Duty & Landed-Cost Intelligence
Model landed cost and sourcing alternatives without a trade attorney. Keep a live readiness inventory of every entry that might be owed a refund. Assess each incoming CBP policy change against your actual product and HTS portfolio — instantly.
The problem
"We're deciding China vs. Vietnam for a part and the duty math is a guess until the first entry actually clears."
Sourcing decisions
"We've almost certainly overpaid duty somewhere. We have no systematic way to find it. Drawback is 'someone should look into that.'"
Unclaimed refunds
"A tariff change hits the Federal Register and we find out when our broker mentions it — if they do."
Regulatory blind spots
How Qubere addresses it
Sourcing Scenarios
Build a scenario by origin country, HTS code, and cost components (unit cost, quantity, freight, insurance). The breakeven and duty differential — the number that used to be a consulting deliverable — computed on the spot.
Customs valuation depth
The full 6-rung 1401a method ladder — transaction value (with first-sale substitution), identical/similar merchandise, deductive, computed — each rung marked accepted / rejected / not-reached with an expandable rationale. A first-sale arrangement registry runs the Nissho Iwai three-prong test (bona fide sale, destined for export, arm's length) against a registered manufacturer→middleman→importer chain, and the 19 CFR 152.103(f) / T.D. 91-7 royalty & proceeds test is walked as a decision tree. None of the three first-sale prongs defaults to "Pass."
| Method (1401a) | Result |
|---|---|
| 1 · Transaction value | rejected |
| 1 · First-sale substitution | under review |
| 2 · Identical merchandise | not reached |
| 4 · Deductive value | accepted |
Duty recovery · opportunity scan
One scan checks each line for: classification review, Section 301 exclusion, trade-agreement claim, first sale, drawback, AD/CVD scope exclusion. estimatedRecovery stays null until an opportunity is actually confirmed — a blank is honest; a made-up number is a liability.
| Opportunity | Entry | Type | Recovery | Deadline |
|---|---|---|---|---|
| First sale on 3 lines | QBR-…3182 | FIRST_SALE | $18,240 confirmed | 141 d |
| USMCA not claimed | QBR-…3204 | TRADE_AGREEMENT | $9,100 confirmed | — |
| Lower-duty HTS alt (GRI) | QBR-…3251 | CLASSIFICATION_REVIEW | analyzing | — |
Drawback matching & Section 301 readiness
Imported lots (created when a filing is accepted, from lines with non-zero duty paid) matched FIFO within HTS code to exports, with over-allocation prevention. Manufacturing = 99% of base + Section 301 duty. The Section 301 inventory shows total duty paid per list, entry counts — and grants auto-create refund opportunities.
| List | Entries | Duty paid | Top HTS |
|---|---|---|---|
| List 3 | 194 | $412,900 | 8541.43, 8537.10 |
| List 4A | 71 | $88,410 | 7616.99 |
| List 1 | 23 | $21,140 | 8504.40 |
PSC & protests
Post-Summary Correction eligibility (entry accepted, before liquidation, material duty impact, no active drawback claim) + an original-vs-corrected duty-delta calc. Full CBP 19 protest workflow — grounds, filing-window tracking, attached entries. Entry-level reconciliation with a "Convert to PSC" action.
Regulatory-change impact
A live feed pulled from the Federal Register API (real fetcher + AI extraction), auto-creating refund-opportunity records where relevant. Assess Impact runs an update against your product and HTS portfolio — the analysis that used to take a trade attorney a week.
What's real, and where the data stops
Unknowns show as "unknown," never a computed guess. That honesty is the trust argument.
Objection handling
| "Do you file the refund with CBP?" | Qubere identifies opportunities, matches drawback lots, and produces the preparation package (claim / PSC / protest). The ACE/CBP submission goes through the filing path (live with ABI credentials) or the broker's system. It's the analysis and the package, done systematically — not an automatic wire. |
| "How real is the recovery detection?" | Models and workflows are built with real Decimal math and over-allocation guards. Two honest data gaps: Section 301 exclusion-text matching and precise PSC deadlines — both show "unknown" rather than a guess. |
| "Is valuation more than just 'transaction value × HTS rate'?" | Yes — the method ladder, first-sale registry, and royalty/proceeds tests are real and wired into Reasonable Care and the AI copilot. None of the three first-sale prongs defaults to "Pass," so an unclear entry surfaces as Needs Review, not a false clean bill. |
| "Is the regulatory feed comprehensive?" | Federal Register CBP notices, live. USITC trade-remedy orders, §232/§301 rate annexes, PGA requirements are on the data roadmap — and the platform is explicit about which datasets are LIVE vs not. |
The close
Next step: give us one sourcing question you're currently sitting on. We build the scenario live — and run the recovery scan against a sample of your filed entries.
Full detail: docs/apps/customs/sales/duty-and-landed-cost.md