Start with transport: return label cost at the contracted rate, not the headline rate, plus the consolidation or linehaul cost if returns route through a hub. Then handling at both ends, the carrier drop, the returns centre receipt, inspection and grading time. Then the duty question, which deserves its own line because the answer varies: duty paid on import may be reclaimable under returned goods or drawback style mechanisms in some circumstances and plainly lost in others, and the difference is paperwork discipline decided at the original entry, long before any customer changes their mind.
Finish with value recovery: the realistic resale price of an opened return in its actual grade, not the new price the spreadsheet prefers, and the write off rate for items that come back unsellable. A reverse cost model that assumes full price resale is a forward model wearing a false moustache. Grade honestly for a quarter and use those rates.
Below a calculable value, returning an item costs more than the item can recover, and the rational instruction is let the customer keep, donate, or dispose of it, with the refund issued on evidence. Brands resist this because it feels like training customers to claim, and managed badly it can. Managed with evidence requirements, category limits, and per customer frequency checks, it is cheaper than freight and kinder than demanding a parcel cross an ocean to be thrown away in a different country.
Publish the threshold internally as a number derived from your stack, reviewed when freight rates move. A keep it decision made per ticket by a support agent under pressure will drift generous. A threshold in the system, with the agents discretion bounded, holds its shape and can be audited against the return fraud it supposedly invites.
Whether duty on a returned import can be recovered depends on the mechanism available, the proof that the returned goods are the goods that were entered, and the state they return in. Serial numbers, order to entry linking, and unaltered condition are the evidence, and they exist only if the forward process captured them. Returns teams discover this backwards: the reclaim fails not on the rule but on the file, because nobody connected the entry number to the order when the goods arrived.
Link entry references to orders as standard, and keep the link through the returns system. It costs a data field. It buys every future reclaim, audit defence, and duty analysis that needs to know which shipment brought a given unit in, which turns out to be most of them once anyone starts asking questions with money attached.
A category with high return rates and a heavy reverse stack is a different business from its forward margin, and the buying, pricing, and market decisions should see it. The product that margins well forward and bleeds on returns might need a price that reflects its true cost, a market it should not be shipped to, or a sizing and description fix that cuts the returns themselves. Reverse cost is not an accounting refinement. It is the number that tells you which cross border sales were worth making.
Report it beside forward landed cost, by category, on the same cycle. The first quarter will be argued about, because the stack contains estimates. The second quarter will be used, because even an approximate reverse cost beats the current figure, which in most businesses is a confident zero.
Keep reading: How should sample and prototype shipments be costed in landed cost models?, How do currency swings change landed cost between order and arrival?.
Return freight, the outbound freight already spent, duty net of any recovery actually achievable, handling and inspection at both ends, repacking, and the resale discount or write off on the returned unit. Each line should use contracted rates and observed grading rates, not list prices and hopes.
Sometimes, through returned goods relief or drawback style mechanisms, depending on the jurisdiction, the goods condition, and the evidence linking the return to the original entry. It is never automatic, and it is usually lost where the entry to order link was not kept.
When the reverse stack for that item exceeds its realistic recovery value. Set the threshold from your own cost stack, apply it in the system with evidence and frequency checks, and review it when freight rates change.
Because duty recovery, audits, and cost analysis all need to know which entry brought a specific unit in. Capturing that link on arrival costs a data field. Reconstructing it during a reclaim is usually impossible.