:
- You answer the comp question with a band, not a single number. No anchoring low.
- You can tell within minutes if the role is under-market or over-budget.
- You negotiate from fact, not feeling. "The median for this role in Denver with 5 years of experience is $115k" lands differently than "I think I'm worth more."
- You decide fast whether to invest hours tailoring an application or move on.
LinkedIn data from late 2025 shows job postings that include a salary range get 42% more applicants. The flip side: candidates who know their market value filter smarter. They apply to fewer roles and convert at higher rates.
Where to find real (not guessed) salary data
Forget the group chat. Forget the anonymous Reddit thread that says "I make $300k as a junior." Here are the sources that hold up:
BLS Occupational Employment and Wage Statistics (OEWS). The Bureau of Labor Statistics runs the most comprehensive salary survey in the US. Updated annually, the OEWS gives you employment numbers and wage percentiles (P10, P25, median, P75, P90) by occupation and metro area. Filter by your SOC code and your city. This is the floor - government data lags by roughly a year, and it captures base wages only. No bonus. No equity. Use it as your floor.
Glassdoor. Self-reported salaries by company, title, and city. The sample size is large for corporate roles in major metros - San Francisco, New York, Austin, Chicago, Seattle, Denver. The bias: people who report tend to be at the extremes (very happy or very unhappy). Cross-reference with other sources. Glassdoor also surfaces total pay estimates that include bonus and additional cash compensation, which gets you closer to the real number.
Salary.com. Aggregates data from employer surveys and HR-reported comp data. Strong for mid-market roles in finance, operations, marketing, and HR. The percentile breakdowns are useful - the 25th, 50th, and 75th give you your band. Their free tools are limited, but the ballpark numbers are directionally correct for non-tech roles.
Payscale. Comp data tied to skills, certifications, and location. Payscale's advantage is the skill-premium layer: you can see how much a specific certification (PMP, CPA, AWS Solutions Architect) adds to the median for your role. If you have a credential that the market values, Payscale helps you price it.
Levels.fyi. The gold standard for the tech industry. Levels.fyi maps actual offer data by company, level, and location. Software engineer, product manager, product designer, data scientist - it shows base, equity, and bonus broken down by level (L3 through L8+). The data is self-reported but verified via offer letter uploads. For FAANG and adjacent companies, Levels.fyi is more accurate than Glassdoor. For startups and mid-size companies, the sample is smaller. Still the first stop for anyone in tech.
Levels TotalComp. A newer comp tracker focused on verified total compensation data with equity detail. Useful for understanding RSU grant sizes, vesting schedules, and refresher policies at specific companies. If you are negotiating a tech offer, cross-reference Levels.fyi and TotalComp to validate the equity component.
LinkedIn Salary. LinkedIn's comp tool is underused. It surfaces salary ranges by title, location, and industry using data from its 900-million-plus member base. Because the data is tied to real profiles and job titles, the sample is enormous. The downside: LinkedIn Salary shows ranges at the title level, not by company. Use it for broad market context, not company-specific number.
Your network. Obvious but underused. Reach out to three or four people in your function, at your level, in comparable companies. Do not ask "how much do you make?" Ask "what range would you consider fair for someone with my background applying in Chicago right now?" Less invasive, more useful. People in your network who recently changed jobs are the best source - their data is fresh.
How to filter the data (region, seniority, remote/hybrid, company size)
National averages are noise. A "senior accountant" median of $92,000 does not tell you anything if you are applying in Tulsa versus Manhattan. Apply layers:
Region and cost of living. The BLS OEWS lets you filter by metropolitan statistical area. Use it. The same job title in San Francisco pays 30-50% more than in Cleveland. That is not because the San Francisco company is generous - it is because a one-bedroom apartment costs $3,200 there and $1,100 in Ohio. Know your local band. If you are interviewing for a remote role at a company headquartered in a high-cost city, check whether the company adjusts comp by location. Many do. Some do not. Ask.
Company size matters. A Series A startup with 30 people pays differently from a mid-size company with 800 employees, and both pay differently from a Fortune 500 with 50,000. Glassdoor and Levels.fyi work best for larger companies. For startups, combine Levels.fyi data with conversations - founders and early hires at VC-backed companies in your space can give you a range. Seed-stage startups often pay below market on base but compensate with equity. Series C and later typically pay at or above market on base to reduce equity dilution.
Seniority is about scope, not title. "Senior Manager" at a Fortune 500 might manage two people and a budget line. "Senior Manager" at a 100-person company might run a department with P&L ownership. Filter by responsibility, not the words on the business card. Are you managing a team? Do you own a budget? Are you setting strategy or executing tasks? These questions determine your band more than the title does.
Remote, hybrid, or on-site. Fully remote roles at distributed-first companies (GitLab, Zapier, Automattic) typically pay a location-adjusted salary. Remote roles at companies with HQ-centric cultures sometimes pay HQ rates regardless of your location. For hybrid roles, the comp is usually tied to the office city. Clarify the policy in the first recruiter call. If the company has a published geo-tier structure (like GitLab's compensation calculator), read it before the conversation.
Industry multiplier. The same functional role pays differently across industries. An FP&A manager at a SaaS company, a hospital system, and a manufacturing plant can have a $30,000 spread. Finance and tech pay at the top. Nonprofits and government pay at the bottom but often offset with benefits, stability, and loan forgiveness programs. Filter by industry if the data source allows it.
Comparing gross vs net (federal/state tax, 401k match, health insurance, equity)
A $120,000 offer letter is not $120,000 in your bank account. The difference between gross and net can be 25-35% depending on where you live and what you elect. Understand the deductions before you compare offers.
FICA. Social Security (6.2%) and Medicare (1.45%) take 7.65% off the top. On $120,000, that is $9,180. There is no way around it for W-2 employees. The Social Security wage base in 2026 is approximately $176,000 - earnings above that are subject to Medicare only (1.45%, plus the 0.9% additional Medicare surtax above $200,000 for single filers).
Federal income tax. Progressive brackets from 10% to 37%. For a single filer taking the standard deduction ($15,000 in 2026), the effective federal rate on $120,000 is roughly 14-16%. That means another $17,000-$19,000 gone. Married filing jointly with dependents shifts the math.
State income tax. The wildcard. Nine states have no income tax: Alaska, Florida, Nevada, New Hampshire, South Dakota, Tennessee, Texas, Washington, and Wyoming. California tops out at 13.3%, New York at 10.9%, Oregon at 9.9%. A $120,000 salary in Texas nets roughly $7,000 more per year than the same salary in California, purely on state tax. Factor this into any cross-state comparison.
401k match. Free money you lose if you ignore it. A 4% match on $120,000 is $4,800 - but only if you contribute at least that much. Some companies offer 50% match up to the IRS limit, some offer full match on the first 6%, some offer nothing. A generous 401k match is equivalent to a $3,000-$5,000 boost in total compensation versus a company with no match.
Health insurance premiums. The invisible compensation line. The average employer-sponsored individual plan premium in 2026 is around $8,500 per year. Employers typically cover 70-80% for individual plans and 65-75% for family plans. The difference between a company that covers 80% of your premium and one that covers 50% is roughly $2,500-$3,500 per year out of your pocket. A Gold plan with a $1,500 deductible is not the same as a Bronze plan with a $7,000 deductible. Read the plan summary. Ask for it before you accept.
HSA and FSA. If the company offers a high-deductible health plan with an HSA, check whether they contribute to it. Employer HSA contributions range from $500 to $2,000 per year. That is tax-free money for medical expenses. FSAs are use-it-or-lose-it but reduce your taxable income.
Build a spreadsheet. Normalize all offers to annual net take-home after tax, after health insurance, after 401k contribution (including the match). Compare the numbers side by side. A $110,000 offer with stellar benefits can beat a $125,000 offer with none.
Bonus and equity: how they fit the calculation
Base salary is table stakes. The comp package gets real when you layer in variable pay.
Annual bonus. Target bonus as a percentage of base salary. Entry-level roles: 0-5%. Mid-level individual contributor: 5-10%. Senior IC and manager: 10-20%. Director and above: 20-50% or more. Company size and profitability matter. Public companies pay bonuses with reasonable predictability. Startups either pay no bonus or tie it to funding milestones. Ask the recruiter: "What was the average bonus payout as a percentage of target over the last two years?" If they cannot answer, discount the target by 50% in your model.
Sign-on bonus. Common in tech, finance, and consulting. Typically $5,000 to $50,000 depending on seniority. Purpose: compensate you for unvested equity or an annual bonus you walk away from at your current employer. If a company cannot budge on base salary, ask for a sign-on bonus. It is a one-time line item that recruiters have more flexibility to approve.
RSUs (Restricted Stock Units). The standard equity vehicle at public companies and late-stage startups. You receive a grant worth a target dollar amount, split over a vesting schedule. The standard schedule: 4 years with a 1-year cliff (25% vests at month 12, the rest vests quarterly or monthly after that). If you get a grant of $80,000 in RSUs over 4 years, that is $20,000 per year in equity - but only if the stock price holds.
The risk: RSUs are taxed as ordinary income at vesting. If your $80,000 grant vests in year one and the stock doubled, you owe tax on $40,000 of income. If the stock dropped 50%, your grant is worth less but you still owe tax on the vest value. Many people sell enough shares to cover the tax ("sell to cover"). Know the policy.
Stock options. More common at startups. You get the right to buy shares at a fixed strike price. If the company exits or goes public above that price, you profit on the spread. If it does not, the options are worth zero. Private company options are illiquid and high-risk. When a startup tells you "we're giving you 20,000 options," the number is meaningless without three data points: the total fully diluted share count (so you know your ownership percentage), the latest 409A valuation (which sets the strike price), and the investors' liquidation preference. If the recruiter cannot give you these, the options are lottery tickets. Price them accordingly.
Refresh grants. Public tech companies issue additional RSU grants annually based on performance. These stack on top of your initial grant and are the primary way total comp grows beyond year 4. Ask about refresh policy. Companies with strong refresh cultures (Netflix, Meta, Google) generate significantly higher total comp over 4+ years than companies that front-load and stop.
ESPP (Employee Stock Purchase Plan). Many public companies let you buy stock at a 10-15% discount through payroll deductions. Free money if you can afford the cash flow reduction. A 15% discount on a stock you can sell immediately is a guaranteed return. Most people ignore ESPP. Do not be most people.
For a full breakdown of how compensation structures compare across employment types, read our guide on W-2 vs 1099 vs contract: which pays more after tax.
Band vs single number: how to present your range without anchoring low
After all the research, you arrive at a number, right? Wrong. You arrive at a band.
A negotiation band has three points:
- Walk-away floor. The number you will not go below, even with perfect benefits. This is your cost of living with a margin, plus the bottom of the market data. Below this, you decline.
- Target. What you believe is fair for your experience, your location, and the role. This should sit near the market median or slightly above if you bring a premium skill.
- Stretch ceiling. Above the market median but still defensible. For candidates with an uncommon combination of skills, domain expertise, or competing offers.
The most common trap: the recruiter asks your expectations and you say "$120,000." They hear the floor. Every time. You just gave away the negotiation.
Here is what you say instead:
"Based on what I've researched for this role in [city], with [X] years of experience and the scope we've discussed, my expectation is in the [floor] to [ceiling] range, depending on the full package. I'd want to understand the benefits structure, bonus target, and equity before narrowing that down."
This answer does three things. It shows you did the work. It gives a range that frames the ceiling as plausible. And it puts the recruiter in the position of justifying why their number belongs at the low end of your band, not you justifying why you deserve the high end.
For more on how to handle the comp conversation during the interview process, read our guide on how to ask for the salary at the interview.
Common mistakes in salary research
I see these over and over. Do not add your name to the list:
Using mean instead of median. One VP earning $400,000 pulls the mean of ten analysts earning $80,000 up to $112,000. The median is $80,000. The mean lies. The median represents the typical person in that role. Prefer medians. When the data source offers percentiles (P25, P50, P75), use them. The BLS OEWS and Levels.fyi give you percentiles. Glassdoor and Payscale often default to mean - dig for the median breakdown if available.
Ignoring regional variance. "Software engineers make $150,000" - where? San Francisco? Austin? Remote from Boise? Without a location, a salary number is useless. Filter every data point by metro area. If you are applying for remote, check the company's location-tier policy and filter accordingly.
Confusing title with scope. "Director" at a 50-person startup sometimes means you have two direct reports and no budget. "Director" at a public company means you run a division. Title inflation is real. Filter by years of experience, team size, and revenue responsibility, not by the words on the card.
Trusting unverified forums. Reddit, Blind, Fishbowl, and Discord have valuable signal - especially Blind for tech comp and Reddit's salary threads. But the data is unverified. Selection bias is extreme: people who post are either flexing or complaining. The middle 80% is silent. Use forum data as a directional sense check, never as a primary source.
Not updating your research. 2024 comp data is stale for 2026. Inflation, layoffs, hiring surges, remote-work policy changes - all of it shifts the band. Refresh your numbers every six months or before each interview cycle. A number that felt aggressive two years ago might be below market today.
Comparing gross only. I have said it, I will say it again: a $130,000 offer with a 6% 401k match, fully covered health insurance, and a 15% bonus target beats a $145,000 offer with no match, a high-deductible health plan, and no bonus. Annualize everything. Build the spreadsheet. Compare net total compensation, not headline salary.
Ignoring the company's moment. A company that just raised a Series C and is scaling fast pays above market. A company that did layoffs in Q3 and is backfilling critical roles pays below. The same role title at the same company in different quarters can have a $15,000-$25,000 spread. Pay attention to the business context.
Sources referenced in this guide (and that you should bookmark)
- BLS Occupational Employment and Wage Statistics (OEWS) - median and percentile wages by occupation and metro area
- Glassdoor Salaries - self-reported pay by company, title, and city
- Salary.com - employer-survey and HR-reported comp data with percentiles
- Payscale - skill-premium and certification-adjusted salary data
- Levels.fyi - tech industry comp broken down by company, level, and equity
- Levels TotalComp - verified total compensation with equity detail
- LinkedIn Salary - broad market ranges from LinkedIn's member base
- Bureau of Labor Statistics - Employment Projections - industry growth outlook and occupation projections
- IRS - 401k Contribution Limits 2026 - elective deferral and total contribution caps
Your career, your number
Knowing what your role pays is not arrogance. It is self-respect. You are not asking for a favor. You are selling time, skill, and output. The bare minimum you deserve is knowing the market price before you sit at the table.
Do the research. Build your band. And when the comp question comes - and it comes in every interview - you answer with data, not a wish.
Analyze your resume against a real job description on Trab and see whether your profile matches the band you are targeting.