1. LandOffer
  2. Blog
  3. All articles
  4. Levels.fyi vs. Payscale for an Offer: Compare Level, Location, and Compensation Components

Job search

Levels.fyi vs. Payscale for an Offer: Compare Level, Location, and Compensation Components

Use one fictional offer to separate base, target bonus, annualized equity and signing pay, then evaluate company-level and title-city data without inventing a market percentile.

10 min readLandOffer team

Levels.fyi and Payscale can both inform an offer discussion, but they may be describing different jobs and different kinds of pay. Levels.fyi can provide a company, level and location context for technology compensation. Payscale’s public salary research can provide a broader job-title and location reference. Before comparing either with an offer, align the role, geography, currency, date and compensation components. A total-compensation figure and a base-salary figure are not competing answers to the same question.

LandOffer.ai publishes this comparison using official public pages inspected October 22, 2026. We did not submit compensation, complete a survey, upload an offer or purchase data. The offer and worksheets below are fictional. You can browse recent roles on LandOffer.ai for discovery, then obtain the actual employer’s role and offer details before choosing comparable data.

A professional aligns two compensation reference cards with one job brief and matching context cues

Start with the offer’s identity, not a salary headline

Write down the employer, job family, internal level if confirmed, actual duties, work location and offer date. Preserve missing information as missing. “Software Engineer,” “Software Engineer III” and “senior” do not establish the same responsibility across employers. Years of experience can add context without proving a particular company level.

The current Levels.fyi Google L4 page for Greater Seattle identifies that company’s L4 as SWE III. In contrast, a standardized framework illustrated on Levels.fyi’s business data page labels its own L4 as Staff. Google’s internal level code and the platform’s standardized framework belong to different ladders. Matching the number four across those pages would create a false comparison.

Payscale’s Seattle Software Engineer page describes a job title and city, not Google L4. Its experience categories are not a substitute for the employer’s internal ladder. Start with the responsibilities and level information you can verify, then record the extent to which each reference matches.

Also distinguish Seattle city from Greater Seattle Area. They overlap, but their geographic populations are not identical. For remote work, ask which location determines the offer’s pay rather than automatically choosing the headquarters city or your preferred city.

Read the component labels before comparing amounts

On the inspected Levels.fyi company-level-location page, the summary used “Average Annual Total Compensation,” “Base Salary,” “First Year Stock” and “Bonus + Sign On.” The same page’s submission table used “Base | Stock (yr) | Bonus,” which does not erase the summary’s explicit first-year and combined-payment labels. The page therefore requires attention to the particular section you are reading; do not silently turn every stock or bonus entry into recurring annual cash.

Payscale’s Seattle headline was labeled average base salary in USD per year. The page displayed separate bonus, profit-sharing and total-pay ranges. A nearby experience paragraph described average total compensation including tips, bonus and overtime. Even within one page, the component being described can change.

Comparison question Levels.fyi sample Payscale Seattle sample
Who is represented? Google, Software Engineer, L4, Greater Seattle Software Engineer title, Seattle
Headline measure Average annual total compensation Average base salary per year
Component caution Summary first-year stock and combined bonus/sign-on labels Base, bonus, profit sharing and total pay are distinct
Level caution Google’s named L4/SWE III No Google-specific L4 designation on this page
What it cannot establish Your employer’s actual offer or guaranteed stock proceeds Your company’s level-specific package or grant terms

Compare base with base first. Then investigate what the remaining measures include and how they are annualized. Do not manufacture a comparable total by adding separate range endpoints: the highest displayed base, bonus and profit sharing may not describe the same people or the same pay period.

Worked worksheet: normalize one fictional offer

Miles has a fictional offer from Cedar Software for a Seattle-based engineering role. The title is Software Engineer III, but the employer has not supplied an internal level code or a cross-company mapping. All amounts below are stipulated USD figures for demonstrating recordkeeping, not market data or a recommended compensation target.

Offer component Fictional stated term Comparable record
Base $160,000 per year $160,000 annual base, before tax
Annual bonus Target of 15% of base, subject to performance and plan conditions $24,000 target; not guaranteed cash
Signing payment $10,000 once, with repayment terms still to clarify Separate first-year item, not recurring salary
Equity Grant described as $80,000 over four years, with 25% scheduled each year $20,000 per scheduled year using the stated reference amount; not assured proceeds
Benefits Employee costs and specific coverage not yet supplied Excluded from the numerical subtotal; questions remain
Identity Seattle role; Software Engineer III; level code unconfirmed No claim of equivalence to Google L4

His arithmetic is explicit: 15% of $160,000 is $24,000. Dividing the stipulated $80,000 reference grant into four equal scheduled portions produces $20,000 per year. Base plus target bonus is $184,000. Adding the grant reference produces a $204,000 annual comparison subtotal; adding the one-time signing payment produces a $214,000 first-year subtotal under those assumptions.

Neither subtotal is guaranteed cash or a tax estimate. The equity figure is a planning convention for this fictional example, not a valuation model. Actual award type, share quantity, reference price, vesting conditions, payout timing and any liquidity restrictions need their own evidence. An uneven vesting schedule would require different yearly entries instead of division by four.

His completed worksheet note says:

Compare $160,000 with a matching annual base measure. Keep $184,000 labeled base plus target bonus. Use $204,000 only where the reference includes a comparable annual equity component and target bonus. Keep $214,000 in a separate first-year column. Do not compare any of these totals with Payscale’s base headline, and do not label Cedar’s title Google L4.

His complete unsent clarification draft reads:

Could you confirm the internal level and scope of this role, the location used for compensation, the bonus conditions, and the equity award’s quantity, reference price and vesting schedule? Please also clarify the signing payment’s repayment terms and employee benefit costs. I want to keep annual recurring, conditional and one-time components separate before discussing the package.

Miles can now ask about the missing terms without accidentally presenting a four-year grant as annual pay or a signing payment as recurring salary. The worksheet does not decide whether he should accept, reject or negotiate the offer.

Four equal grant-reference tiles, a calendar and a separate one-time gift box illustrate different compensation periods

Treat methodology as context for the reference set

Levels.fyi’s published data explanation describes different evidence weights: supporting documents receive the most weight, followed by accounts and verified work emails, with anonymous submissions screened. It also describes spam and duplicate removal, flagged-data review and estimated levels where company ladders are not mapped. These are the provider’s descriptions, not an independent audit or proof that every visible entry has documentary verification.

That page distinguishes raw site data from subscription tools that add normalized cross-company benchmarking. A public company page does not establish that you have used those business features. Nor does a platform’s standardized level automatically become the employer’s confirmed level for your offer.

Payscale’s methodology page describes standardization and a MarketMatch process involving pay factors and recent matching profiles. It also discusses multiple datasets and employer subscription analytics. The current data-products FAQ explicitly separates its individual consumer salary survey from the employer HRIS-sourced Peer Global dataset behind its benchmarking products. Do not borrow an enterprise product’s refresh cadence or data source to characterize a public consumer salary page.

Save the source, exact page and displayed measure beside every figure you plan to discuss. “From Payscale” or “from Levels” alone leaves too much of the methodology and population unknown.

Check the date and denominator behind the sample

Payscale’s Seattle example showed 734 salary profiles and a July 10, 2026 update date. Its separate US Software Engineer page showed 25,510 profiles and a July 13 update date. The national count supports a broader reference; it does not resolve Seattle geography, Cedar’s responsibilities or a company-level match.

The Levels.fyi sample showed three unmasked recent submissions followed by an “unlock by adding your salary” prompt and masked rows. We did not submit anything or inspect the gated rows. Three visible rows are not evidence that the aggregate has only three observations. We also did not establish a denominator or exact observation window for the displayed summary from those rows alone.

Record a page’s update date separately from the date you opened it. A “2026” title is not the date of every underlying observation. Recent submissions do not establish that all records in a summary are new-hire offers; an existing employee’s reported compensation can answer a different question from a new offer.

A percentile describes a position within a defined distribution. It is not a probability that an employer will agree to your request, and it cannot be assigned from a mismatched headline comparison. If the summary’s observation window or sample count is not established, write “not established” beside it instead of borrowing the business-page demo’s last-12-months label.

Worked comparison memo: decide what each source can support

Elena is a fictional reader reviewing Miles’s completed worksheet. She uses the actual page scopes described above, without copying their salary numbers into his fictional offer or claiming either source generated his terms. Her completed source-fit record is:

Reference Fit to Miles’s offer Permitted use in her memo
Levels Google L4/Greater Seattle Specific employer and level, but Cedar-to-Google mapping unconfirmed; metro differs from city Illustrate a technology employer’s component structure; keep the level match unresolved
Payscale Seattle Software Engineer City and broad job family align; company and internal level do not Provide broader title/city context and inspect base separately
Payscale US Software Engineer Broad title and country; geography less specific Background context only; do not substitute national scope for Seattle
Cedar offer worksheet Same fictional offer, periods and assumptions recorded Source of offered terms, not market distribution evidence
Employer clarification Level, equity terms and benefit costs remain unanswered Identify the evidence needed before tightening the comparison

Her first memo sentence was: “The Levels total proves this offer is below market, while Payscale proves the opposite.” Her corrected paragraph is:

These pages describe different populations and pay measures. The Google reference has a company-level context that has not been mapped to Cedar. The Seattle Payscale page is geographically closer but not Cedar-level-specific. Neither supports a percentile for Miles’s offer. First resolve Cedar’s level and equity terms; then compare annual base with annual base and a documented equivalent total with the matching period. Keep one-time signing pay separate.

Her final decision record says: “Retain both references for different questions. Do not average their headline figures, choose the larger figure as a negotiating entitlement or invent a blended market rate. The next useful evidence is an employer clarification and a closer reference set, not a verdict based on title resemblance.”

This memo improves the comparison while leaving the personal employment decision to the candidate. It also explains why two useful platforms can disagree without either one establishing the right amount for a particular offer.

Choose the reference that answers your next question

Use a company-level-location view when you can verify the relevant ladder and want to examine technology compensation components. Use a broader title-and-city view to investigate a different reference population, while reading its specific pay labels and dates. Confirm the actual page’s scope each time; neither platform substitutes for written employer terms.

Keep the offer worksheet beside the source-fit memo so the next comparison preserves its component definitions, time periods and unresolved level mapping. For the wider search, browse recent roles on LandOffer.ai, then use the employer’s responsibilities and location to choose the data worth comparing.

Sources and Further Reading