Before we read a single profit report, you have one job. Enter your cost of goods.
It is tempting to skip ahead to the dashboards and the charts. Don’t. If you take nothing else from this course, take this: a profit number built on a missing cost is not a profit number. It is a guess wearing a suit.
Profit has two halves, and Amazon only knows one
Strip away all the jargon and profit is just this:
What you got paid, minus what the product cost you.
The Data Hub knows the first half cold. Every dollar Amazon paid you, every fee they took, every refund they clawed back, every storage charge, every dollar of ad spend. All of that is captured for you.
But the second half, what the product actually cost you to buy or make, Amazon has no idea. It never touches their system. They don’t know if your $30 product cost you $4 or $24. Only you know that.
So until you tell the Data Hub your cost, it is working with half the equation. And half an equation can’t give you a real answer.
What happens when you leave COGS blank
Here is the part that catches people off guard. When you don’t enter a cost, the report does not just show a blank or refuse to calculate. To avoid a zero-cost situation — which would make every product look like pure profit — Seller Labs fills in a default cost of 25% of your revenue so the reports still run and stay sane.
Be clear on what that number is and isn’t. It is a Seller Labs placeholder, chosen only so there’s never a missing cost. It is not a fee Amazon charges, not an industry benchmark, and almost certainly not what your product actually costs you. It is a stand-in, and it is quietly wrong for almost everyone:
- If your real cost is 40% of revenue, the 25% assumption makes you look more profitable than you are. You will keep products you should be questioning.
- If your real cost is 10% of revenue, the assumption makes you look less profitable than you are. You might kill a winner.
Either way, you are now making keep, kill, and reorder decisions on a number that was invented to fill a gap. The whole point of this course is to give you numbers you can trust. That starts with replacing the assumption with reality.
You do not have to do all of it today
Good news: you do not need perfect cost data on all 2,000 SKUs before this is useful. Profit is concentrated. A small handful of products usually drives most of your sales.
So start there. Enter accurate COGS for your top sellers first. Get the products that actually move the needle onto real numbers, and the rest can follow over time. Even getting your top 10 to 20 SKUs done changes everything about how much you can trust your venue-level profit.
What “cost of goods” should include
Keep it simple to start, but be consistent. Your COGS is what it costs you to land one unit, ready to sell:
- The unit price you pay your supplier or manufacturer.
- Inbound freight and shipping to get it to Amazon (or to you).
- Any per-unit prep, packaging, or inspection cost.
You do not need to bake Amazon’s fees, storage, or ad spend into COGS. The Data Hub already tracks all of those separately and subtracts them for you. COGS is just your product cost. Double-counting fees here would throw the numbers off in the other direction.
Where to enter your COGS
You have two ways to update your costs:
- On the Bulk Update Products page in Seller Labs. You can set costs for many products at once instead of one at a time.
- In Claude, through the Seller Labs Amazon MCP. Tell Claude the new costs, check the preview it shows you, and approve it.
Either way, combined with the “top sellers first” approach above, you can get your highest-impact SKUs onto real numbers in a single sitting. New costs apply going forward right away. To also correct past profit, tick Update Historical COGS on the Bulk Update Products page. Updates made through the Seller Labs Amazon MCP always recalculate past profit. Some profit reports catch up after your next Amazon settlement.
For step-by-step instructions for both options, see Uploading Your Cost of Goods Sold (COGS) into the Product Catalog.
Try this in your account
One thing to know before you start: you won’t find a SKU with a truly blank cost, because Seller Labs auto-fills the 25% placeholder. And the system can’t tell you which costs are real versus still on the placeholder — there’s no flag for it, and since the 25% is a share of a price that changes, you can’t reliably spot it by the number either. You are the only one who knows which products you’ve entered real costs for. So the goal here isn’t to make the MCP detect the placeholder. It’s to build a prioritized worklist and check your own coverage against it.
Open Claude with the Seller Labs Amazon MCP connected and ask:
“Rank my SKUs by sales over the last 90 days, highest first, and show each one’s current cost of goods.”
That’s your priority order: the products that move the most money sit at the top. Go down that list and, for each one, ask yourself the only question the data can’t answer for you — did I actually enter this cost, or is it still the 25% default? Anything you haven’t personally set, enter now, either on the Bulk Update Products page or by asking Claude right in the same chat:
“Set the COGS for SKU ABC-123 to 4.50 in my [venue name] venue.”
Claude will show you a preview first and only saves the change after you approve it.
Then come back and ask:
“For my top 20 SKUs by last-90-day sales, show me each one’s cost of goods as a percentage of its average sale price.”
This is your sanity check. Any product sitting at exactly 25% is a strong candidate for “still on the placeholder” — worth a second look. It won’t catch everything (a real cost can happen to land near 25%), but it’s a useful flag for the high-impact products that matter most. When your top sellers are all on real, entered costs, you’re ready to trust the reports we build in the rest of this course.
Part 2 of 7 in Level 1: Understanding Amazon Profitability.
← Previous: Why Amazon Profitability Is So Confusing | Next: The Fee Maze: SKU-Level vs Account-Level Fees →