CRI Ireland is designed around a simple methodological principle: keep imported company-register fields separate from editorial interpretation, preserve stable company identities, clean unusable values before publication and explain how navigation layers are derived.
Company identity
The company number is treated as the stable record key. Updated imports use that key to update an existing company record rather than generating a duplicate profile because a name, address or status has changed.
Source normalisation
Imports trim whitespace, standardise usable values and remove known placeholders such as no-address markers, repeated asterisks, null-like strings and malformed location data. Unreliable location strings are not used to create public town pages or breadcrumbs.
Directory geography
County and town fields are used only where a usable location can be derived from the imported source. A registered office is a register field and should not automatically be interpreted as the only place in which a company operates.
Sector mapping
Broad business sectors are mapped from NACE prefixes to help users navigate the directory. A sector label is therefore a discovery layer, not a substitute for the company’s official filings or a professional classification opinion.
Updates
Large CSV datasets are processed in batches. The importer tracks created, updated, skipped and failed rows. Aggregate directory counts and cached sector data are refreshed as imports complete.
Editorial separation
Business guides, grants and support pages are WordPress editorial content, while company records remain in the custom directory database. This separation allows company data to scale without turning hundreds of thousands of companies into WordPress posts.
Import pipeline
The company importer is designed for large datasets. Files are validated, queued and processed in batches so the site does not need to load hundreds of thousands of rows into memory in a single web request. Each row is normalised and matched against the existing directory by company number. A new number creates a record; an existing number updates the existing record. Import logs make it possible to distinguish created, updated, skipped and failed rows.
Stable URLs
Company profile URLs are intended to remain stable even when a company name changes. The company number provides the strongest link between refreshed records and an existing profile. Where a slug changes for usability, the directory should preserve or redirect older routes where practical rather than fragmenting the same company across multiple pages.
Search and filters
Directory filters query indexed fields in the custom company table. This keeps the company dataset separate from ordinary WordPress posts and allows paginated searches across a very large number of records. Important directory dimensions such as county, town, sector, type and status can have dedicated landing pages, while arbitrary combinations of multiple filters are not intended to create unlimited thin SEO pages.
Derived content
CRI may derive a short plain-language company summary from structured fields. The summary can state the company name, CRO number, legal type, imported status, sector mapping and available location, but it should not invent products, clients, employee numbers, revenue, directors or services that are not present in the source data.
Statistics
Statistical pages use supplied aggregate datasets and can be regenerated when replacement data is installed. Visualisations are designed for exploration, with clear source notes and supporting tables where practical. CRI does not silently merge unrelated datasets or label estimates as official totals.
Quality review
Data quality is an ongoing process. CRI uses conservative cleaning rules, correction reports and refreshed imports to improve the directory over time. The methodology page is updated when a material new data source, mapping rule or enrichment process is introduced.