Calculating profit per product on Amazon FBA: the net margin per SKU that actually holds up

Profit per product on Amazon FBA stands or falls with the quality of your COGS input, not with the formula. The formula is well known and simple. The number it produces is only as reliable as the cost price you feed into it, and that is exactly where things go wrong for almost every seller I talk to.

Short answer. Calculate profit per product on Amazon FBA as: selling price minus referral fee (8-15%), FBA fulfillment fee, storage per unit, PPC per sale, return impact and your landed cost (COGS plus inbound, prep, customs). What remains is your net margin per SKU. The calculation is only as accurate as your COGS input.

The formula that matters for an established seller

You already know the happy-path version. For a scaling seller I write it like this:

Net profit per SKU = selling price − referral fee − fulfillment fee − storage per unit − PPC per sale − return impact − landed cost.

The referral fee sits around 15% for most categories, around 8% for electronics, higher for apparel. The fulfillment fee depends on size tier and weight. Every calculator covers this much, including the FBA calculator from Helium 10. The problem is not in these lines. It is in the two at the bottom: return impact and landed cost.

Landed cost is where the number is won or lost

COGS is not your purchase price. Landed cost is COGS plus inbound freight, prep, customs and import duties, tooling spread out across your units where relevant, and a buffer. Anyone who only enters the supplier's invoice price structurally overstates their margin. With a container that has wildly varying freight costs, or a product with expensive prep, your real margin per SKU can easily shift by 5 to 8 percentage points. That is the difference between a SKU you scale up and one that costs you money while your dashboard stays green.

Tools like Sellerboard calculate net profit neatly once the COGS is entered. Sellerboard itself points out that landed cost is more than purchase alone. But the tool does not know what your landed cost is. It has to come from somewhere, per batch, per supplier, per inbound shipment. That is exactly the input that for most sellers is manual, outdated or estimated.

Return impact and storage: the silent margin eaters

Returns lower your effective selling price, and you do not always get all the fees back. A SKU with 12% returns and damaged return stock carries a very different margin than the same SKU on paper. Storage belongs allocated per unit, not as a monthly lump. Slow movers with a low IPI devour storage, and that cost has to land on the SKU where it belongs, otherwise you are subsidizing your losers with your winners without seeing it.

Why a standard tool stops here

Sellerboard and Helium 10 cover the happy path excellently: one seller, Amazon, clean COGS. My clients are past that point. They also sell on bol and their own webshop, buy in batches with varying landed cost, and their cost price lives in a different system than their margin tool. The question is not "which tool", the question is: does the margin per product hold up across all channels combined, fed by COGS that is current?

That is what I build. First I work through the process: which steps in your COGS chain are grown-out manual work that could be done smarter or removed. Often half a day of spreadsheet work per month turns out to be an integration nobody ever set up. Then I build the solution that does handle your exceptions: a landed-cost feed per batch into the margin tool, or a custom dashboard that puts Amazon, bol and webshop side by side at SKU level. The calculation has to happen. I make sure it holds up where the standard tool does not cover your situation.