Skip to content
Published on

Compensation Benchmarks & Salary Transparency Tools 2026 Complete Guide - Levels.fyi · Carta Comp · Pave · OpenComp · Glassdoor · Blind · JobPlanet · Saramin · Wanted · OpenWork · Ten-shoku Kaigi Deep Dive

🇰🇷KO🇺🇸EN🇯🇵JA
Split View필사 모드
Authors

Intro — May 2026, the era of "working without knowing your number" is ending

Until the early 2020s, "do not ask coworker salaries" was a workplace default. As of May 2026, that default is flipping fast. The EU Pay Transparency Directive becomes binding on June 7, five US states (New York, California, Colorado, Washington, Illinois) require salary ranges in job postings, and Gen Z treats "we do not disclose" as a red flag for low pay.

The tooling landscape shifted too. Levels.fyi has become the de facto standard for big-tech compensation data, Pave acquired OpenComp in February 2024 and solidified its lead in B2B comp benchmarking, and Carta extended from cap tables into Carta Comp plus Carta AI. In Korea, JobPlanet strengthened its salary dataset and Wanted layered AI matching onto both hiring and comp. In Japan, OpenWork (formerly Vorkers) holds the JobPlanet-equivalent position, while Ten-shoku Kaigi rides the Recruit Holdings brand to scale.

This post is not a marketing matrix. It is an honest map of where to get which data, how far to trust it, and how to actually use it in negotiation.

Why pay transparency became the 2026 default — law, generation, market

Three forces converged.

  1. Law: The EU Pay Transparency Directive becomes binding across all member states on June 7, 2026. Job postings must publish salary ranges, employees can request same-job average pay, and employers with 200+ headcount must report gender pay gaps. In the US, state-level rules in NY/CA/CO/WA/IL together cover roughly 35% of the labor force.
  2. Generation: Gen Z routinely shares salaries on Levels.fyi, Blind, and Reddit. "We do not disclose" reads as a signal of below-market pay.
  3. Market: Big tech's heavy RSU mix and complex option structures push workers to ask "what is my package actually worth?" Base-only comparisons no longer suffice.

When these three forces meet, compensation data stops being HR's secret and becomes a public good on the labor side.

Crowdsourced comp data — why Levels.fyi became the de facto standard

The crowdsourced camp (employees self-reporting their own packages) shakes out as follows in 2026.

  • Levels.fyi: The de facto standard for big-tech SDE/PM/designer comp. Key differentiator: it breaks Total Comp into base + stock (RSU annualized over 4 years) + signing + bonus + perks.
  • Blind: Anonymous corporate network with email-domain verification. Self-reported salary plus discussion boards.
  • Glassdoor: The oldest player. Stronger in general industry, sales, HR, and management roles than in big tech.
  • Indeed Salary: Estimated pay derived from Indeed's job-posting corpus. Direct beneficiary of US state-level disclosure laws.
  • PayScale: A veteran from the 2000s, with both employee and employer SaaS.
  • Salary.com: US job-level pay statistics with consulting attached.
  • Comprehensive.io: Founded 2022. Crawls and aggregates salary ranges published in job postings, surfacing "what is actually posted." Direct beneficiary of US pay disclosure laws.

Crowdsourced data has well-known limits.

  1. Top-bias self-selection: Successful negotiators submit more often.
  2. Big-tech bias: Over 70% of Levels.fyi submissions are SDE roles.
  3. Snapshot mismatch: Initial offer value and current value (after stock-price moves) are easily confused.

These limits are filled in by the B2B survey camp covered below.

Levels.fyi deep dive — the five-tier verification system and "TC" as the unit of comparison

What makes Levels.fyi different from other crowdsourced sites is twofold. First, it fixed the standard package representation as "Total Compensation = Base + Stock (annual) + Bonus + Signing (amortized)," making cross-offer comparison possible. Second, it labels each submission with a verification tier (email, offer letter, paystub, work email, peer-verified).

Levels.fyi's core surfaces:

  • End-to-End: Distribution of packages by company + level + role + location (median, 25/75 percentile, mean).
  • Negotiation School: A negotiation guide, including paid 1:1 coaching.
  • Internships: Hourly/weekly/stipend data for intern roles.
  • Calculator: A side-by-side compare tool with FX, tax, and cost-of-living adjustments.

The screen shape looks roughly like the JSX sketch below.

// Levels.fyi /companies/google/salaries/software-engineer page (conceptual)
type Comp = {
  totalYearlyCompensation: number
  baseSalary: number
  stockGrantValue: number
  bonus: number
  level: string
  location: string
  yearsAtCompany: number
  yearsOfExperience: number
}

export function CompTable({ rows }: { rows: Comp[] }) {
  return (
    <table>
      <thead>
        <tr>
          <th>Level</th>
          <th>YoE</th>
          <th>Location</th>
          <th>Base</th>
          <th>Stock</th>
          <th>Bonus</th>
          <th>TC</th>
        </tr>
      </thead>
      <tbody>
        {rows.map((r) => (
          <tr key={r.level + r.location + r.yearsAtCompany}>
            <td>{r.level}</td>
            <td>{r.yearsOfExperience}</td>
            <td>{r.location}</td>
            <td>{r.baseSalary}</td>
            <td>{r.stockGrantValue}</td>
            <td>{r.bonus}</td>
            <td>{r.totalYearlyCompensation}</td>
          </tr>
        ))}
      </tbody>
    </table>
  )
}

Practical tip: When reading a company's data on Levels.fyi, require (a) submissions within the last 12 months, (b) at least 30 entries, and (c) the IQR at your level and location. Do not get swayed by a single outlier.

Blind — anonymous community plus self-reported comp

Blind is a work-email-verified anonymous network that launched in Korea in 2013 and expanded into US, Indian, and Japanese big tech. The comp data is not as cleanly tabular as Levels.fyi, but Blind is strong in two areas.

  1. Company board sentiment: Manager reputation, team vibes, layoff rumors — qualitative signal. Adoption among Korean IT firms is very high.
  2. Anonymity: Work-email verification plus domain masking. Small companies, however, easily lose anonymity.

Blind's downsides: (a) boards skew toward gossip with a poor signal-to-noise ratio, and (b) the data is hard to analyze outside the app. Treat Blind as a complement to Levels.fyi.

Glassdoor — the veteran for US general industry, reviews, and ratings

Glassdoor launched in 2008, was acquired by Recruit Holdings (Indeed's parent) in 2018, and has lost ground to Levels.fyi in big-tech comp. It still leads in:

  • General-industry pay: Healthcare, education, government, retail, manufacturing.
  • Company reviews and ratings: CEO approval, work-life balance, diversity scores.
  • Interview reviews: The largest database of interview questions.

Glassdoor's data limits: self-reports without verification, and very wide intra-role variance. Read distributions and medians, not point values.

Comprehensive.io — the rising aggregator of posted salary ranges

Founded in 2022, Comprehensive.io is a direct beneficiary of US state pay-disclosure laws. It crawls and aggregates salary ranges from job postings, organized by company, role, and location.

Differences from crowdsourced sites:

  1. Source is the company's own posting, not a self-report — no submission bias.
  2. Range (min~max) centric, not TC.
  3. Strong coverage of newer companies: Levels.fyi skews big tech; Comprehensive.io captures Series A through C startups.

Caveats: (a) almost no non-US coverage, and (b) significant gaps for roles where disclosure is not required (most managerial and senior individual contributor titles).

B2B comp SaaS — Pave's OpenComp acquisition locked in the lead

The B2B comp benchmarking SaaS used by HR teams to design internal pay policy saw a big event in February 2024: Pave acquired OpenComp, effectively cementing Pave's market lead. As of 2026, the lineup looks like this.

  • Pave: Founded 2020. Market leader after the OpenComp acquisition. Direct HRIS integrations (Workday, BambooHR, Rippling) and real-time benchmark refresh.
  • Carta Comp: Strength in equity-inclusive package benchmarks, combining cap-table data.
  • Mercer Comp Surveys: A global consulting giant founded in 1937. Best industry and country depth.
  • Aon Radford Surveys: The de facto standard for big-tech comp surveys. Started as Radford in 1976.
  • WTW (Willis Towers Watson) comp surveys: Strong in global executive comp.
  • Salary.com CompAnalyst: Aimed at mid-market and SMB.

Pave's conceptual API call looks like this.

# Pave benchmark API request (conceptual)
import requests

resp = requests.get(
    "https://api.pave.com/v1/benchmarks",
    headers={"Authorization": "Bearer $PAVE_API_KEY"},
    params={
        "job_family": "Software Engineering",
        "level": "L4",
        "location": "San Francisco Bay Area",
        "company_stage": "Series C",
        "headcount_range": "201-500",
    },
)
data = resp.json()
print(data["base_salary"]["p25"], data["base_salary"]["p50"], data["base_salary"]["p75"])
print(data["equity_value"]["p50"], data["total_comp"]["p50"])

The core advantage is direct HRIS integration. Once a customer connects Workday/Rippling/BambooHR, Pave streams anonymized data into the benchmark pool in near-real-time. This live-data model reflects market moves 6–12 months faster than Mercer/Radford's annual surveys.

How OpenComp ended up inside Pave — the February 2024 merger

Like Pave, OpenComp started in the early 2020s as a startup-friendly comp benchmarking SaaS. Its differentiators were aggressive pricing and an equity dilution simulator. After funding pressure in late 2023, OpenComp announced its merger with Pave in February 2024, and the standalone OpenComp product is being wound down into Pave.

The OpenComp domain now redirects to Pave's marketing pages, and existing customers have been migrated to Pave contracts — a clear signal of market consolidation.

Traditional surveys — where Mercer, Aon Radford, and WTW still dominate

Even as Pave/Carta take share with live SaaS, traditional surveys remain the standard in:

  1. Public-company executive comp: SEC filings plus Mercer/Radford surveys are the comp-committee baseline.
  2. Global multinationals: Mercer is deepest for 50+ country comparisons.
  3. Non-tech industries: WTW leads in manufacturing, energy, and financial-services-specific surveys.

Representative surveys:

  • Mercer Total Remuneration Survey (TRS): 50+ countries, 800+ job codes.
  • Aon Radford Global Technology Survey: The big-tech comp standard. Quarterly updates.
  • WTW Global 50 Survey: Executive comp across 50 markets.

These are (a) expensive (tens of thousands to hundreds of thousands USD per year), (b) bundled with consulting contracts, but (c) effectively required for IPO prep and comp-committee documentation.

Equity calculators — why Carta grew from cap tables into a full platform

In an era when more than half of a startup package is equity, cap-table-plus-equity-math has become the heart of comp tooling. Camp lineup:

  • Carta: #1 cap-table SaaS. Took a credibility hit in 2024 around alleged misuse of secondary-sale data, but rebounded with the Tactyc acquisition strengthening LP analytics.
  • Pulley: A Carta challenger with simpler UX and lower pricing.
  • Shareworks (Morgan Stanley): Strong in public-company ESPP/RSU administration.
  • EquityZen, Forge Global: Secondary markets for private shares.
  • Index Ventures Founder Salary Tool: Founder pay guidance from seed through Series C.

Carta leads not just for cap tables but because (a) 409A valuations, (b) equity exercise simulation, and (c) tax-optimization recommendations all live under one roof.

RSU / NSO / ISO / ESPP — equity 101 in one pass

Terminology:

  1. RSU (Restricted Stock Units): Standard at public companies. Taxed as ordinary income plus payroll taxes on vest. Value = stock price.
  2. ISO (Incentive Stock Options): Employee stock options at private companies. Strike price is fixed; exercise can trigger AMT. If you meet the 2-years-from-grant plus 1-year-from-exercise holding rule, sale is taxed at long-term capital gains.
  3. NSO/NQSO (Non-Qualified Stock Options): Fewer issuance restrictions than ISOs. Ordinary income at exercise; capital gains on sale.
  4. ESPP (Employee Stock Purchase Plan): A program where public-company employees contribute payroll percentages over an offering period to buy company stock at a discount. Up to 15% discount under IRS Section 423.
  5. 409A valuation: A fair market value valuation under IRS Section 409A for US private companies, refreshed quarterly to annually, used to set ISO/NSO strike prices.
  6. Single-trigger vesting: Acceleration on a single event such as acquisition.
  7. Double-trigger vesting: Acceleration requires both an acquisition and an involuntary or good-reason termination — better for retaining acquired employees.

Without this vocabulary you cannot evaluate an offer letter accurately.

US state pay transparency laws — five states cover a third of the workforce

US state-level job posting salary disclosure status as of May 2026:

  • Colorado Equal Pay for Equal Work Act: Effective January 2021. Postings must include pay range plus benefits. Violations cost USD 500–10,000 per occurrence.
  • New York Pay Transparency Law: Effective September 2023. Employers with 4+ employees must include pay ranges in postings.
  • California SB 1162: Effective January 2023. Employers with 15+ must publish pay ranges in postings; 100+ must file demographic pay reports.
  • Washington State: Effective January 2023. Employers with 15+ must publish pay range plus benefits.
  • Illinois (SB 1480 follow-on): Effective January 2025. Employers with 15+ must publish pay ranges.

Summed, the five states cover roughly 35% of US labor — which is exactly why a tool like Comprehensive.io exploded.

EU Pay Transparency Directive — binding from June 7, 2026

The EU Pay Transparency Directive (2023/970/EU) must be transposed into national law by all 27 member states by June 7, 2026. Key provisions:

  1. Salary ranges in job postings: Candidates must receive pay information up front.
  2. Right to information: Employees may request average pay for the same job plus gender pay gap data.
  3. Gender pay gap reporting: Every 3 years for 100+ headcount, every year for 250+.
  4. Joint pay assessment when gap exceeds 5%: If an unjustified gender pay gap of 5% or more persists, a joint assessment with worker representatives is required.
  5. Reversed burden of proof: In pay-discrimination litigation, the burden shifts to the employer.

The EU rules are significantly stricter than US rules, and global firms are trending toward unifying global policies to the EU standard.

Korean salary platforms — JobPlanet, Saramin, Wanted, Rocketpunch

Korea does not yet have US/EU-style legal disclosure mandates (public-sector employers excepted). The de facto market is formed by the following platforms.

  • JobPlanet: Founded 2014. Korea's Glassdoor equivalent. Company ratings, salary data, and interview reviews. Strong on unlisted Korean IT firms.
  • Saramin: Korea's #1 general job board. Salary information attached to job postings.
  • Wanted: AI-recommendation job platform. Recommends salary ranges by back-solving from past accepted-offer data.
  • Rocketpunch: Startup and tech focus. Series-stage average comp guidance.
  • JobKorea: One of Korea's two largest job boards alongside Saramin. Pay guides and role medians.
  • Blind: Very high adoption among Korean IT employees.
  • Jobaliyo (잡알리오): Korea's national public-sector job-information network.
  • ALIO: Public-institution management information disclosure system, including pay and benefits.

Korean market characteristics:

  1. Salary negotiation culture is less developed than in the US or EU; signing bonuses and RSUs are rare outside big tech.
  2. Public-sector pay is 100% disclosed via ALIO, providing a private-sector comparison anchor.
  3. JobPlanet salary data is a complement: For big-tech packages, Levels.fyi is deeper.

JobPlanet deep dive — Korea's standard for salary data

JobPlanet uses a give-to-get model: you must write your own company review or salary entry to unlock others. Data structure:

  • Salary: Median and average by role, tenure, and employment type.
  • Company reviews: Five axes scored out of 5 — work-life balance, benefits, culture, leadership, promotion opportunity.
  • Interview reviews: Questions, difficulty, accept/reject result.

Limits: (a) weak coverage of big-tech RSU and signing bonus structures, (b) no verification of self-reports, (c) anonymity easily breaks for small companies.

Wanted — AI matching and salary recommendations

Wanted, founded in 2015, is a Korean AI-driven hiring platform. On the comp side:

  1. Salary recommendations trained on past offers: It learns from apply → accept → join data to estimate "what salary band someone with your profile is likely to receive at this company."
  2. Direct salary comparison: Distribution of accepted offers among candidates with similar role, tenure, and skill profile.
  3. Headhunter matching: For senior roles, salary negotiation runs through partnered headhunters.

Wanted's limits: high accuracy in large-sample roles (dev, design, PM) but lower reliability in small-sample roles.

Japanese salary information market — OpenWork, Ten-shoku Kaigi, Lighthouse, CariCone

Japan is structurally different from the US and Korea.

  • OpenWork: Launched in 2007 as Vorkers, rebranded to OpenWork in 2019. The JobPlanet equivalent for Japan. Company scores, salary, and employee reviews.
  • Ten-shoku Kaigi (転職会議): Operated by Recruit Group. Plugged into Recruit's massive recruiting data.
  • Lighthouse (ライトハウス): Operated by Lightup. Company ratings and interview reviews. An OpenWork challenger.
  • CariCone (キャリコネ): Operated by Grouplus. Reviews plus pay information.
  • e-shains: Niche, foreign-capital and IT focus.
  • Japan Ministry of Health, Labour and Welfare Basic Wage Structure Survey: Official statistics on wage distribution by industry, region, age, and gender.

Japanese market characteristics:

  1. New-graduate batch hiring (shinsotsu-ikkatsu-saiyo): All April joiners start at the same compensation level, so negotiation leverage is near zero.
  2. Base + multi-layer allowances (kihonkyu plus shokumu-teate, tsukin-teate, jutaku-teate): Multi-layered structure. Base-only comparison is misleading.
  3. Retirement allowance (taishoku-kin): Lump-sum payout at separation, based on tenure × company multiplier. Unlike US RSUs, it is paid at exit.
  4. Gender pay gap disclosure mandatory since July 2022: Employers with 301+ employees must disclose annual gender pay differentials.

OpenWork deep dive — the JobPlanet-equivalent leader in Japan

OpenWork (formerly Vorkers) holds roughly 60% of the Japanese job-review market. Data structure:

  • Company score: Eight axes (overtime, fairness of evaluation, employee morale, leave, work environment, growth, diversity, management) scored out of 5.
  • Salary: Average and distribution by role. Base + bonus + overtime are sometimes blurred.
  • Employee reviews: Free-form qualitative text.

Limits: (a) weak coverage of foreign-capital IT firms, (b) almost no big-tech RSU data, (c) significant self-report bias.

Ten-shoku Kaigi — Recruit's Japanese Glassdoor

Ten-shoku Kaigi is operated by Recruit Holdings, Japan's largest recruiting platform. Strengths:

  1. Plugged into Recruit's massive job-posting and accepted-offer dataset: Postings + outcomes + reviews in one place.
  2. Advertiser revenue model: Some employers pay to boost their pages — a potential bias vector.
  3. Parent brand: Recruit itself is the center of Japan's recruiting market.

Downsides: parent-company dependence raises (a) subsidiary-data trust questions and (b) data depth is shallower than OpenWork.

AI in comp — Pave forecasting and Carta AI

Compensation SaaS adopted AI in earnest between 2024 and 2026.

  1. Pave Benchmark Predictions: Given company stage, role, location, and headcount, it forecasts the market median 12 months out — useful for designing forward-looking pay policies.
  2. Carta AI: Combined with cap-table data, it recommends "what RSU grant should this employee get in the next round."
  3. Levels.fyi Peer Comparison: Enter your package and see where you sit in the distribution at your company, level, and location.
  4. Personalized Offer Optimization: Some SaaS dynamically price offers using candidate-specific signals on negotiation propensity and mobility.

AI recommendation limits: training-data bias transfers directly. If the training corpus contains a gender pay gap, the recommendation reproduces it. That is the rationale for tools like Syndio and Trusaic.

Syndio, Trusaic — pay equity analytics

Syndio (founded 2017) and Trusaic (founded 2010) lead the pay equity SaaS category. Core features:

  1. Regression-based gender/race pay gap analysis: Controlling for role, tenure, and location to surface "unjustified" gaps.
  2. Remediation simulation: Compute the cost of closing the gap to zero.
  3. EU Directive compliance reports: Auto-generated.
  4. CA SB 1162, IL PEPA compliance reports.

Syndio is used by Netflix, Slack, Salesforce, Adobe, and others. Trusaic is strong in mid-market and industrial customers.

The EU Pay Transparency Directive coming into force in June is expected to drive explosive growth in this category.

Compensation negotiation tools - Levels.fyi Negotiation, Candor, TBD

A negotiation-side toolset for offer recipients also formed.

  • Levels.fyi Salary Negotiation: 1:1 coaching (15–30 minute calls, typically a few hundred USD). Levels.fyi's data edge directly powers the coaching content.
  • Candor: Negotiation coaching and offer analysis. Free guides plus paid 1:1.
  • TBD: A newer negotiation SaaS using AI-driven offer analysis.
  • Ten Mile Career, FAANG Coach, and others: Big-tech interview-and-negotiation coaching boutiques.

The core value of negotiation tools: (a) interpreting market data, (b) quantifying the "cost of not negotiating," and (c) telling you the negotiation room (seller's vs buyer's market) per company.

Founder comp — Index Ventures and Kruze Consulting

Founder pay (CEO/CTO/CFO) follows different guidelines than regular employees. Reference materials:

  1. Index Ventures Founder Salary Tool: Stage-by-stage, country-by-country recommended ranges from seed through Series D. European and US data.
  2. Kruze Consulting CEO Comp Report: US seed through Series C CEO median and mean.
  3. First Round Capital State of Startup Pay: Stage-by-stage role-by-role comp.
  4. Pilot.com Salary Reports: Published by Pilot using its own customer data.

The key founder comp tradeoff: lower base in exchange for higher equity, but too low and you cannot live; too high and you burn runway.

Korean negotiation culture — the reality of yeon-bong-hyeop-sang

In Korean, "yeon-bong-hyeop-sang" usually refers to negotiation at (a) hire and (b) annual renewal.

At-hire negotiation reality:

  1. Big tech and foreign-capital: US-style signing bonuses and RSU negotiation are possible. Levels.fyi data is a direct bargaining chip.
  2. Korean IT majors (Naver, Kakao, Coupang): Base plus RSU (Coupang and others) negotiable. JobPlanet and Blind are the primary intel sources.
  3. Traditional Korean conglomerates: Strong seniority/grade systems narrow base negotiation rooms.
  4. Startups: Series-stage average plus equity negotiation. Rocketpunch and JobPlanet are the data sources.

Public-sector and public-institution pay is fully disclosed via ALIO (www.alio.go.kr), serving as a benchmark for private-sector negotiation.

Japanese negotiation culture — the shadows of taishoku-kin, shushin-koyo, and shinsotsu-ikkatsu-saiyo

Japan's compensation structure differs from both US and Korean.

  1. Shinsotsu-ikkatsu-saiyo: New-graduate batch hiring. April joiners start at uniform pay — negotiation room is near zero.
  2. Base salary + multiple allowances + biannual bonus: Base + role/commute/housing allowances + summer/winter bonuses. Base-only comparison misleads.
  3. Taishoku-kin (retirement allowance): Lump-sum payout at separation, tenure × company multiplier. Unlike US RSUs, it is paid only when leaving.
  4. Legacy of shushin-koyo (lifetime employment): The convention has weakened, but the assumption that "changing jobs lowers pay" still lingers.
  5. Mid-career hires (chuto-saiyo): Foreign-capital and IT firms have meaningful negotiation room. OpenWork and Ten-shoku Kaigi are primary sources.

The Ministry of Health, Labour and Welfare's Basic Wage Structure Survey is refreshed every Q4 with official pay distribution by industry, region, age, and gender.

Compensation comparison checklist — 13 items to evaluate every offer

A 13-item checklist when an offer arrives.

  1. Base salary: Distinguish gross from net.
  2. Signing bonus: Lump sum, often heavy first-year tax hit.
  3. Annual bonus (target): Percent of base, weighted by company + individual performance.
  4. Stock grant value: Total over four years, plus annualized.
  5. Vesting schedule: One-year cliff plus monthly/quarterly. Public companies often use 1/4/4/4 distribution.
  6. Vesting triggers: Single vs double trigger on acquisition or termination.
  7. ESPP benefit: Public companies may offer up to 15% discounted stock purchase.
  8. Pension/retirement: 401(k) match (US), National Pension (Korea), taishoku-kin (Japan).
  9. Health insurance: Employee contribution share (US); social insurance (Korea, Japan).
  10. PTO: Unlimited, accrual, sabbatical.
  11. Remote-work policy: Fully remote, hybrid, on-site.
  12. Childcare and tuition: US big tech often covers fertility, adoption, and dependent care.
  13. Post-termination option exercise window: 90 days is standard, some companies extend to 10 years.

Listing these 13 items per offer in one table avoids base-only comparison traps.

Five common mistakes in pay negotiation

  1. Comparing base only: Without TC, two identical bases may differ massively in value.
  2. Accepting the first offer immediately: Even at companies with room to move, first offers tend to land near the low end of market.
  3. Negotiating without a competing offer: Without a BATNA, leverage is near zero.
  4. Negotiating only over email: Calls or meetings close gaps much faster.
  5. Looking only at numbers: Great pay does not survive a bad manager or team. Always read Blind, JobPlanet, OpenWork reviews.

Transparency burdens on the employer side

For employers, pay disclosure is non-trivial.

  1. Fairness audit cost: Pay gap analysis and remediation by role and tenure.
  2. Rising dependence on external data: Pave, Carta, Mercer-type SaaS dependencies grow.
  3. Upward pressure on pay: Once employees see the same-role median, anyone below the 25th percentile will push for a raise.
  4. Legal exposure: The EU directive shifts the burden of proof to employers — adopting pay-equity analytics SaaS becomes effectively mandatory.

This burden is, in turn, the comp SaaS growth engine. As of May 2026, Pave is approaching nine-figure ARR; Carta Comp trails but is closing.

Outlook for the next 12 months — consolidation, AI, and the EU directive

Three changes are likely.

  1. Consolidation: Pave + OpenComp was the flare. Smaller comp SaaS will likely be absorbed by Pave or Carta.
  2. AI recommendations as default: Levels.fyi, Pave, and Carta will bake AI recommendations into base features — negotiation simulation, 12-month forecasts, automated remediation.
  3. EU directive effects: After June, global firms are expected to unify policies to the EU baseline, driving explosive growth in pay-equity SaaS.

Korea and Japan lack the same legal teeth, but JobPlanet and OpenWork are producing the same effect through crowdsourcing.

Closing — "what am I worth?" now has a data-backed answer

In May 2026, "what am I worth" is no longer decided by a manager's verdict. The Levels.fyi distribution, the JobPlanet median, the OpenWork quarterly distribution, and the disclosure dataset the EU directive will create — workers now hold more cards than ever.

Knowing the tools gives you leverage. Leverage forces employers to design pay policies honestly. The result is better wage efficiency across the market. May this map of tools, laws, and culture be useful in your next conversation.

References

Discuss on Twitter • View on GitHub