What this does and does not tell you

Attribution windows overlap, so the same order can be counted by every platform that touched it — which is why the claims can add up to more than the money that arrived.

Blended figures cannot be inflated that way: net sales and total ad spend are each counted once, whatever the dashboards say.

An overlap is a decision defect, not a saving. Finding it puts no money back — what it changes is which claims you can trust when you decide where the next dollar goes.

What it tells you

• whether your platforms' claims add up to more than the revenue that reached you, and by at least how much
• your blended marketing efficiency against the ROAS your platforms average out to
• whether a first order covers what it cost to acquire the customer

What it cannot tell you

• these are averages across every order, so they show the shape of a problem, not where it comes from
• margin by product, channel, discount code or country needs order-level data; no blended number reaches it
• when platforms claim less than your net sales, this test cannot see overlap at all — overlap between platforms can still be hiding inside the claims
• claims and sales are counted over different windows: attribution straddles month ends and claims are usually counted before refunds
• the answer is only as good as the margin you entered; a gross margin, or one struck on VAT-inclusive prices, overstates profit per order before ads and payback

Net sales excludes VAT and shipping. The margin you enter is taken after product cost, delivery, payment fees and packaging, so profit per order before ads is struck on net sales.

Every one of those needs order-level data. That is what I do. A full month of that output, on an illustrative brand, is here: a sample month. Send me what the calculator showed you and I will tell you what I would check first: