
The model below has seven inputs. Three ofthem are usually estimated, one is usually omitted, and the omission is the onethat most often changes the answer.
Landed cost is the number that tells youwhether a product is profitable. It is also the number most importers computeonce, badly, and then never revisit.
Component | Basis | Common error |
Goods cost | Factory price at the agreed Incoterm | Comparing FOB to CIF prices as though they were the same |
Freight | Ocean, and any air or rail leg | Using a rate rather than an annual average |
Insurance | Cargo value plus a percentage, at a stated rate | Omitting it because it is small |
Duty | Classified value multiplied by the tariff rate | Using the wrong classification or the wrong valuation basis |
Federal fees | Merchandise processing and harbor maintenance | Assuming the ad valorem rate applies without the caps |
Destination and inland | Terminal, drayage, chassis, delivery | Treating it as a one-off rather than a per-unit cost |
Carrying cost | Inventory days multiplied by your cost of capital | Ignoring it entirely |
The seventh is the one that changesdecisions. Two products with identical freight and duty can have very differentlanded costs once the inventory they require is priced in.
The carrying cost is the omitted one. It isthe value of the inventory multiplied by the cost of capital, multiplied by thefraction of a year the goods are held. On a product with a 25 percent marginand a 60-day pipeline, that is a material number, and it is never on a freightinvoice.
FOB, CIF and DDP prices are not comparable.A DDP price includes duty and destination costs, so it looks higher and is not.Before any landed cost comparison, convert every quotation to the same term.
The conversion requires knowing thefreight, the insurance, the duty rate and the destination charges — which isexactly the data you need for the landed cost anyway. Doing it once gives youboth.
The conversion in the other direction isalso useful. If you buy DDP and want to know whether the supplier's price isreasonable, subtract the duty, the destination charges and a plausible freightfigure, and see what remains as the goods cost. A residual that is below themarket price for the product is a signal that something in the DDP price is notwhat it appears.
US customs value for most goods istransaction value: the price actually paid or payable, adjusted for assists,selling commissions and certain royalties, and excluding international freightand insurance. That definition is specific, and using a different one producesa duty figure that will not survive an audit.
The classification matters equally. Dutyrates vary widely by tariff line, and a plausible-looking classification thatis wrong can produce a rate error in either direction. Both directions createliability.
Two valuation adjustments catch importersregularly. Assists — tooling, moulds, design work or materials that the buyerprovides free or at reduced cost to the supplier — must be added to the pricepaid. So must selling commissions and, in some cases, royalties. Omitting themproduces an understatement that is a violation even when the omission isaccidental.
1. Take one shipment. Record everyinvoice line against the seven components above.
2. Divide the total by the numberof units in the shipment. That is the landed cost per unit at that volume.
3. Repeat for three shipmentsacross different seasons, because freight and destination costs move.
4. Use the highest of the threefor pricing and the average for planning. Pricing on the average means sellingat a loss half the time.
The fifth step is to separate the variablefrom the fixed. Duty and goods cost scale with units; destination charges andentry fees do not. That split is what allows a per-unit landed cost to beprojected to a different volume, which is the question every pricing decisionactually asks.
The model should also carry a date. Landedcost calculated on a rate from eighteen months ago is not a cost model, it is ahistorical record. Rebuilding it annually, and re-running it whenever a tariffrate or a gateway changes, takes an hour and prevents a year of pricing onstale inputs.
Once carrying cost is included, a slowerand cheaper routing often turns out to be more expensive than a faster one,because the extra inventory required outweighs the freight saving. Thatconclusion is invisible in a freight-only comparison and obvious in a landedcost model.
The model does not need to be precise to beuseful. It needs to be consistent, so that comparisons between routings,suppliers and products are made on the same basis.
The model also exposes the categories wherethe freight decision is irrelevant. For a product with a 3 percent duty rateand a low value-to-weight ratio, freight dominates and routing matters. For onewith a 25 percent duty rate and a high value density, the classificationmatters far more than any rate negotiation.
Knowing which category each product sits intells you where to spend the next hour.