Skip to main content

Glassdoor vs Levels.fyi: which salary tool should you use?

When you start researching salaries, you quickly encounter Glassdoor and Levels.fyi. Both are free, both are popular, and both will show you different numbers for the same role. This article untangles the two: what each database covers, how they differ, why they sometimes conflict, and which to trust for different scenarios. The goal is to know when to lean on each and how to combine them for a complete picture.

Quick definition: Glassdoor is a general salary database covering all industries and roles; Levels.fyi is a specialized tech-compensation database that breaks down equity and benefits in detail. They serve different purposes and are best used together, not as alternatives.

Key takeaways​

  • Glassdoor covers all industries and roles; Levels.fyi covers only tech and adjacent fields
  • Levels.fyi has more granular compensation breakdowns (base, equity, bonus, sign-on); Glassdoor gives broader ranges
  • Glassdoor data skews optimistic on bonus and equity; Levels.fyi is more precise about timing and vesting
  • If you are in tech, check both. If you are outside tech, Glassdoor is your primary source
  • The two databases often show different ranges because they measure different populations

Glassdoor: breadth and accessibility​

Glassdoor (glassdoor.com) is the largest general salary database. It aggregates compensation information from millions of employees across all industries, geographies, and role levels.

Who uses Glassdoor:

  • Employees at any company looking to understand their market value
  • Job seekers comparing offers
  • Employers benchmarking their compensation

What data Glassdoor provides:

  • Base salary range (usually with 25th, 50th, 75th percentiles)
  • Bonus range (as a percentage of base or dollar amount)
  • Stock awards/equity (for public and some private companies)
  • Total compensation estimate
  • Company-specific reviews (including pay satisfaction comments)
  • Role-specific and location-specific data

How to read Glassdoor:

Search "Software Engineer" in "San Francisco, CA" and you will see:

  • Base salary: $150,000–$185,000 (median: $167,500)
  • Bonus: 10–25% (median: 15%)
  • Stock awards: $30,000–$100,000 annually (median: $65,000)
  • Total compensation: $210,000–$330,000 (median: $245,000)

This gives you a range, but the ranges are quite wide. The base is more reliable; the equity estimate is often optimistic.

Strengths:

  • Covers every occupation and industry
  • Company-specific data (search "Facebook Software Engineer" vs. "Google Software Engineer")
  • Large dataset (millions of reports)
  • Free and easy to use
  • Good for entry-level, mid-level, and executive roles
  • Shows company reviews alongside salary data
  • Role progression data (how salary changes as you advance)

Limitations:

  • Equity and bonus estimates are often overstated. Glassdoor reports are typically self-reported maximum comp, not what people actually realized (which depends on stock vesting, bonus achievement, etc.)
  • Data lags reality by 6–12 months
  • Heavily skewed toward tech and finance (if you work in healthcare or manufacturing, data is sparser)
  • Self-reported, so accuracy varies by individual and by company
  • The ranges are wide, making it hard to pin down a specific target

Example of Glassdoor optimism:

An engineer reports they received a "sign-on bonus of $50,000" and "stock grant of $100,000 annually." Glassdoor captures both. But in reality:

  • They might only keep that equity if they stay four years (vesting over time)
  • The stock price might drop 30% (vested value becomes $70,000)
  • The sign-on bonus was to cover leaving a previous bonus; it is not typical annual comp

Glassdoor shows the headline numbers; it does not show the caveats.

Levels.fyi: precision for tech​

Levels.fyi (levels.fyi) is a specialized database focused exclusively on technology and adjacent high-paying fields. It collects compensation data from engineers, product managers, designers, and data scientists at startups, growth companies, and Big Tech.

Who uses Levels.fyi:

  • Tech workers comparing offers and planning moves
  • Engineers deciding between startups and Big Tech
  • People trying to understand equity grants and vesting

What data Levels.fyi provides:

  • Base salary
  • Sign-on bonus (year 1 only)
  • Annual stock grant (typical vesting schedule)
  • Refresher grants (year 2, 3, 4 equity)
  • Annual bonus (percentage or dollar amount)
  • Total comp breakdown (year 1 vs. ongoing)
  • Level-specific bands (e.g., L3, L4, L5 at Google)

How to read Levels.fyi:

Search "Software Engineer" at "Google" in "Mountain View":

  • Level 4 (mid-level):
    • Base: $150,000–$165,000
    • Sign-on: $50,000–$60,000 (year 1 only)
    • Annual equity grant: $150,000–$170,000 (vests over 4 years, so year 1 = ~$40,000)
    • Annual bonus: 15–20% of base
    • Year 1 total: $280,000–$345,000
    • Year 2+ total: $245,000–$290,000 (no sign-on, but continuing equity)

This level of detail is invaluable for understanding how total comp is structured and changes over time.

Strengths:

  • Highly specific on equity (vesting schedule, refresher grants, true annual comp)
  • Accurate for Big Tech (Google, Meta, Amazon, Apple, Microsoft) and growth-stage startups
  • Distinguishes between "year 1 comp" (with sign-on) and "ongoing comp" (equity continuing to vest)
  • Shows how compensation structure differs by company (Google equity is larger; Amazon cash is larger; etc.)
  • User base is tech-savvy and reports accurately
  • Free
  • Helpful for level progression (see how much a jump to L5 gains)

Limitations:

  • Limited to tech and tech-adjacent roles (no accountants, lawyers, doctors, etc.)
  • Heavy bias toward large tech companies and well-funded startups (sparse data for mid-market or older-school tech companies)
  • Concentrated in major hubs (SF, NYC, Seattle) with less data for secondary markets (Austin, Denver, etc.)
  • Self-reported, though the community is usually accurate
  • Does not include company-specific reviews or culture data

Example of Levels.fyi precision:

Two offers, both showing "$250,000 total comp" on the surface:

Offer A (Google L4):

  • Base: $160,000
  • Sign-on: $50,000 (year 1 only)
  • Annual equity grant: $150,000 (vest over 4 years)
  • Year 1 comp: $310,000
  • Year 2+ comp: $260,000

Offer B (startup series C):

  • Base: $150,000
  • Sign-on: $30,000 (year 1 only)
  • Annual equity grant: $70,000 (vest over 4 years, but startup fails)
  • Year 1 comp: $220,000
  • Year 2+ comp: $170,000

On paper, both say "total comp" around $250,000, but the structures are very different. Levels.fyi helps you understand which offer is actually better and how much risk is in the equity portion.

Why do they show different numbers?​

Glassdoor and Levels.fyi often show different salary ranges for the same role at the same company. This is not because one is wrong; it is because they measure different populations.

Reasons for divergence:

  1. User base: Glassdoor has a broad user base (all industries); Levels.fyi users are primarily tech workers. Tech workers at Levels.fyi may be more savvy negotiators, resulting in higher reported comps.

  2. Recency: Levels.fyi is updated more frequently (users report recently-negotiated offers), while Glassdoor data is older.

  3. Definition of compensation: Glassdoor's "total comp" is an estimate; Levels.fyi's is disaggregated (base + bonus + equity), making it clearer.

  4. Self-selection: People who report high salaries on Glassdoor may be disproportionately at the top of their company. People reporting to Levels.fyi are reporting a recent offer (a more representative sample of what is actually attainable).

  5. Equity volatility: Glassdoor estimates equity based on grant size; Levels.fyi shows the grant but notes vesting. Stock appreciation/depreciation is not reflected in either, but Levels.fyi users understand this better.

Example of divergence:

You search "Senior Software Engineer" at "Meta" in "Menlo Park":

Glassdoor:

  • Base: $170,000–$200,000 (median: $185,000)
  • Bonus: 20–30% (median: 25%)
  • Stock awards: $200,000–$400,000 (median: $300,000)
  • Total comp estimate: $350,000–$600,000 (median: $470,000)

Levels.fyi:

  • Base: $175,000–$190,000
  • Annual equity grant: $250,000–$350,000 (vest over 4 years)
  • Year 1 comp (with sign-on $100,000): $425,000–$540,000
  • Year 2+ comp: $325,000–$440,000

Why the difference?

  • Glassdoor's upper estimates are inflated (some outliers reporting much higher; some conflating total comp with realized gains)
  • Levels.fyi's range is more recent and based on actual recent offers (not years-old reports)
  • Glassdoor shows "stock awards $400,000" but that does not account for vesting; if you leave after 2 years, you never realized it
  • Levels.fyi explicitly shows vesting, so you know $250,000 grant vests at $62,500 per year

For this role, Levels.fyi is more trustworthy because its user base has recent, verified offers.

How to use both together​

The best practice is not to choose one, but to use both, understanding their strengths.

Decision tree:

Is your role in tech (software engineer, product manager, data scientist, designer)?
├─ Yes: Use Levels.fyi as primary, Glassdoor to cross-check and see the broader market
└─ No: Use Glassdoor as primary. Check BLS if available. Do not use Levels.fyi

For tech roles (use both):

  1. Check Levels.fyi first. Get the granular breakdown of base, equity, and bonus.
  2. Cross-check on Glassdoor. See the broader range and company reviews.
  3. If they conflict: Levels.fyi is usually more current and accurate. Glassdoor is conservative on base, optimistic on equity.
  4. Adjust for specifics: If Levels.fyi shows data for "Mountain View" and you are negotiating in "Denver," reduce by the geographic premium (typically 10–20%).

For non-tech roles (use Glassdoor primarily):

  1. Check Glassdoor for the role and location.
  2. Cross-check with BLS data (official government stats).
  3. Do not use Levels.fyi (it will not have your role).

Specific comparison: Google Software Engineer​

Let me walk through a real example to show how both databases work in tandem.

Scenario: You have an offer from Google for a mid-level Software Engineer role (L4) in Seattle.

Glassdoor data (search "Software Engineer" at "Google" in "Seattle, WA"):

  • Base salary: $155,000–$180,000 (median: $167,500)
  • Bonus: 20–30% of base (median: 25%)
  • Stock awards: $200,000–$350,000 (median: $275,000)
  • Total comp (Glassdoor estimate): $350,000–$510,000 (median: $432,500)

Levels.fyi data (search "Google" → "SDE" → "L4" → "Seattle"):

  • Base: $160,000–$175,000
  • Sign-on bonus: $60,000–$75,000 (year 1 only)
  • Annual equity: $200,000–$280,000 (4-year vest, so year 1 = $50,000–$70,000)
  • Year 1 comp: $270,000–$320,000
  • Year 2+ comp: $210,000–$245,000

Interpretation:

  1. Base salary: Both databases show $160,000–$175,000. This is reliable.
  2. Year 1 comp: Levels.fyi shows $270,000–$320,000 (accounting for sign-on and equity vesting). Glassdoor shows $432,500 (inflated because it estimates full annual equity, not vested value).
  3. Ongoing comp: Levels.fyi shows $210,000–$245,000. Glassdoor's estimate is overstated because Glassdoor does not adjust for vesting schedules.
  4. What to negotiate: Your anchor should be Levels.fyi's year 2+ comp ($210,000–$245,000 total). If you want to push, you can negotiate on sign-on bonus ($65,000 is reasonable) or base ($170,000 is reasonable). Asking for higher equity is less effective because equity vesting is fixed.

Real-world examples​

Case: Marcus, senior engineer at Meta, 2023

Marcus received a Meta offer and saw conflicting information:

  • Glassdoor said: $350,000–$550,000 total comp
  • Levels.fyi said: $300,000–$400,000 year-1 comp, $250,000–$320,000 ongoing comp

Marcus realized Glassdoor was inflating equity (not adjusting for vesting), and Levels.fyi showed the realistic ongoing number. He negotiated to $180,000 base (vs. the offer of $175,000) and a $90,000 sign-on (vs. $75,000). The Levels.fyi data gave him confidence that these asks were in-market. Glassdoor would have made him think the total comp was already very high and he should not push.

Case: Priya, product manager at Stripe, 2024

Priya saw Glassdoor data showing $250,000–$350,000 for her PM role. Levels.fyi showed $200,000–$250,000. The discrepancy concerned her. She checked Levels.fyi for context and saw that Stripe's compensation is more base-heavy (less equity) compared to Google. She negotiated to $160,000 base (Levels.fyi range: $145,000–$165,000) and was satisfied.

Common mistakes​

Mistake 1: Using Glassdoor equity estimates as firm numbers. Glassdoor says "$200,000 stock awards" but does not account for vesting, dilution, or stock price changes. Use it as a rough indicator, not a forecast.

Mistake 2: Treating Levels.fyi as universal. Levels.fyi only has tech jobs. If you are a PM at a non-tech company, Levels.fyi will not help.

Mistake 3: Ignoring the sign-on bonus. Sign-on is year 1 only and is often negotiable. If Levels.fyi shows "$80,000 sign-on," you can push for more. If it shows "$30,000," you have room.

Mistake 4: Comparing across levels. A L4 and L5 are different roles with different comp. Use the correct level in Levels.fyi.

Mistake 5: Not adjusting for location. Both databases support location filtering. Use it. A Seattle engineer earns less than an SF engineer at the same company.

FAQ​

Is Levels.fyi more accurate than Glassdoor?​

For tech roles, yes—Levels.fyi has more recent, detailed reports and a user base that is more likely to accurately understand their comp. For non-tech roles, Glassdoor is your only option (Levels.fyi has no data).

Can I use Levels.fyi data when negotiating with a non-Big-Tech company?​

Cautiously. Levels.fyi is heavily skewed toward high-paying companies. A startup Series C might pay 20% less than a Levels.fyi Big Tech comp. Use Levels.fyi as the high end and adjust down if the company is smaller or earlier-stage.

What if Glassdoor and Levels.fyi show very different numbers?​

Trust Levels.fyi for tech if the data is recent (posted in the last 6 months). Check the post date. Older Levels.fyi data can be stale too. If both are recent and conflict, triangulate with a third source (BLS, a recruiter, a peer).

Should I report my salary to Glassdoor or Levels.fyi?​

That is your choice. Contributing anonymous data helps others. Many people report to both. Be accurate; do not inflate. If you received a $50,000 sign-on and $100,000 equity grant, report both, not a rounded "$150,000 comp."

Can I use Levels.fyi data from a company I did not work at?​

Yes. Levels.fyi is specifically designed to help people compare external companies and understand what they should be paid at a target employer. That is its primary use case.

What if my specific company is not on Levels.fyi?​

Use Glassdoor. Levels.fyi has data for ~50 large tech companies and is sparse for others. If your company is not listed, you can search "similar companies" (similar size, stage, industry) and estimate an adjustment.

Summary​

Glassdoor is the broad, general salary database; it covers all industries but skews optimistic on equity and bonus. Levels.fyi is the specialized tech database; it is precise on equity vesting and year-by-year compensation but only covers tech and high-paying roles. For tech roles, use Levels.fyi as your primary source and cross-check Glassdoor. For non-tech roles, use Glassdoor and BLS. Both databases are self-reported, but Levels.fyi's tech user base is typically more accurate. Combine them to avoid over-relying on either one.

Next​

→ The counter-offer script