If you already run volume on Amazon, you don't need COGS explained from scratch. You need it pinned down precisely — because the number quietly drives your gross margin, your tax return, and every "can I afford to discount this?" decision.
Most ranking guides stop at the textbook formula and a vague list of "costs to include." This one draws the line the accountants actually draw, then shows how fees and ad spend turn a healthy COGS into a thin per-order profit.
The Amazon COGS formula
The accounting formula is the same one every ranking page cites:
COGS = Beginning Inventory + Purchases − Ending Inventory
Say you start the month holding $6,400 of inventory, buy $18,200 more, and finish with $5,100 on the shelf. Your COGS for the month is $6,400 + $18,200 − $5,100 = $19,500. That $19,500 is the cost of the units that left your account — nothing more.
The formula works because it backs into "what did I actually sell?" without counting every unit by hand. It assumes your records of purchases and your month-end inventory count are accurate — which is exactly where most sellers leak money.
What belongs in COGS — and what doesn't
Strictly, COGS is the landed cost of the product: the factory or supplier price, plus inbound freight, duties, and any prep or inbound shipping to get units into Amazon's network. These are the costs of making the unit ready to sell.
Operating expenses sit outside COGS. Rent, software subscriptions, your salary, and accounting fees are overhead, not the cost of a specific unit — a distinction worth getting right, which our operating expenses definition breaks down. Long-lived purchases like equipment are a third bucket entirely; see the line between capital and operating expenses.
The gray zone is Amazon's own fees, and the ranking guides blur it. The cleaner treatment: referral and FBA fees are selling costs, not COGS. They're triggered by the sale, not by producing the unit. Lumping them into COGS makes your gross margin look worse than it is and muddies which lever to pull.
The fees that are NOT COGS (but kill your margin anyway)
Keeping fees out of COGS doesn't mean ignoring them — it means tracking them in the right line so you can see them. And on Amazon they're large.
The referral fee is 15% of the sale price for most categories, ranging from 5% to 45% depending on category, according to Feedvisor's referral fee guide. On a $40 order in a standard category, that's $6 gone before you've touched product cost.
FBA fulfillment fees stack on top, charged per unit by size and weight tier for picking, packing, and shipping. Add storage fees, returns processing, and advertising, and the selling-cost layer often dwarfs the per-unit COGS itself.
Inventory valuation: FIFO, LIFO, or weighted average
Because unit costs change between purchase orders, you need a rule for which cost "leaves" when a unit sells. The three standard methods are FIFO (first in, first out), LIFO (last in, first out), and the weighted average method, which blends every unit's cost into one rolling figure.
FIFO is the most common and usually the simplest to defend. If your supplier prices drift up over time, FIFO assigns your older, cheaper costs to COGS first — which raises reported gross profit relative to LIFO.
For multi-batch sellers, the rolling-average approach often tracks reality best; our walkthrough of weighted average cost of goods sold shows the arithmetic. Pick one method and stay consistent — switching mid-year distorts every comparison you'll try to make.
Worked example: COGS vs. true per-order profit
Here's where a clean COGS number stops being enough. Say you're running about 420 orders a month at a $40 average order value in a standard 15% referral category.
Take one $40 order with a $16 landed unit cost:
| Line item | Amount |
|---|---|
| Sale price | $40.00 |
| COGS (landed unit cost) | −$16.00 |
| Gross profit | $24.00 |
| Amazon referral fee (15%) | −$6.00 |
| FBA fulfillment fee (illustrative) | −$5.40 |
| Ad spend allocated to the order | −$7.00 |
| True per-order profit | $5.60 |
The referral fee (15%) is sourced above; the rest are illustrative operating numbers. The point stands either way: your COGS looks like a 60% gross margin, but once selling costs and ad spend land, you keep $5.60 — about 14% of the sale. Across 420 orders that's roughly $2,350 for the month, not the $10,000 the gross-margin line implied.
That gap is why COGS alone is a dangerous number to steer by. For context on where online stores actually land, one NYU retail dataset cited by SPS Commerce puts average online-retail gross margin near 42.5% and net margin under 5% — a reminder that the distance between gross and net is where businesses live or die.
The POD wrinkle: no inventory to count
If you fulfill through print-on-demand, the inventory formula barely applies. You hold no stock, so there's no beginning or ending inventory to net out — your COGS is simply the per-unit supplier charge on each order you actually fulfilled.
That makes per-order COGS cleaner to compute but more punishing when something goes wrong. A refunded POD item can't be restocked, so the supplier cost is unrecoverable — you eat the full COGS on top of the refund. There's no "it goes back on the shelf" cushion.
It also means your COGS moves with every supplier price change and shipping-rate update, order by order. Getting it onto the books correctly each month is its own discipline, covered in recording cost of goods sold.
Why operators get this wrong
The usual failure isn't the formula — it's stale inputs. Supplier prices rise, freight spikes, a category's referral tier changes, and the COGS in your spreadsheet keeps using last quarter's numbers.
The second failure is channel blindness. If you sell on Amazon and run a Shopify storefront, each channel has its own fee stack and its own true per-order profit, and a blended average hides the one that's actually losing money. The broader framing of these levers lives in our ecommerce ops economics guide.
The fix is to compute profit at the order level, per channel, on live costs — not to admire a gross-margin percentage once a quarter.
Where Victor fits (and where it doesn't)
If you also run a Shopify store alongside Amazon, that's where PodVector AI's AI employee, Victor, earns its keep. Victor computes true per-order profit on your Shopify orders by combining live supplier costs from Printify, Printful, and Gelato with your Meta Ads and Google Ads spend — the same stack-it-all-up math as the worked example above, run continuously instead of once a month.
To be clear: Victor works your Shopify business, not Amazon — it doesn't pull Seller Central data. But if Amazon taught you to respect the gap between COGS and profit, Victor closes that gap on the store you own, and delivers the reports to your Google Drive. See what Victor does with your real numbers.
FAQs
Does Amazon COGS include FBA and referral fees?
Strictly, no. COGS is the landed cost of the product — supplier price plus inbound freight and duties. Amazon's referral and FBA fees are selling costs triggered by the sale, so cleaner books track them as operating/selling expenses, not COGS. Some guides fold fulfillment fees into COGS; it's defensible for internal profit math, but keeping them separate shows you which lever to pull.
What is the COGS formula for Amazon sellers?
Beginning Inventory + Purchases − Ending Inventory = COGS. It calculates the cost of the units you sold during the period without counting each one individually, as long as your purchase records and inventory counts are accurate.
How is COGS different for print-on-demand sellers?
POD sellers hold no inventory, so there's nothing to net out. Your COGS is just the per-unit supplier charge on each fulfilled order. The catch: a refunded POD unit can't be restocked, so that supplier cost is a total loss on top of the refund.
Which inventory valuation method should I use?
Most Amazon sellers use FIFO because it's simple and maps to how goods actually move. Multi-batch sellers with shifting unit costs often prefer the weighted average method. Whichever you choose, apply it consistently — switching methods mid-year breaks every period-over-period comparison.
Is a high gross margin enough to know my store is healthy?
No. Gross margin only subtracts COGS. Once Amazon fees, advertising, and overhead come out, net margin is often a fraction of gross — frequently under five percent for online retail. Judge an operating store on true per-order profit, not the gross line.