Amazon book ads work on one idea: a reader searches Amazon, your book appears in the results or on another book’s page, and you pay only if they click. Setting up the first campaign takes about ten minutes. Running it well comes down to two things: the numbers you work out before you start, and how patient you are afterwards.
Before you spend anything
An ad buys a visit. The book page makes the sale. If the cover, title, description, price and reviews do not turn browsers into buyers now, paid visitors will not do better, and you will simply pay to find that out.
The second thing to settle is how much a sale is worth to you. Your break-even ACOS is the most you can spend on ads, as a share of the sale price, before the ad loses money:
royalty per sale ÷ sale price × 100
A $4.99 eBook on the 70% royalty plan earns about $3.40 a sale after the small delivery fee, so it breaks even near 68%. A $12.99 paperback earning $2.50 breaks even near 19%.
That one number sets every limit that follows: what you can bid, how long you can let a keyword run, and when a campaign is working. What ACOS is, and how break-even works.
Last, decide what the campaign is for. Profit on this one book, or visibility for the first book of a series you expect readers to finish? A series can justify a higher ACOS on book one, because each reader it buys is worth more than one sale. That is a deliberate choice, not a licence to ignore the numbers.
Where Amazon book ads are set up
In your KDP Bookshelf, choose Promote and Advertise for the book, then create an ad campaign in the marketplace you want — Amazon.com, Amazon.co.uk, Amazon.de and so on. That takes you to the Amazon Ads console, where the campaigns live.
Book ads can appear on Amazon, on Kindle devices and on Goodreads, and only when Amazon is the featured offer for the book. Each marketplace has its own campaigns and budget. Start in the one where most of your sales already come from; you can copy what works to the others later.
Step 1 — one Sponsored Products campaign, automatic targeting
Choose Sponsored Products, add the book, and choose automatic targeting. With automatic targeting Amazon decides which searches and product pages to show your book on, working from your book’s own details. The point is not to be efficient in week one. The point is to learn which searches actually find and buy your book.
Automatic campaigns split the traffic into four groups, and each can carry its own bid or be switched off:
- Close match — searches closely related to your book.
- Loose match — searches more loosely related to it.
- Substitutes — pages of books a reader might buy instead of yours.
- Complements — pages of books a reader might buy alongside yours.
For bidding strategy, pick Dynamic bids – down only. Amazon can then lower your bid when a click looks unlikely to sell, but will never raise it. The “up and down” option can raise a bid by up to 100%, which is a bad fit for a campaign whose job is to gather data cheaply.
Step 2 — a daily budget you could lose for three weeks
The daily budget sets what you can spend on average. On busy days Amazon may spend up to double it by default (you can limit that to 25% in the campaign settings), but a month never costs more than the daily budget times the days in it. Choose a figure you could lose in full over three weeks and still be fine, because that is the price of the data.
| Daily budget | Three weeks | Clicks at $0.40 | Sales at 5% conversion |
|---|---|---|---|
| $5 | $105 | ~260 | ~13 |
| $10 | $210 | ~525 | ~26 |
| $20 | $420 | ~1,050 | ~52 |
Those are ceilings, not forecasts. A new book on a modest bid often cannot spend its whole budget, because it wins too few auctions. That is not a problem to fix in week one.
Step 3 — bid from your royalty, not Amazon’s suggestion
Amazon shows a suggested bid. It tells you what other advertisers are paying, not what your book can afford. Your own ceiling comes from two numbers:
royalty per sale × conversion rate
A $3.40 royalty at a 5% conversion rate justifies about $0.17 a click. A $6.99 royalty at 5% justifies about $0.35.
Start at about two-thirds of that figure and let the results tell you where to raise it. If you do not know your conversion rate yet, 5% is a reasonable placeholder for a book with a good cover and a few reviews. Use your real figure as soon as you have one.
Free toolAmazon PPC bid calculatorA starting bid from your price, royalty and conversion rate.Step 4 — leave it alone for two to three weeks
This is the step most authors skip, and it costs them the most money. Two things make early numbers misleading:
- Sales arrive late. For books, Amazon credits a sale to an ad click from up to 14 days earlier, and order data can take up to 12 hours to appear. The last few days always look worse than they will end up.
- Small numbers lie. At a 5% conversion rate, one keyword gets zero sales from its first 20 clicks about a third of the time, purely by chance. A keyword that looks dead after 20 clicks is often just unlucky.
How many clicks it takes before a missing sale actually means something is covered in judging a keyword.
Step 5 — mine the search term report
After two or three weeks, download the search term report. It lists the exact words shoppers typed before they clicked your ad, and which of those searches turned into orders. It is the most valuable free keyword research you will ever get, because it comes from real buyers of your book.
- Searches that sold at an ACOS under your break-even go into a manual campaign as exact-match keywords, each with its own bid.
- Add those same searches to the automatic campaign as negative exact keywords, so your two campaigns stop bidding against each other.
- Searches with plenty of clicks and no orders become negative keywords, so you stop paying for them.
Step 6 — change one thing at a time
Once the manual campaign is running, there are three levers: the keyword bids, the daily budget, and the placement adjustments. Placement adjustments let you bid more for top of search or for product pages without changing the base bid.
Change one lever, write down the date, and judge the change only on data from after that date. If you change three things in a week and ACOS moves, you will never know which change did it. A weekly routine is plenty for most books. Checking every day mostly produces changes you then have to undo.
How to tell whether it is working
| Question | The number that answers it |
|---|---|
| Is this keyword worth its spend? | Its ACOS against your break-even, once it has enough clicks |
| Is the campaign profitable? | Campaign ACOS against break-even, over at least 30 days |
| Are ads growing the book? | TACOS — ad spend ÷ all sales — falling over several months |
| Is the book page the problem? | Plenty of clicks, very few orders, across every keyword |
The difference between ACOS and TACOS is explained in ACOS vs ROAS vs TACOS.
Common questions
How do I advertise my book on Amazon?
In your KDP Bookshelf, choose Promote and Advertise for the book, pick a marketplace and create a Sponsored Products campaign. Start with automatic targeting and a modest daily budget, let it run two to three weeks, then build a manual campaign from the searches that sold.
Can I advertise my book on Kindle devices?
Yes. Sponsored Products ads for books can appear on Amazon, on Kindle devices and on Goodreads, as long as Amazon is the featured offer for the book. You set them up the same way, from Promote and Advertise in your KDP Bookshelf.
How much does it cost to advertise a book on Amazon?
There is no fee to start and no minimum contract. You pay per click, usually somewhere between $0.20 and $0.80 for book keywords, and your daily budget caps what you can spend. What matters more is the cost of a sale: 20 clicks at $0.40 is $8 of spend for one sale at a 5% conversion rate.
Can I advertise my book on Amazon for free?
Not with ads — every click is paid. If your eBook is in KDP Select you can run free-book promotion days or a Kindle Countdown Deal, which cost nothing but a lower price for a few days. They are promotions, not advertising, and they work well alongside it.
How long does it take for Amazon book ads to work?
Expect two to three weeks before the numbers mean much for a new campaign. Sales are credited to clicks from earlier days, and individual keywords need dozens of clicks before a lack of sales is evidence of anything.
Should I advertise the eBook or the paperback?
Advertise the format that leaves more room under its break-even ACOS — usually the eBook, because a 70% royalty leaves far more room than a paperback that earns a few dollars after printing. Advertise the paperback on its own when it sells well by itself, as gift books, colouring and puzzle books often do.
Do I need a tool to run Amazon book ads?
Not to start. One or two campaigns are easy to run in the Amazon Ads console. Tools earn their keep once you have several books, hundreds of keywords, or no time to check bids every week.
- Work out break-even ACOS — royalty ÷ price — before you spend anything. It sets every other limit.
- Start with one Sponsored Products campaign on automatic targeting, dynamic bids down only, and a budget you could lose for three weeks.
- Bid from your royalty: break-even bid = royalty × conversion rate. Start at about two-thirds of it.
- Leave the campaign alone for two to three weeks; early numbers under-count sales and over-react to luck.
- Move searches that sell into exact-match manual campaigns and negate them in the automatic campaign.
- Change one thing at a time, and judge each change only on data from after it.