Free tool · no signup
How much of your ad spend
is buying nothing?
Three numbers off your Seller Central dashboard and you have the answer. It is your target, not ours — the calculator only does the arithmetic you would do yourself if you had ten minutes.
Run the numbersYour account, last full month
Spend buying nothing
34% of your spend. This is what 20 000 € costs above what 52 632 € of ad revenue needs at 25% ACoS.
Mirox Growth
20 ASINs across 2 marketplaces fits Growth. Flat fee — never a percentage of the spend it is meant to reduce.
Left over
The subscription is 5.8% of the overspend. Recover that share and it has paid for itself.
Sanity check. Independent SaaS PPC research puts wasted spend at 30–40% of the average Amazon ad budget — 6 000 € to 8 000 € at your spend. Your own target implies 6 842 €. Where the two disagree, trust yours: it is built from your account, not an average of everyone’s.
No email required for the number above. Simulation Mode runs 30 days on your real account, read-only, without a card — it replaces this estimate with the bids Mirox would actually have placed.
The arithmetic, in full
ACoS is ad spend divided by the revenue those ads produced, so the revenue is recoverable from the two numbers you already have: revenue = spend ÷ ACoS. The spend that same revenue would need at your target is revenue × target, and the gap between the two is the figure at the top of the page. Written as one line: overspend = spend × (1 − target ÷ current).
Two things that formula deliberately does not do. It does not assume revenue holds while spend falls — it assumes the opposite, that the same revenue is bought more cheaply, which is the conservative reading. And it does not credit Mirox with recovering all of it. What it gives you is a ceiling on the prize, calculated from your own stated target, which is the only honest thing a calculator on a vendor’s website can offer before it has seen your account.
If your target and your current ACoS are the same, the answer is zero and the page says so. A calculator that always finds waste is a sales brochure with an input field.
Where the gap usually lives
Bids that never come down
A keyword that converted in March gets the same bid in September because nothing recalculated it. Rules-based tools re-run the same rule; they do not notice that the underlying conversion rate has halved. Most of the recoverable waste in an account is here, spread thin across hundreds of keywords rather than concentrated in an obvious few.
Search terms nobody negated
Every automatic campaign buys traffic for phrases that were never going to convert. They are individually small and collectively expensive, and they compound: an unnegated term keeps buying for as long as the campaign runs.
One threshold for every marketplace
A target set for Amazon.de applied unchanged to Amazon.fr and Amazon.it prices three different auctions as if they were one. Sellers running cross-marketplace campaigns without per-region tuning report an average 28% ACoS penalty for it.
None of the three is visible from a monthly ACoS figure, which is why the calculator can only bound the problem. Seeing which keyword, which search term and which marketplace needs a per-bid decision trace — there is a real one to read, no login required.
Questions this raises
- How is the waste figure calculated?
- One subtraction. Your ad revenue is your spend divided by your current ACoS. The spend that revenue would need at your target ACoS is that revenue multiplied by the target. The difference is what you are paying above your own target — no model, no assumed uplift, nothing of ours in the number except the arithmetic.
- Why does it ask for my target ACoS instead of telling me a good one?
- Because there is no good ACoS in the abstract. 25% is comfortable on a 45-point gross margin and loss-making on a 20-point one. The ACoS calculator works out your break-even from your margin; this one prices the distance between where you are and where you said you want to be.
- Is the 30–40% waste figure yours?
- No, and that is why it is shown as a cross-check rather than as the answer. It comes from independent SaaS PPC research into the average Amazon ad budget. An average cannot be wrong about your account because it was never about your account, so the estimate that drives this page is built from your numbers instead.
- Do I have to give you an email to see the result?
- No. The number appears as you type and the page never posts anything anywhere. If you want the real version rather than an estimate, Simulation Mode runs 30 days on your live account, read-only, without a card — and that does need an account.
- Will Mirox actually recover the whole amount?
- Almost certainly not, and anyone promising it should be treated with suspicion. Some of the gap is structural — a category where CPCs are simply high, a launch you are deliberately overpaying for. What Simulation Mode gives you is the honest version: the bids Mirox would have placed on your real keywords for 30 days, with the reasoning attached, before a cent of your money moves.
From estimate to evidence
An estimate is a guess.
Simulation Mode is a record.
Connect your account read-only for 30 days. Mirox scores every keyword and writes down the bid it would have placed and why — without placing it. At the end you compare its decisions against what actually happened.
30 days free on Simulation Mode · no card required · connect and disconnect in one click