technical guide
How Vehicle Data Is Matched
Understand how vehicle records are matched and sourced.
Data architects, model-risk reviewers, and analysts responsible for reproducible vehicle-derived decisions.
why it matters
What this helps solve
Vehicle specifications are easy to copy but difficult to reconcile because similarly named trims can differ by market, year, engine, battery, or approval record.
coverage and freshness
Coverage for the current release
Share of export-cleared variants linked to at least one hard external identifier — a source-owned key that resolves to exactly one variant, so joins never rely on name or specification matching.
Share of export-cleared variants with a sourced combined CO2 emissions value in grams per kilometre.
Share of electric and plug-in hybrid variants with a sourced WLTP electric range value.
Maximum age 8 days. Oldest measured source refresh: 2026-09-15.
Apex is designed for weekly source-aware releases, but automated publication remains pending explicit production activation; each source keeps its measured fetch date and may follow a longer upstream publication cadence.
details
What to evaluate
Start with record grain
Treat a canonical row as a modeled vehicle variant, then verify the identity fields and matching quality needed for the target workflow before joining emissions or EV attributes.
Keep source context
Retain source URLs, field provenance, snapshot version, units, and caveats in downstream tables so reviewers can distinguish a normalized value from a regulatory assertion.
scope boundaries
Keep these limits visible
- Similar model names are not sufficient evidence of variant equivalence across jurisdictions or model years.
- Source definitions for peak power, system power, range, and emissions can differ even after unit normalization.
- Not a regulatory type-approval or emissions-compliance certification.
- Not a consumer valuation, repair, or vehicle-purchase recommendation service.