Your supplier invoice says a unit costs $12. Your ERP agrees. Your dashboard shows a healthy 58% gross margin. All three are wrong. A real landed cost calculation adds duties, brokerage, insurance, and the cost of holding the goods — and when you run it, that 58% often drops to 51%. The gap between the two numbers is where your profit hides, and most systems never show it to you.
That understatement runs 3–12% of COGS for the typical importer. On a $2M cost of goods base, that is $60K to $240K of "profit" that was never there. Here is how the gap opens, and how to close it.
Invoice price plus a freight guess is not your cost
Most ERPs compute margin the fast way: supplier invoice price, plus a flat freight estimate, and stop. That works if freight is trivial and everything else is zero. For anyone importing physical goods, neither is true.
True landed cost has five layers, not two:
- Product cost — the supplier invoice.
- Freight — ocean, air, or ground, plus fuel surcharges.
- Duties and tariffs — assessed at the border by HS code.
- Brokerage and insurance — customs clearance, cargo insurance, port fees.
- Carrying cost — capital tied up while inventory sits.
Skip layers 3 through 5 and your recorded cost is fiction. As Shopify's guide to inventory valuation puts it, the value you assign to inventory drives every margin number downstream. Start that number too low and every report inherits the error.
The timing problem: freight arrives weeks after the goods
The harder issue is not which costs to include — it is when they show up. A shipment clears customs in March. Duties post in March. But the freight invoice from your forwarder arrives in April, and the brokerage bill in May.
So your ERP records the goods at invoice-plus-estimate in March, ships and sells half of them, then receives the real freight cost after the margin is already booked. The recorded product cost no longer matches reality, and there is no clean way to walk it back.
This is a real-time data problem dressed up as an accounting problem. The IRS guidance on accounting periods and methods requires that inventoriable costs attach to the goods — but it does not make the timing convenient. You either estimate accurately up front and true up later, or you carry a rolling error.
The fix is disciplined: book a landed cost accrual at receipt using known rates, then reconcile against actual invoices within the same close. Track freight-in as part of COGS, not as a floating operating expense that disappears from your unit economics.
Duties allocate unevenly, and that distorts SKU margin
Here is the second timing trap. When a container holds twelve SKUs, duties and freight get allocated across them — usually by value or by weight. Choose the wrong basis and you overstate margin on some SKUs and understate it on others.
A heavy, low-value SKU allocated by value looks more profitable than it is. A light, high-value SKU allocated by weight carries costs that belong elsewhere. Both errors survive into your pricing decisions.
Allocate by the driver that actually generated the cost. Duties follow declared value by HS code. Ocean freight follows volume or weight. Insurance follows value. Mixing them into one pooled percentage is the shortcut that produces the 3–12% distortion at the SKU level, even when your blended number looks fine.
Getting this right is the whole point of proper COGS tracking — your blended margin can be accurate while your per-SKU margins are quietly upside down.
Carrying cost is the layer everyone forgets
Freight and duties are visible bills. Carrying cost is not, which is why it gets left out — and why it matters most for importers with long lead times.
Money spent on a container that sits in transit for six weeks, then in your warehouse for another eight, is money not available for anything else. That has a real cost: your cost of capital, plus storage, insurance, shrinkage, and obsolescence risk. For slow-moving inventory it can add several points to true landed cost on its own.
You do not need to capitalize carrying cost into inventory value the way you do freight and duties. But you do need to see it when you price and reorder. BigCommerce's ecommerce accounting primer is blunt about the risk of treating inventory as a static number — the cash it consumes is dynamic, and it compounds.
Landed cost changes how you value inventory
Your valuation method decides how landed cost flows into COGS as you sell. Under FIFO, the oldest costs — including the freight and duty rates from an earlier shipment — hit COGS first. That matters when rates move, and freight rates have moved violently in recent years.
NerdWallet's breakdown of FIFO vs. LIFO walks through the mechanics. The point for landed cost is this: if your per-unit cost only captures invoice price, your FIFO layers are built on the wrong foundation and the method choice cannot save you.
Whatever method you run, the definition of cost of goods sold stays the same — the full cost of getting sellable product in the door. Landed cost is simply COGS done honestly. For the wider picture on how these numbers roll up, see our guide to ecommerce margins decoded and the deeper work on true margins.
What to do this quarter
You do not need to rebuild your ERP. You need to see the gap and act on it. Three steps:
- Run one real landed cost calculation on your top SKU. Add every layer — invoice, freight, duties, brokerage, insurance, carrying cost. Compare it to what your ERP shows. The difference is your baseline error.
- Fix your allocation basis. Move from a single pooled percentage to per-driver allocation: value for duties, weight or volume for freight. Do it for your top ten SKUs first.
- Accrue at receipt, reconcile at close. Book estimated landed cost when goods arrive, then true up against actual invoices before you close the month. No more April surprises on March margin.
Shopify's retail financial management guide makes the case that margin discipline is a monthly habit, not an annual audit — and this is where the habit starts. When you have your unit costs clean, revisit your contribution margin by SKU and your work across the inventory costing hub.
The takeaway
Invoice price plus a freight guess is a number that feels precise and is quietly wrong. A true landed cost calculation folds in duties, brokerage, insurance, and carrying cost, allocates each by its real driver, and trues up when the late invoices land. That is the difference between a margin you can price against and a margin that evaporates at year-end.
This is exactly the kind of gap CentSight surfaces. Sitting on top of QuickBooks and your bank, it reads your live numbers, watches your margins continuously, and flags the drift before it becomes a $60K surprise. If what it surfaces confirms what you already know, you're out $95. If it doesn't, it just found the profit your ERP was hiding. Start with the ecommerce finance hub.
