Compensation & hiring evidence · live panel

What employers state they pay. Measured, not surveyed.

Stop defending numbers you can't trace. PayEvidence keeps a point-in-time record of employer-published pay ranges and hiring activity, so compensation, strategy and data teams can put a public source behind every figure they present.

See the evidence — free Get my coverage review
$100k$200k$300kmedian $118k
Every stroke is one employer-stated salary range in the live sample (n=36,766, annual USD, reference FX as of 2026-10-09). Drawn from the same rows the explorer shows.
1,563 companies 132,760 open roles 37,493 employer-stated ranges 67 countries Updated under an hour ago 100% employer-stated data · every row source-linked
One measured signal

52.8% of open roles state pay Measured across 132,760 open roles at 1563 companies. EU roles in this sample disclose at 44.9% (n=31311) — the gap the EU Pay Transparency Directive is designed to close. This is source-specific evidence, not a labor-market census.

Inspect the underlying rows Open the full dashboard EU transparency league →
Built for decisions

From job rows to usable signals

One evidence layer, three workflows. Every aggregate can be traced back to public source rows.

01

Investment & strategy

Track hiring direction, function mix and compensation posture before slower corporate disclosures.

See hiring movers →
02

Compensation teams

Benchmark employer-stated ranges by company, level, function and market — without survey lag.

Compare stated engineering pay →
03

Data & AI teams

Query normalized, point-in-time rows through a stable API with source evidence and explicit limitations.

Read the API contract →
Salaries

What employers state, by market

Live benchmarks built only from employer-published ranges — each with its sample size in the open.

Engineering · United States n=6036 Sales · United States n=3452 Data & ML · United States n=2684 Operations · United States n=2413 Product · United States n=1143 Finance · United States n=806 Marketing · United States n=771 Consulting · United States n=765 All benchmarks →

What this panel is

  • Employer-published job facts, observed directly on public ATS boards and snapshotted append-only.
  • Stated pay only: a salary enters the panel when the employer wrote one, never modeled or imputed.
  • Source-linked: every row keeps its public URL, observation time and extraction confidence.

What it is not

  • Not a labor-market census — the sample skews to technology-sector employers; the live sector mix is published per company.
  • Point-in-time by construction: every series grows from its first observation and is never backfilled — the exact window of each series is stated on the methodology page.
  • Not survey data — no self-reported salaries, no scraped personal profiles, no private data of any kind.

Coverage, field completeness and known biases are published per company. Methodology →

Why it compounds

History that cannot be bought back

Anyone can copy today's job boards. Nobody can reconstruct the day-by-day record of what was open, at what stated pay — we measured only ~0.09% of it to be recoverable from archives. Every collection cycle widens that gap.

Methodology API
Every cycle writes history that can never be reconstructed later — archive evidence reaches back to 2022, and the live full-panel series compounds daily.

Sparse archive captures are kept as isolated evidence, never blended into the live series. Synthetic history is never mixed into production.