Stable company identity
The eight-character Companies House number is the canonical identifier and URL. Company names may change and therefore never control the canonical path.
M Methodology
Facts, matches and gaps are different things
BritDB is designed so the interface cannot silently turn missing coverage into a clean bill of health.
Atomic source imports
Official files load into staging tables, receive indexes and record checks, then merge into production only after parsing completes. Users never browse a half-imported batch.
Evidence-first matching
Company numbers and regulator-supplied identifiers are definitive. Name and postcode combinations create candidate matches only. Fuzzy name-only matching is never displayed as fact.
Deliberate search indexing
Only canonical active-company pages enter the initial XML sitemaps. Search results, experimental designs, query variants and later pagination do not compete with company profiles.
Corrections at the right layer
If the underlying public record is wrong, the original registrar must correct it. BritDB can correct its import, matching, labels or presentation and always provides the source link.
Privacy-preserving visit counts
BritDB counts human page views and daily unique hosts without cookies. IP addresses are converted to a one-way keyed hash and are never stored in raw form; known crawlers and monitoring tools are excluded.
Financial facts stay source-limited
BritDB parses supported facts from electronically filed iXBRL/XBRL accounts, preserves the reporting period and source batch, and keeps missing values distinct from zero. Figures are company-filed and are not audited by BritDB.