Construction Equipment Auction Results Database Data: Dataset Structure and Field Coverage

An auction results database answers a question dealer listings cannot: what did a specific class of used machine actually sell for, on a dated event, at one row per lot. In this industry that record exists in exactly one place Datadory catalogs - Ritchie Bros. Auctioneers - auction results for heavy equipment, the world's largest heavy equipment auctioneer, operating as RB Global across its Ritchie Bros. and IronPlanet marketplaces.

The grain matters as much as the name. Coverage granularity is one record per equipment lot per sale event, refreshed daily, spanning construction machinery, trucks and mining equipment through rbauction.com and regional IronPlanet domains. At verification, IronPlanet alone listed 39,911 construction items, and roughly 1,400 sitemap URLs per chunk refresh each day - an enumeration surface no official statistics series offers.

Set it against the rest of the 22 primary datasets Datadory catalogs for Construction Machinery & Heavy Transportation Equipment: the US Census Annual Survey of Manufactures (NAICS 333120) measures factory shipments, not second-owner transactions, and Eurostat PRODCOM DS-066341 counts EU production annually since 1995. Neither carries a price for an individual used machine. Only the auction channel does.

Each lot resolves into the attributes a valuation model needs: equipment identity (category such as excavators, dozers, cranes or mining equipment, plus trucks), condition, sale-event date and location, and the realized outcome for that lot. Because coverage is one row per lot per event, a collector can rebuild category-level price histories by aggregating across events rather than waiting for any published index.

Three differences decide which source a workflow needs. First, semantics: a listing is an unsold machine with a wishful number - many show Call instead of a figure - while an auction lot closes at an observed transaction. Second, retention: both marketplaces drop units as dealers remove them and hold no dated archive, whereas auction events are dated by design. Third, geography: the auction record is global out of the box, while the marketplaces center on the United States with localized mirrors for Canada, the UK, Australia and Europe.

A working pipeline runs in six steps, ordered so every step names a real source:

Run steps 2 and 3 continuously and steps 4-6 annually or weekly as noted. The blend produces category-and-region residual-value curves no single source provides alone.