1. Data ingestion
Job listing and salary observation data is imported from public job boards and aggregated salary sources. Each observation is tagged with role, seniority, location, contract type, and currency.
Methodology
How salary benchmarks are built and published.
PayBench aggregates publicly available salary and job listing data, normalizes it across sources and markets, and computes benchmark snapshots that feed structured SEO pages.
2. Normalization
Observations are normalized to annual gross salary equivalents in local currency. Outliers are clipped to reduce noise. Seniority and contract-type adjustments are applied using a consistent multiplier model.
3. Rollup computation
Normalized observations are grouped by job family, country, and city. For each group, the system computes the 25th percentile (low), median, and 75th percentile (high) of the annual salary distribution.
4. Confidence scoring
Each rollup receives a confidence score based on sample size, source diversity, and data freshness. Routes with low confidence or stale data are marked noindex and excluded from sitemap generation.
5. Publishing thresholds
Salary routes require at least 10 observations within 90 days. Freelance routes require at least 3 observations. Job market routes require at least 3 active listings within 45 days. Routes that drop below these thresholds are automatically suppressed.
Methodology notes
Observed vs modeled policy
PayBench distinguishes between observed public-market inputs and modeled outputs derived from those inputs. When a route depends on modeled backing, the route should say so visibly and keep the supporting methodology link close to the benchmark evidence.
Source lineage and route facts
Each published route is backed by a rollup snapshot with freshness, sample size, source count, and public route facts. The machine-readable endpoint is part of the public verification layer, not a hidden internal-only API.
Example reference: salary/software-engineering/germany.
Review cadence
Methodology, trust-page copy, and publishing thresholds are reviewed when source mix changes, when route-facts fields change, or when freshness and indexability policies are tightened. Current documented review baseline: June 2026.
Currencies and regions
Germany, Austria, and Switzerland use EUR. Netherlands uses EUR. Sweden uses SEK. All rates are expressed as annual gross equivalents unless explicitly noted as day rates (freelance benchmarks).
Limitations
Benchmarks reflect aggregated market data, not individual compensation outcomes. They should be used as directional anchors, not guarantees. Sample sizes are shown on every page so users can calibrate their confidence appropriately.