ACOS vs TACoS: The Amazon Ads Number That Decides

Almost everyone advertising a book on Amazon looks at ACOS and decides with it: under some number the campaign is working, over it the campaign isn’t. The number is usually 30%, inherited from ecommerce. The problem is that 30% means nothing on its own in KDP, because ACOS measures spend against the book’s list price — and you don’t earn the list price. You earn a fraction of it that depends on format, royalty tier, and, in print, the cost of printing. A 30% ACOS can be an excellent business on a 70% eBook and a loss on a paperback.

The two metrics, without ambiguity

	ACOS	TACoS
Formula	Ad spend ÷ ad-attributed sales	Ad spend ÷ total sales (ad + organic)
What it measures	Campaign efficiency, in isolation	How much of your whole business advertising carries
Rises when	The ad converts worse or clicks get pricier	Advertising supports a growing share of your sales
Falls when	The ad converts better	Organic sales grow faster than spend
Good for	Bid, keyword and negative decisions	Deciding whether the book stands on its own

The practical difference is the denominator. ACOS completely ignores sales the ad caused indirectly: someone sees your ad, doesn’t buy, comes back two days later searching your title by name, and buys without touching the ad. That sale is organic in the report and never shows up in your ACOS — but you paid for it.

On Amazon that effect isn’t marginal, because rank feeds on sales velocity. The sales you buy with advertising push the book up the results, and that position generates more sales with no ad attached. It’s the loop that makes KDP advertising closer to buying rank than to buying sales.

Your break-even ACOS is your royalty rate

Here’s the calculation almost nobody runs. The point where a campaign stops losing money isn’t 30% — it’s the percentage of the price that actually reaches you. Four cases at the same 30% ACOS:

Format and price	Royalty per sale	Break-even ACOS	Profitable at 30% ACOS?
eBook at $4.99, 70% tier	≈ $3.40	≈ 68 %	Yes, with wide margin
eBook at $2.99, 35% tier	≈ $1.05	35 %	Yes, but barely
eBook at $0.99, 35% tier	≈ $0.35	35 %	Yes, though each sale nets cents
Paperback at $14.99, 60% minus print	≈ $5.20	≈ 35 %	Barely

An eBook’s royalty is the price times the tier, minus the per-megabyte delivery fee; a paperback’s is 60% of list minus printing cost, which depends on page count and ink and comes from your own KDP dashboard. The figures above are illustrative, with an assumed print cost — yours has to be computed with your numbers. How the two tiers work, and the delivery fee the dropdown hides, is covered in KDP royalties: 35% vs 70%.

The conclusion is uncomfortable but simple: the same 30% ACOS leaves a huge margin on an eBook inside the 70% band and almost nothing on a paperback. Copying another author’s ACOS target without knowing their format and price is copying the answer to a different sum. Which is why the ACOS target gets set after the price decision, not before — and why the price shouldn’t be copied either, as covered in Kindle eBook pricing.

The page reads ACOS can’t see

If the book is in KDP Select, there’s a second hole. A Kindle Unlimited reader who arrives through your ad doesn’t generate a sale — they generate page reads, paid at a rate Amazon sets monthly. Amazon Ads doesn’t attribute that revenue to the campaign, so it shows up as a click that cost money and sold nothing.

In a long genre title with readers who finish what they start, page reads can be a large share of real revenue. Judging that campaign on ACOS alone guarantees you pause it exactly when it was working.

How to use both

1. Use ACOS for keyword-level decisions. It’s the right metric for the small stuff: which search term deserves a higher bid, which comes down, which gets paused, which becomes a negative. There, ACOS’s isolation is a feature, not a flaw.
2. Use TACoS for book-level decisions. If TACoS falls month over month while total sales rise, advertising is building a book that increasingly stands on its own. If it rises, you’re renting sales.
3. Compute your break-even ACOS before setting a target. One number per format and price, written down somewhere. Without it, any target is superstition.
4. Harvest search terms before touching bids. The search term report tells you what people actually typed. Converting terms get promoted to exact-match campaigns; terms that spend without selling become negatives. It’s the highest-return work and the least-done.
5. Give every change enough clicks to mean something. A keyword with six clicks says nothing. Daily noise on Amazon is high, so compare weeks against weeks, the same way you would with Book Performance metrics.

FAQ

What’s a good ACOS for a book? It depends on your royalty per sale, not on an industry benchmark. Compute your break-even ACOS (royalty ÷ price) and set the target below it, keeping whatever margin you want. For an eBook inside the 70% band that break-even sits near 70%; for a paperback it’s usually under 40%.

What if I’m willing to lose money during launch? That’s a legitimate decision when it’s deliberate and has an end date: you’re buying sales velocity to push rank and the organic sales that follow. What doesn’t work is discovering six months later that you’ve been paying for every sale since spring.

Does Markfolio calculate TACoS? Ads Optimization works from your Amazon Ads report: it computes ACOS, ROAS, CPC, CTR and conversion, and the bid, pause and negative recommendations come out of those. TACoS additionally needs your organic sales, which live in your KDP royalty reports — that’s the side Profit Engine covers.

Why does my ACOS get worse when I raise the budget? Because the extra budget usually buys less relevant impressions — the tightest searches were already covered. Raising budget without widening the term list almost always degrades average efficiency.

Are automatic campaigns worth running? Yes, as a data source. An automatic campaign on a contained budget surfaces search terms you’d never have guessed; once proven, those terms move to exact-match manual campaigns. Using it as your main long-term campaign is what gets expensive.