How to calculate true Amazon FBA profit per SKU
Last updated
Sale price minus product cost minus the referral fee is the shortcut that hides losing SKUs. Here is the full stack, and how far down it you actually need to go.
The full per-SKU formula
A fully loaded SKU profit figure subtracts eight things from net sales, not three:
SKU net profit = net sales, minus:
- landed cost (or your COGS basis)
- referral fees
- FBA fulfillment fees
- advertising cost
- returns and damage losses
- storage and inventory carrying cost
- other Amazon fees
Landed cost is itself more than the supplier invoice: unit cost plus inbound freight to Amazon, plus duty and customs clearance, plus inspection, packaging, labeling, and prep. The FBA profit calculator takes all of these per unit, and the fee calculator supplies the referral and fulfillment numbers for a specific size tier and shipping weight.
Then three decision metrics fall out of it:
Per-unit net profit = SKU net profit / net units soldNet margin = SKU net profit / net salesROI = per-unit net profit / per-unit landed cost
The common shortcut leaves out fulfillment, PPC, returns, storage, inbound freight, and the cash tied up in inventory. That is how a losing SKU reads as profitable for months.
Which method fits your situation
| Your situation | Start with | Watch |
|---|---|---|
| Evaluating a new product | A per-unit profit calculator | Net profit, margin, ROI, breakeven TACOS |
| Ranking SKUs that already sell | SKU Economics or Profit Analytics in Seller Central | Sales, fees, ad spend, net proceeds |
| ACOS looks fine, profit does not | Convert ACOS to TACOS | Profit after ads, breakeven TACOS |
| Replenishment costs keep changing | Weighted-average COGS, or FIFO when the gap is material | Landed cost by batch |
| Returns are eating the margin | A returns reserve per unit | Reserve versus actual return loss |
| Inventory turns slowly | Profit at 30, 90, 180, 365 days | Storage, aged-inventory surcharge, cost of capital |
| SKU profit does not match the payout | Payments reports as the control total | Unallocated fees, timing differences |
| Closing the books or filing taxes | Settlement statements and the general ledger | Accounting completeness, not SKU margin |
Operating analysis and accounting reconciliation are different jobs. The first decides what to do with the product. The second confirms the money was recorded correctly. Most sellers only need the second one at month end.
Questions
Which costs are most often missing from an FBA profit calculation?
Inbound freight, duty, PPC spread across all units, return losses, and storage. Those five account for most of the gap between a spreadsheet that says a SKU makes money and a bank balance that says it does not.
Below them sit the charges that are easy to forget because they arrive monthly rather than per order: aged-inventory surcharges, inbound placement fees, low-inventory-level fees, storage utilization surcharges, and the cost of capital on stock bought weeks before it sells. The FBA fee reference lists the current US rates for each.
Does the SKU Economics report show true profit?
It shows net proceeds, which becomes net profit only if your off-Amazon costs are complete. Amazon's report brings together sales, Amazon fees, advertising, and net proceeds at the FNSKU, MSKU, ASIN, and parent-ASIN levels, so it is genuinely useful for ranking SKUs by performance.
What Amazon cannot supply is COGS. Supplier pricing, freight, duty, brokerage, and prep are seller-entered, and a report running on incomplete cost data will overstate profit on every unit. Short reporting windows can also mislead, because orders, refunds, ad spend, and fees post on different dates.
Should profit calculations use ACOS or TACOS?
TACOS, for anything measuring average per-unit profitability. ACOS divides ad spend by ad-attributed sales, so it measures campaign efficiency. TACOS divides the same spend by total sales, which is what actually happens to your margin, because ad spend has to be carried by every unit sold, organic ones included.
The conversion is one multiplication. At 30% ACOS with 40% of sales ad-attributed, TACOS is 30% x 40% = 12%. On a $30 sale price that is about $3.60 of advertising per unit, not $9.00. Getting this backwards is the single most common reason a SKU with acceptable ACOS still loses money.
Why is my profit falling when my ACOS has not changed?
Because a shrinking share of your orders is organic. ACOS only looks at ad-attributed sales, so if PPC starts carrying a larger portion of total volume, campaign ACOS can sit perfectly still while TACOS climbs and margin erodes underneath it.
The figures worth tracking are TACOS, per-unit profit after advertising, breakeven TACOS, the ad-attributed share of sales, and whether organic order volume is declining month over month. A rising ad-attributed share with flat total sales is the early warning.
Is weighted-average COGS or FIFO more accurate for one SKU?
Weighted average is enough when replenishment costs are stable, and it is one division: total cost of all batches divided by total units in all batches. Most SKUs with comfortable margins never need more than this.
FIFO earns its extra work when batch costs diverge. With 500 units at $6 landed and 500 at $8, the weighted average is $7. While you are still selling the $6 batch, $7 understates your current profit; once the $8 batch starts moving, it overstates it. Switch methods when that difference would change a pricing, reorder, or discontinuation decision, not for theoretical precision.
How do I work out a returns reserve per unit?
Divide total historical return losses by net units sold in the same period. That gives a per-unit figure you can subtract in a sourcing estimate before you have any return history for the new product.
The loss is more than the refunded sale price. A single return can involve reversed revenue, a refund administration charge Amazon retains, a returns processing fee, an item that comes back unsellable, inspection or repackaging or disposal cost, and PPC spend that is gone regardless. Subtract the recovery value of units that can be resold. Apparel, footwear, fragile goods, and fit-sensitive products should never use a generic marketplace return rate.
How much does slow-moving inventory cost per unit?
Three costs compound with time: monthly storage, aged-inventory surcharges once stock passes the long-term bands, and the cost of capital on money sitting in a warehouse. Peak storage rates apply from 15 October 2026 through 14 January 2027, so the same unit costs more to hold in Q4 than in June.
Model the SKU at 30, 90, 180, and 365 days rather than at one point. When the extra cash recovered by continuing to sell is less than the future storage, aging, and financing cost, a markdown, removal, or liquidation is the cheaper outcome. The ROI calculator prices holding time against turns.
Can running inventory too lean also cost money?
Yes. Amazon's low-inventory-level fee is based on historical days of supply, and can apply when both the short-term and long-term historical days of supply are below 28 days. Cutting stock to dodge storage fees can therefore trade one charge for another.
That makes inventory planning a comparison rather than a rule. Price the storage and capital cost of holding more against the fulfillment surcharge and stockout risk of holding less, per SKU, rather than adopting "lean is always better" across the catalog.
Can I keep using last year's profit model after the 2026 FBA fee changes?
No. Amazon applied a 3.5% fuel and logistics-related surcharge to FBA fulfillment fees in the United States and Canada from 17 April 2026. It is calculated on the fulfillment fee, not the sale price, so a $5.04 fulfillment fee becomes about $5.22.
Recalculate whenever fees change, packaged dimensions or shipping weight change, the item crosses into another size tier, a supplier reprices, duty or freight moves, promotions change your average selling price, TACOS rises materially, return behaviour shifts, or inventory enters a new aging band. Dimensions deserve the most attention: a packaging tweak can move a product into a costlier tier and remove the margin without touching the listing price.
Why does my SKU profit not match my Amazon payout?
Usually timing, not error. Orders, shipments, ad attribution, refunds, and fees post on different dates, so a SKU rollup and a settlement statement rarely agree inside a short window. Compare over a longer period before treating the gap as a problem.
For reconciliation, use three layers: SKU-level profitability to explain product performance, Amazon settlement or Payments totals as the control total, and bank deposits plus the general ledger to confirm the cash. Investigate only differences that are persistent, material, and unexplained. Reimbursements, removals, and inventory adjustments are the usual culprits when they are real.
When is a calculator enough, and when do I need full reconciliation?
A per-unit calculator is enough for sourcing, pricing, advertising, and inventory decisions, which is most of what you do week to week. It answers whether the product still works under realistic assumptions, and it does not need to match a bank deposit to do that.
Reconcile properly when you are closing the books or filing taxes, when a SKU looks profitable but cash keeps declining, when refunds or reimbursements are material, when purchase costs vary a lot between replenishments, or when a bookkeeper needs support for the number. Define the decision first, then use the lightest tool that answers it.
How often should I recalculate SKU profitability?
Monthly as a floor, and immediately after any material change in Amazon fees, price, PPC, returns, supplier cost, product dimensions, shipping weight, or inventory age.
A workable rhythm is weekly checks on price, fees, and TACOS with a flag on any SKU below your profit floor; a monthly pass that updates COGS and inbound freight, reviews SKU Economics, and compares the rollup with settlement totals; and a full re-test of price, COGS, TACOS, and expected inventory time before every reorder. Reordering on last year's sale price and cost basis is how a SKU quietly goes underwater.
Run your own numbers: the free FBA profit calculator models fees, storage, and cost of capital per unit - no signup, nothing leaves your browser.