Unused merchandise drawback covers imported goods exported in the same condition, the classic case for a distributor whose US warehouse serves Canadian or overseas customers. Manufacturing drawback covers imported inputs made into an exported article. Rejected merchandise drawback covers imported goods that did not conform and are exported or destroyed under the rules. Each has its own matching method, direct identification or substitution where the law allows it, and its own clock: claims generally run up to five years from import, which means a backlog of eligible entries is a recoverable asset, not ancient history.
The matching is the work. A claim ties the export to the import and proves the duty was paid, through entry records, inventory accounting, and export documentation that agree with each other. Companies fail drawback on records, not on eligibility. If your inventory system cannot say which import lot fed which export shipment, build that trail first, on a substitution basis where available and properly approved, before filing anything.
A landed cost model that expenses all duty on entry overstates the true cost of every unit that later exports. Sophisticated operators book an estimated drawback recovery against the duty line for the export share, trued up as claims pay. The effect is not cosmetic: a product that looks unviable for international sales at full duty can clear its margin hurdle at one percent of the duty, and a pricing team working from the unadjusted cost will kill that channel on wrong numbers.
Do not net the estimate to zero risk. Claims can be denied in part, substitution rules have conditions, and accelerated payment is a privilege with a bond, not a default. Book a conservative recovery rate built from your own first claims, and let the rate rise on evidence.
Drawback has its own applications: rulings for manufacturing drawback, waivers and approvals for the accounting methods, and for many claimants, a broker or drawback specialist who files on software built for it. The specialist fee, often a share of recovery, is worth modelling against an in house build unless your volume is large and steady. Either way, somebody must own the import to export data chain, and that ownership belongs in the job description, not in a year end scramble.
Destruction claims deserve a process note: destruction must happen under customs supervision or an approved alternative, with notice given in advance. The warehouse that quietly landfills unsold imports has destroyed the goods and the claim in the same act. Put the drawback check into the disposal workflow, before the skip is booked.
Pull two years of entries, mark the export shipments from your order system, and estimate the overlap value with your broker. That one exercise sizes the prize and exposes the record gaps while they are still cheap to fix. Confirm current drawback rules and claim procedures with your broker or a drawback specialist, because the regulations are detailed and the filing requirements are unforgiving of improvisation.
Generally up to 99 percent of the duties, and certain fees, paid on the imported merchandise covered by an approved claim. The one percent retention and the claim costs are why the recovery belongs in the model as an estimate, not as a full reversal.
Claims are generally filed within five years of the import date, which lets a new programme recover on several years of past entries if the records support the matching. That window is closing continuously, so the oldest eligible entries should be worked first.
Size it before deciding. A small export share of a high duty product can still clear the setup cost, especially with back years claimable. A low duty, low export business may not. The two year overlap estimate answers it in your numbers, not in general.
Yes, under the unused or rejected merchandise provisions where the destruction is done under customs supervision or an approved procedure with advance notice. Unrecorded disposal does not qualify, so the process has to start before the goods are destroyed.