Data & Valuation Methodology

Listings are indexed from publicly accessible government auction and agency sources, then normalized into a single schema (category, asset type, year, make, model, mileage or hours, condition, location, bid, closing time). Every record links back to its source listing, which is always authoritative.

Value estimates are built from indexed sold comparables for the same asset type, adjusted for year, usage (mileage or hours) and condition. When too few comparables exist, we fall back to category-level modeling and report the confidence as Low. A value range accompanies every midpoint estimate.

Confidence is reported as High, Medium or Low based on the number and closeness of comparables. Low confidence means the range is wide and should be treated as indicative only.

We add estimated buyer fees and estimated transport to the current bid to produce an all-in cost. The modeled spread is estimated market value minus all-in cost; modeled savings and discount percentage are derived from the same figures. Fees and transport are estimates and vary by auction house, location and asset size.

The score is deterministic and weighted as follows:

• Market discount — 30%
• Comparable support — 20%
• Modeled spread size — 20%
• Demand signal for the asset type — 10%
• Competition level (bid activity) — 10%
• Data confidence — 10%

The same inputs always produce the same score. A high score means the modeled economics look favorable — not that an asset is sound, roadworthy or resellable.

Condition is often self-reported or absent. Mechanical faults, missing keys, title status and salvage history are frequently undisclosed. Photos may be stale. Where a field is missing, we report it as unavailable rather than filling it with a guess.

Any figure labeled Surplus Asset Intel Network Data describes only records indexed by this platform. It is not a measure of the entire government surplus market.