Advertising
Amazon Ads for KDP books: a beginner's system that doesn't burn the royalty
How to run profitable Amazon Ads for self-published books: when a book is ready for ads, the ACOS break-even math, the three campaign types, a starter structure with real budgets, and the weekly optimization loop.
Amazon Ads for KDP books:
Advertising
Amazon Ads is the only ad platform where book buyers are already standing in the bookstore with a card in hand. That proximity makes it the highest-intent channel an indie publisher has — and also the easiest place to quietly shred thin royalties, because the auction does not care that your paperback only nets $3.70.
The publishers who make ads work treat them as arithmetic with a feedback loop: know your break-even ACOS before the first campaign, structure campaigns so the data tells you something, and harvest what converts. The ones who lose treat ads as a hopeful boost button.
This guide gives you the readiness checklist, the math, a starter campaign structure with actual budgets, and the weekly loop — sized for indie books earning $2–$7 per sale, not for venture-funded brands.
When a book is ready for ads (and when ads just subsidize a weak listing)
Ads multiply the listing's conversion rate. A page that converts 10% of visitors gets ten sales from a hundred paid clicks; a page that converts 1% gets one. Same spend, tenfold difference — which is why advertising an unready book is donating money to Amazon.
Readiness means: a cover competitive against the category's top 20 at thumbnail size, a description structured as sales copy, at least 5–10 honest reviews, and a price inside its category's normal range. Each weak element taxes every click you buy.
The good news: ads on a ready book produce a second dividend. Paid sales improve organic rank, which produces organic sales the ads never paid for. That organic halo is what makes moderately-performing campaigns worthwhile overall.
- Ads multiply conversion; they cannot create it.
- Readiness: thumbnail-grade cover, sales-copy description, 5–10 reviews, normal price.
- Paid sales lift organic rank — the halo often decides overall profitability.
- Fix the listing first; it is the cheapest optimization available.
The ACOS math: know your break-even before spending a cent
ACOS (Advertising Cost of Sales) is ad spend divided by ad-attributed sales revenue. Your break-even ACOS is royalty divided by list price: a $4.99 ebook earning $3.40 breaks even at 68%; a $12.99 paperback earning $4.10 breaks even at 32%. Below break-even, ads print money; above it, they consume royalty.
Two corrections make the simple math honest. Kindle Unlimited page reads attributed to ad clicks add invisible revenue — KU-heavy books can profitably run above their visible break-even. And the organic halo means a campaign at slightly-above-break-even ACOS can still grow total profit. Start strict, loosen with evidence.
Practical bid sizing follows: with a $3.40 royalty and a 10% conversion rate, a click is worth $0.34. Starting bids of $0.30–$0.45 for ebooks and $0.40–$0.60 for higher-royalty paperbacks are sane defaults you adjust with data, not vibes.
- Break-even ACOS = royalty ÷ list price. Compute it per format.
- KU reads and organic halo justify modest overshoot — with evidence.
- Click value = royalty × conversion rate; bid below it.
- Starter bids: $0.30–$0.45 ebooks, $0.40–$0.60 paperbacks.
The three campaign types and the only one beginners need
Sponsored Products — ads in search results and on product pages — is where books win. It carries the buyer's search intent, offers keyword and product targeting, and supports both automatic and manual modes. This is 90% of indie book advertising.
Sponsored Brands (the banner with multiple books) requires three-plus titles and rewards series and catalogs; revisit it when you have them. Lockscreen ads (Kindle device placements) are cheap but low-intent — an occasional experiment, never the backbone.
Within Sponsored Products, the two targeting modes split the work: automatic campaigns let Amazon mine your metadata for search terms (research), and manual campaigns bid on the keywords and competitor ASINs you choose (precision). The starter structure uses both deliberately.
- Sponsored Products = the indie workhorse; start and stay here.
- Sponsored Brands needs 3+ titles; lockscreen is a side experiment.
- Auto campaigns do research; manual campaigns do precision.
- Product targeting (competitor ASINs) is manual mode's second weapon.
A starter structure: three campaigns, $7 a day
Campaign one — auto discovery, $2/day: default bids slightly low, runs continuously as your keyword research engine. Campaign two — manual keyword, $3/day: 15–25 phrases from your KDP keyword work and the category's vocabulary, exact and phrase match. Campaign three — manual product targeting, $2/day: 10–20 ASINs of comparable books whose buyers are your buyers.
Resist the urge to launch ten campaigns: small budgets spread thin produce no statistically meaningful data anywhere. Three campaigns at $7/day total generate interpretable numbers within two to three weeks on most books.
Negative keywords are the unsung tool: from day one, add obvious mismatches (free, used, unrelated genres) and weekly, move spending-but-not-converting search terms to negatives. Most wasted ad budget in book campaigns exits through terms nobody negated.
- Auto $2 + manual keywords $3 + product targeting $2 = $7/day starter.
- Few campaigns, enough budget each — data quality over coverage.
- Negate aggressively: 'free', wrong genres, spending non-converters.
- Two to three weeks before structural judgments.
The weekly loop: harvest, negate, adjust
Once a week, fifteen minutes: open the search-term report. Harvest — search terms with sales move into the manual campaign as exact match with a slightly raised bid. Negate — terms with 15+ clicks and zero sales become negatives. Adjust — keywords above break-even ACOS get bids cut 20%; keywords well below it get bids raised 10% to buy more volume.
Judge slowly. Attribution lags days, KU reads accrue for weeks, and book sales are low-volume events: a keyword with 8 clicks and no sale is not yet information. Decide on multi-week windows, never on yesterday.
Scale what works and starve what does not: budget flows toward the campaign with the best post-halo economics. And remember the catalog effect — ads on book one of a linked series or a visible pen-name catalog earn read-through revenue no single-book ACOS captures. Advertising rewards exactly the catalog structure that publishing rewards.
- Weekly: harvest converters, negate spenders, adjust bids ±10–20%.
- Multi-week judgment windows; low-volume data lies daily.
- Budget flows to post-halo winners.
- Series and catalogs make the same ACOS worth more.
Operational checklist
- Listing ready: competitive cover, sales-copy description, 5–10 reviews, normal price.
- Break-even ACOS computed per format before launch.
- Three starter campaigns live: auto, manual keyword, product targeting.
- Starting bids set below computed click value.
- Obvious negative keywords added on day one.
- Weekly fifteen-minute loop scheduled: harvest, negate, adjust.
- Judgments made on 2–3 week windows, royalty-based not unit-based.
- KU page-read revenue and organic halo included in profitability review.
FAQ
How much should I spend on Amazon Ads as a new author?
$5–$10 per day across two or three campaigns is enough to generate real data within weeks without material risk. Scale only what the search-term report proves; total spend should follow evidence, not ambition.
What is a good ACOS for books?
Relative to your break-even, which is royalty ÷ price: roughly 68% for a $4.99/70% ebook, near 30% for typical paperbacks. Campaigns under break-even are profitable immediately; slightly above can still win once KU reads and organic halo are counted.
Why do my ads get impressions but no clicks?
The ad shows but loses to neighbors — almost always a thumbnail or relevance problem. Check the cover against the books beside it in search results, and tighten targeting toward queries where your book plausibly belongs on the page.
Why clicks but no sales?
The ad works and the product page does not. Description, reviews, price and look-inside carry conversion; fix them before raising spend. Also verify the clicked terms actually match the book — misleading relevance produces curious clicks, not buyers.
Do Amazon Ads work for Kindle Unlimited books?
Often better than for purchase-only books: a click that becomes a borrow earns page-read revenue that ACOS does not display, so KU-enrolled books can sustainably run at higher visible ACOS. Estimate KU revenue per borrow when you review profitability.