Your Amazon profit report is probably wrong. Here's what it's missing.
Open almost any Amazon profit dashboard and you’ll get a clean number next to each SKU: revenue, cost, profit, margin. It looks precise. For most sellers, it’s also wrong — not because the arithmetic is broken, but because several real costs never make it into the calculation in the first place.
Here are four of the biggest, in roughly the order they tend to get missed.
1. Advertising spend (10–25% of revenue)
This is the big one. Amazon PPC spend settles on a completely different report and schedule than order data, so most dashboards that build “profit” straight from orders and fees simply don’t have ad spend in the calculation at all — or they show it as one lump total for the account, not attributed to the products it was actually spent on.
For a seller running sponsored product campaigns at anything close to typical levels, advertising can run 10–25% of revenue on the SKUs it’s applied to. A product that looks like it’s making a healthy 22% margin without ad spend factored in might be making 5%, or losing money, once its actual advertising cost is attributed to it.
The attribution problem is real — Amazon’s ad reports don’t map cleanly to individual orders — but “hard to attribute precisely” is different from “safe to ignore.” A dashboard that shows margin with no ad spend line isn’t showing you a conservative number; it’s showing you a wrong one.
2. Inbound placement fees ($0.30–$0.80/unit)
When Amazon changed its FBA inbound model to let sellers choose between sending inventory to a single fulfilment centre (cheaper storage split, extra placement fee) or letting Amazon distribute it (no placement fee, different storage economics), it created a real per-unit cost that lives on a separate report from the standard fee breakdown most tools import.
At $0.30–$0.80 per unit, this looks small next to a $25 item. It’s less small next to a $8 item with a 40% margin, where it can be the difference between a product that’s worth reordering and one that isn’t. Because it’s on a different report, it’s one of the most commonly dropped costs in profit calculations that only pull the “standard” fee types.
3. Long-term storage fees
Standard monthly storage fees are usually captured correctly — they’re on the main fee report everyone imports. Long-term storage fees, charged on inventory that’s sat in a fulfilment centre past a certain age threshold, often aren’t, because they post separately and less frequently.
For a seller with reasonably fast-moving inventory this might be a rounding error. For anyone carrying slower SKUs, seasonal stock, or inventory built up ahead of a promotion that didn’t move as fast as planned, long-term storage can meaningfully erode the margin on exactly the products least able to absorb it.
4. Returns processing costs
A return isn’t just a refunded sale. There’s a processing fee, and depending on the product’s condition on return, it may come back as unsellable, sellable-but-discounted, or requiring disposal — each with a different cost outcome that a simple “refund the order” calculation doesn’t capture.
Categories with high return rates (apparel, anything with a sizing or fit component) can see this cost compound: the lost sale, the processing fee, and the diminished or zero value of the returned unit, stacked on top of each other. A profit report that treats a return as “subtract the refund amount” is understating the true cost every time.
Why this keeps happening
None of these costs are hidden, exactly — they’re all in Amazon’s reporting somewhere. The problem is that they live on different reports, on different schedules, in different formats, and a tool built to import “the fees report” and stop there will always miss the ones that don’t live in that one place.
The result is a profit number that’s internally consistent (the math is right, given its inputs) but not actually correct (the inputs are incomplete). That’s more dangerous than an obviously wrong number, because it looks trustworthy right up until you make a pricing or inventory decision based on it.
What accounting for it properly actually requires
Getting per-product margin right means pulling from more than one Amazon report, attributing costs that don’t come pre-attributed to a specific order, and doing it consistently enough that the number means the same thing this month as it did last month.
That’s the difference between a dashboard, which summarises what’s easy to summarise, and books, which are built to be complete. SellerTally posts advertising cost, placement fees, long-term storage, and return processing as their own accounts in your ledger — not folded into a single “fees” bucket, not left off because the report they live on is inconvenient to import.
If your current profit numbers have never included ad spend or placement fees, it’s worth checking what your real per-SKU margin looks like once they’re in. For a lot of sellers, it changes which products are actually worth reordering.