Reference guide

Amazon Ads for KDP authors: a practical guide

A neutral, step-by-step reference for self-published authors running Amazon Sponsored Products: campaign structure, bidding from your royalty, ACOS vs TACoS, keyword vs product targeting, auto vs manual, negatives and budget. Formulas are shown so you can check every number yourself.

Last reviewed by the InteliAds team.

What Amazon Ads are for a KDP book

Amazon Ads lets you pay per click to show a book in Amazon search results and on other product pages. For KDP authors the core format is Sponsored Products: an ad for one book (one ASIN per ad) that looks like a normal listing with a Sponsored label. You pay when a shopper clicks, not when the ad is shown.

Ads do not fix a book that does not convert. Cover, title, description, price, reviews and the look inside decide whether a click becomes a sale. Ads only buy the click, so get the product page right before spending.

Related: Sponsored Products (glossary)

Campaign structure that stays readable

An Amazon Ads account holds campaigns; each campaign holds ad groups; each ad group holds the ad (the book) and its targets (keywords or products). Budget and bidding strategy live on the campaign. Bids live on each target.

A structure most authors can maintain: one auto campaign per book to discover search terms, one manual keyword campaign, and one manual product-targeting campaign. Keep different formats (ebook, paperback) and different marketplaces in separate campaigns, because their royalties and therefore their break-even points differ.

Name campaigns so the book, format, marketplace and targeting type are visible in the name. You will filter by those words for months.

  • Auto: discovery, low bids, harvest converting search terms
  • Manual keyword: proven search terms, exact and phrase match
  • Manual product: competing or comparable book ASINs and categories

Start from the royalty, not the list price

Amazon reports ACOS against the list price of attributed sales. Your profit depends on the royalty you keep. A $4.99 ebook at the 70% option keeps roughly $3.49 minus delivery cost; a $12.99 paperback at 60% keeps 60% of the list price minus printing cost. The same ACOS can be profitable on one and a loss on the other.

Break-even ACOS = royalty per sale ÷ list price × 100. If your ACOS is below it, the ad sales covered their own click cost on a per-sale royalty basis. If it is above, each ad sale lost money unless read-through or page reads make it back.

Max cost per click at break-even = royalty per sale × conversion rate. If a book earns $3.40 per sale and converts 8% of clicks, paying more than about $0.27 per click loses money on average.

Related: KDP royalty calculator · Break-even ACOS calculator

ACOS, ROAS and TACoS: what each tells you

ACOS (advertising cost of sales) = ad spend ÷ ad-attributed sales × 100. It judges the ads in isolation. ROAS is the inverse: ad sales ÷ ad spend, so 25% ACOS equals a ROAS of 4.

TACoS (total advertising cost of sales) = ad spend ÷ total sales (ad plus organic) × 100. It shows how dependent the book is on ads. A falling TACoS while ad spend holds steady usually means organic sales are growing alongside the ads.

Use ACOS against your break-even ACOS to judge individual targets. Use TACoS over weeks to judge whether advertising is building the book or just renting sales.

Related: ACOS, ROAS and CPC calculator · TACoS calculator

Auto vs manual campaigns

Auto campaigns let Amazon choose where to show the book from its content and metadata. Amazon groups auto targeting into close match, loose match, substitutes and complements, and you can bid on each group separately. Auto is the cheapest way to learn which real shopper searches and which product pages lead to sales.

Manual campaigns target what you choose: keywords or products. They give control over bids per target and are where proven terms go once auto data shows they convert.

A common loop: run auto, read the search term report, move converting search terms into a manual campaign, and add them as negatives in the auto campaign so the two do not bid against each other.

Keyword targeting vs product targeting

Keyword targeting bids on what shoppers type. Match type controls how close the search has to be: broad matches related searches in any order, phrase matches searches containing the phrase in order, exact matches the term itself and close variants.

Product targeting bids on specific ASINs (for example a comparable book in your subgenre) or whole categories. It places the book on product pages and in search results related to those products. For fiction, comparable-author ASINs are often more precise than broad genre keywords.

Neither is better in general. Compare them per book on cost per click, conversion rate and ACOS against break-even, not on impressions.

Related: Match types (glossary)

Bidding: how to set and change bids

Start new targets at a bid you can afford at break-even, using the max-CPC formula above with a conservative conversion rate. Amazon's suggested bid reflects competition, not your royalty.

Campaign bidding strategies: dynamic bids down only (Amazon may lower the bid when a sale is less likely), dynamic up and down (may raise or lower it), and fixed bids. Placement adjustments can raise bids for top of search or product pages on top of the base bid.

Change bids on evidence, not on a day of data. Ads and KDP reports lag, and attribution for a click can arrive days later. A practical rule: decide after enough clicks to expect at least one or two orders at your normal conversion rate. Lower bids on targets well above break-even ACOS; raise bids carefully on targets well below it that are limited by impressions.

Negative targeting

Negative keywords and negative products tell Amazon where not to show an ad. Use negative exact to block one search term, negative phrase to block every search containing a phrase, and negative product targets to stop showing on specific ASINs.

Negate search terms that have spent more than a sale is worth without converting, and terms that are clearly irrelevant (a different genre, a free-book search, an author name you do not want to bid on). Negate terms you have moved to a manual campaign inside the auto campaign.

Review negatives before adding them in bulk. A term with no sales after a handful of clicks may simply not have had enough clicks yet.

Related: Amazon KDP ads: clicks but no sales

Budget

A daily campaign budget caps spend per day across the campaign. If a campaign runs out of budget early in the day, it misses later shoppers; if it never comes close, the budget is not the constraint and bids or targets are.

Size total spend to what the catalog can earn back. Divide monthly ad spend by monthly royalties (including page reads you can measure) per book. If that ratio keeps rising while sales do not, the extra spend is not paying for itself.

Put budget on the books and targets with proof first. A small test budget per new book is enough to learn which targets convert.

Related: What KDP ads cost

A weekly routine

Once a week, per book: check spend against royalties, sort targets by spend with no orders, compare each target's ACOS to break-even ACOS, harvest converting search terms from auto into manual, add negatives for proven waste, and note every change so you can judge it next week.

  • Spend vs royalty per book (not just ACOS)
  • Targets and search terms with spend and no orders
  • ACOS vs break-even ACOS per target
  • Harvest, negate, then adjust bids
  • Log what changed and why

Where InteliAds fits

Everything above can be done in the Amazon Ads console and a spreadsheet. InteliAds automates the reading and the logging: it connects Amazon Ads through the official API and KDP royalties through the KDP Sync Chrome extension, shows true profit per book, lists wasted keywords, targets and search terms in one Targeting hub, and runs IF/THEN rules on an hourly schedule with Dry Run, bid min/max guardrails, cooldowns and revert.

Related: KDP ads optimizer · KDP ads automation · Pricing

See profit per book in your own account