Methodology

How we built the dataset.

The UK Construction Salary Benchmark is a continuously-updated, crowdsourced dataset of white-collar pay across the United Kingdom. Submissions come from individuals (employees) and hiring teams (employers).

Currency & geography

All salaries are captured and reported in GBP. Day rates are flagged separately and never mixed with permanent salaries. The tool is scoped to the UK only.

Location: UK counties

Every submission is bucketed into one of ~92 UK locations: the 48 ceremonial counties of England, 32 Scottish council areas, 22 Welsh principal areas, and 6 Northern Irish counties. For user submissions we always capture the county where the job is based — not where the contributor lives.

Job-board rows (Reed) are resolved to a county via a four-step fallback: (1) exact match on a county name or common alias, (2) lookup against a curated ~250-entry UK town/city → county map, (3) postcode outward-area → county (e.g. M3 → Greater Manchester, EH7 → City of Edinburgh), (4) free-text scan over the full location string. Rows we can't resolve with confidence are kept in the raw tables for later re-parsing but are excluded from the benchmarks.

Data sources

Benchmarks are calculated across three independent source tables, unified at query time: (1) direct user submissions from employees and employers, (2) curated research data compiled from public recruiter reports and agency surveys, and (3) live Reed UK job postings. Each source is stored separately and never overwritten, so the mix behind any figure is auditable.

Privacy floor

The dashboard refuses to display any percentile for a segment with fewer than two data points. The county heatmap uses a single-point floor so the map stays useful at the edges. Individual rows are never returned to the browser — only server-side percentile aggregates.