Drawback money is lost in spreadsheets. Matching imports to exports at SKU level, with the five year clock tracked, turns eligibility into a system.
Duty drawback refunds up to 99 percent of duties on imported goods that are later exported or destroyed, and it is chronically underclaimed. The reason is rarely eligibility. It is bookkeeping. The import happened two years ago in one system, the export left last month from another, and nobody can prove the exported units are the imported ones to the standard CBP requires.
Multiply that by thousands of SKUs, multiple suppliers feeding the same product, and staff turnover, and the refund becomes archaeology. Successful drawback programs are not better at paperwork at claim time. They capture the linkage as a byproduct of normal operations, when the facts are fresh.
Every import line needs five things stored where exports can find them: the entry number and line, the SKU and its classification, the quantity and duty actually paid, the supplier lot or batch identifier, and the date the five year claim clock started. Duty paid is the figure people most often estimate later and regret; capture the liquidated reality, not the estimate.
Substitution drawback loosens the physical matching for many goods, letting commercially interchangeable products qualify, but it raises the classification and documentation bar rather than lowering it. Either route, unused merchandise, manufacturing, or substitution, lives or dies on whether your records connect import lines to export events without a human rebuilding the chain from emails.
When an export ships, the system should consume eligible import inventory under a stated rule, usually first in first out within the SKU and classification, and record exactly which import lines it drew from. That running ledger is the claim before it is a claim: quantities matched, duties attached, deadlines visible.
Exceptions deserve their own queue. Exports missing commercial invoices, imports with classification changes mid stream, and goods re-exported after repair or warranty work all break naive matching. A small weekly exception review keeps the queue from becoming the next archaeology project, and it surfaces process errors while the people who made them still remember why.
Drawback claims must be filed within five years of the import date, and eligibility dies quietly line by line as that window closes. A useful dashboard shows recoverable duty by expiry quarter, so the business can prioritize: accelerate claims for aging lines, and where sensible, prioritize exporting older eligible stock before its window shuts.
Treat recovered drawback as a measured program, not found money. Track match rate, claim acceptance rate, and dollars expiring unclaimed. That last number is the honest scoreboard. Driving it toward zero is worth more than any single large claim, because it proves the system, not a heroic one off effort, is doing the work.
Claims must be filed within five years of the import date of the goods. After that window closes, the duties on those import lines can no longer be recovered.
Entry documentation showing duties paid, proof of export, and records linking the exported goods to the imported ones, or to commercially interchangeable goods under substitution rules.
No. Substitution allows commercially interchangeable merchandise to qualify, but the classification, timing, and documentation requirements are strict and must be tracked just as carefully.
Usually record linkage, not eligibility. Imports and exports live in different systems, and without SKU level matching captured as operations happen, proving the chain years later becomes impractical.