NHTSA publishes new campaigns on Mondays and the API reflects them within hours. A 7-day SleekRank cache means the corpus tracks NHTSA within a week, which matches how the federal portal updates. For breaking high-profile campaigns, drop the cache on the specific page group to 24 hours.

The format is YY V/E/T NNN-XXX where YY is the year, V is vehicle (E for equipment, T for tires), NNN is the campaign sequence within the year, and XXX is typically 000. SleekRank uses the lowercased dash-formatted version as the slug, e.g. 23v-456-000 for 23V-456-000.

Yes. NHTSA hosts the chronology, defect notice, and remedy instructions as PDFs at predictable URLs per campaign. Link them from the per-campaign page via tag mappings on the relevant URL fields in the API response. Hosting copies on the operator's domain is also fine since the documents are public records.

Transport Canada Defect Investigations and European Rapid Alert System (Safety Gate) have their own feeds. Run parallel page groups at /transport-canada-recall/{slug}/ and /eu-safety-gate-recall/{slug}/. The three corpora can be cross-linked when the same vehicle is recalled in multiple jurisdictions.

Mega campaigns affecting millions of vehicles are still single campaigns in NHTSA's record but reference many makes and models in the affected_vehicles list. SleekRank renders the page with the full affected-vehicles table; the make/model clusters surface the campaign on every relevant manufacturer and model page automatically.

Yes. NHTSA provides a VIN decoder API. Add a VIN lookup form on the per-campaign page that calls the NHTSA endpoint and confirms whether a given VIN falls within the recall scope. The lookup runs client-side against NHTSA's API; SleekRank just renders the form and ID context.

Safercar.gov is the consumer face of the NHTSA recall data, but the underlying API is publicly documented and explicitly intended for third-party use. SleekRank reads the API directly, so the corpus is upstream of any consumer portal and not dependent on screen scraping that breaks when NHTSA's site changes.

Each campaign record is 3-8 KB in the items table depending on the size of the affected_vehicles array. Thirty thousand campaigns is well under 1 GB of MySQL footprint, fine on any production database. Even with the make/model/component parallel page groups, the total corpus stays small.