Why your Amazon Ads ACOS is lying to you

30 July 2026 · 6 min read · All guides

Not because it's broken. Because it's measuring money that was never yours.

A while ago I put a small budget behind a book on Amazon Ads. After a few weeks the console showed me a keyword with a return on ad spend of 14.65×. Fourteen and a half dollars back for every dollar in. In any other advertising context that's a number you'd frame.

My royalty statement did not look like a book that was winning.

It took sitting down with both reports side by side to find the gap, and once you see it you can't unsee it. That keyword lost money on every single sale it made. So did every other keyword the console was praising. The dashboard wasn't malfunctioning — it was answering a different question from the one I was asking.

The console reports what the customer paid Amazon

When Amazon Ads says your campaign produced $59.96 in "sales", it means customers handed Amazon $59.96. It does not mean you received $59.96. It doesn't know what you received, because your royalty rate, your printing cost and your delivery fee live in KDP, which is a different system with different reports.

So ACOS — advertising cost of sale — is your spend divided by retail revenue. For a retailer selling their own product at a known margin, that's a reasonable yardstick. For an author earning a fraction of the cover price, it's measuring the wrong money.

Here's the same campaign viewed both ways. A paperback listed at $14.99, on the 60% royalty plan, costing $4.85 to print, so the author earns $4.14 a copy:

That example still turns a small profit. Change one variable — make it a full-colour picture book where printing eats most of the cover price — and the same 4.76× return puts you deep underwater.

The number you actually need

Break-even ACOS is not an Amazon setting or an industry benchmark. It's arithmetic, and it's specific to each of your books:

Run that for a few common formats and the spread is startling:

A 30% ACOS is comfortably profitable for the first book, roughly break-even for the second, a small loss for the third, and catastrophic for the fourth. Any advice that names a target ACOS without asking what you earn per copy is guessing.

That last row is where I started. On a $0.52 royalty you need to return twenty-one times your ad spend just to stand still — which is why a 14.65× return was a loss. Some books cannot be advertised into profit at any bid, and finding that out before you spend is worth more than any optimisation trick.

Now the part that cuts the other way

If your book is in Kindle Unlimited, the console is understating you, and possibly by a lot.

When someone borrows your book after clicking your ad, you earn money for every page they read. That income lands in a different KDP report, on a different schedule, and the ads console records the borrow as zero revenue while still charging you for the click. A campaign that looks like a disaster in the console can be quietly profitable once page reads are counted.

So the error runs in both directions: retail pricing flatters your ads, and Kindle Unlimited hides real income from them. Depending on your book, either effect can dominate. The only way to know is to put both reports in the same view.

There's a further wrinkle worth knowing: nobody can tell you your Kindle Unlimited earnings for the current month, including Amazon. The per-page rate depends on the size of a shared monthly fund and the total pages everyone read, so it isn't set until after the month closes. Pull a KDP royalty statement mid-month and the earnings column for page reads reads "N/A". Any tool showing you a confident figure before then is estimating, whether or not it says so.

Why your numbers never quite match

One more thing that drives people to distraction: the ads console and your KDP dashboard will disagree about how many books you sold, and both are right.

• Different clocks. Amazon credits an ad-attributed sale to the day of the click, not the purchase. Your royalty report uses the royalty date.
• Different scope. The ads console counts only sales it attributes to ads. Your KDP reports count everything.
• Settling. Attribution keeps adjusting for about 48 hours.
• Different identifiers. Your paperback and your ebook are separate ASINs, even though readers see one book.

That deserves its own piece, and it has one: why your KDP and Amazon Ads reports never agree.

What to do about it

1. Work out your break-even ACOS per book, before you look at another campaign. It takes thirty seconds and it changes what "good" means.
2. Stop optimising on ROAS alone. A number that looks like a winner against retail can be a loser against royalties.
3. If you're in KU, don't judge on ACOS at all. Use TACOS — total ad spend against all income, page reads included.
4. Give new campaigns room. Early campaigns run hot while Amazon figures out who to show your book to.
5. Fix economics before bids. If your break-even ACOS is under about 10%, no amount of clever bidding rescues it. Price, page count, print colour and format move that number far more than any campaign setting.

AuthorTally is an independent tool and is not affiliated with, endorsed by, or connected to Amazon.com, Inc. Royalty rates, printing costs and Kindle Unlimited payouts change; the arithmetic here doesn't, but check your own figures against your own reports.