Facebook and Instagram ads do one thing Amazon Ads cannot: put your book in front of a reader who was not looking for a book. That is the whole case for them, and the whole risk. A reader mid-scroll has no buying intent, so the ad has to create it, and most of the money authors lose on Meta goes on ads that ask a cold reader to buy on the spot.
This is the version of this guide I would want to read before spending my first £100: what to set up, what to send the click to, what it costs in pounds, and how to find out whether it worked. Where a figure comes from a vendor’s page or a named survey, it says so with the date. Where I could not verify something on the vendor’s own page, it says that too.
Rebuilt and reviewed by Robert Prime, 16 September 2026.
What the surveys say, before the advice
Two sources worth more than any guru’s screenshot:
- BookBub surveyed 200 authors who advertise (page updated 2 March 2026): 82% agreed they could reach a large audience of readers with Meta ads, against 76% for Amazon Ads. Reach, note, not profit. BookBub publishes no profitability figure.
- Written Word Media’s 2025 survey of 1,346 indie authors puts Facebook ads in its “mixed, challenging, but can work” group, with the note “for those who learn the system”, and lists strategic Facebook ads among the tactics its highest earners use.
Read together: the audience is there, the return depends on doing it properly, and the people making it pay are the ones who treat it as a system rather than a button.
The three decisions that matter
1. The objective
Meta’s Ads Manager makes you choose one of six campaign objectives: Awareness, Traffic, Engagement, Leads, App promotion and Sales. The objective tells Meta’s system what to optimise for, so it decides who sees the ad. For authors only two are worth using:
- Sales when you send readers to a page where a purchase or a sign-up can be measured (your own store, a BookFunnel page, or an Amazon Attribution link, covered below).
- Leads when the ad’s job is an email address, using either a landing page or Meta’s own instant form.
Traffic buys clicks from people who click on anything. Engagement buys likes. Neither pays for a book. And the blue “Boost post” button under a Facebook post is an Engagement campaign in disguise; use Ads Manager.
2. Where the click lands
Three destinations, in order of how well they usually pay:
- A reader-magnet page. A free novella or first-in-series delivered by BookFunnel in exchange for an email. You own the address forever, and every later book sells to that list for nothing. The reader magnet guide covers the set-up.
- A wide store or your own shop. Apple, Kobo or a direct store, where you can see the sale and Amazon Ads cannot reach.
- Amazon, through an Amazon Attribution link. Fine for a 99p first-in-series or a launch week, as long as you use the tracking link and not a bare Amazon URL.
Sending cold traffic to a bare Amazon link is the mistake behind most “Facebook ads do not work” posts. You cannot see which ad sold what, so you cannot learn.
3. The creative
The cover, shown large, does most of the work. A short vertical video of the cover with three lines of on-screen text does more. The copy leads with the reader, not the author: “If you read slow-burn fantasy romance, this one is 99p this week” beats any sentence that starts with “Award-winning”. Test four or five images and hooks at once and switch off the losers after five days; the winner is rarely the one you would have picked.
Tracking Amazon sales: Amazon Attribution
Amazon Ads’ own announcement, Amazon Attribution launches for Kindle Direct Publishing authors, is dated 30 September 2022. It lets KDP authors measure “their non-Amazon ads and understand the impact to sales in Amazon’s store”, and see “the total number of pages read” and “the Royalty amount earned” from those ads, through the advertising console.
In practice: create an Attribution tag for each Facebook ad, use that link as the ad’s destination, and the console shows you sales, Kindle Unlimited pages read and royalties per ad. It is the only way to attach an Amazon sale to a Meta ad. Without it, you are guessing from the sales graph. Amazon’s page does not state that it is free, and I could not find a charge anywhere in the console, so treat it as free with that caveat.
What it costs, in pounds
My starting point is £10 to £20 a day per ad set, for five days, before judging anything. That is £50 to £100 per test, and one test is one audience with four or five creatives. Two tests a month is £100 to £200 a month, which is close to the $250 a month Written Word Media reports its successful authors spending on ads.
What to measure:
| Destination | The number | What good looks like |
|---|---|---|
| Reader-magnet page | Cost per email address | Under £1 for fiction is common; over £2 and the page or the offer is wrong |
| Amazon via Attribution | Cost per sale against royalty per sale | A 99p book pays 35p, so a direct profit is unlikely; judge on pages read and reviews |
| Wide store | Cost per sale against royalty per sale | Break-even on book one is fine if there is a series behind it |
What I could not verify on Meta’s own page: several ad-industry sites state Meta’s technical minimum as about $1 a day for ad sets billed on impressions and about $5 a day for ad sets optimised for clicks or conversions. Meta’s help page on minimum budgets would not load without JavaScript on 16 September 2026, so I have not confirmed those figures. They do not change the advice: at £1 a day the system cannot learn, and you learn nothing either.
The VAT point most author guides miss
Meta invoices UK advertisers from Ireland. UK accountancy firms describe the position the same way: a UK VAT-registered business that gives Meta its VAT number receives an invoice with no VAT and accounts for it under the reverse charge on its own return; an advertiser who has not given a VAT number may be charged VAT. If you are registered, put the number in Ads Manager before the first campaign. If you are not registered, budget for the VAT on top. I have not read HMRC’s own guidance for this page, so check with your accountant; the self-publishing tax guide covers the wider picture.
When not to run them
- One standalone book, no list, no series. The economics do not close. Build the list with a free book first.
- No page to land on. Set up the reader-magnet page before the ad, not after.
- A launch week you could fill with your list and one promo newsletter. The newsletter promo sites hit readers in buying mode; Facebook interrupts them. For a pure sales spike, promos are usually the better first pound.
The authors who win treat Facebook as list-building first and selling second, then let email and series read-through do the earning. If you want the whole system laid out for your own book, that is what the Sell More Books course is for; my company, my course, weigh accordingly.
Frequently asked questions
Are Facebook ads better than Amazon Ads for authors?
Different jobs. Amazon Ads catch readers already shopping; Facebook ads reach readers who are not. In BookBub’s 2026 survey more authors said they could reach a large audience on Meta (82%) than on Amazon (76%), but neither survey publishes a profit comparison.
What budget should I start with?
£10 to £20 a day per ad set for five days, then keep only what hits your cost-per-email or cost-per-sale target.
How do I know a Facebook ad sold a book on Amazon?
Use an Amazon Attribution link as the ad’s destination. Amazon Ads opened Attribution to KDP authors on 30 September 2022 and it reports sales, pages read and royalties per link.
Why is my Facebook ad not selling books?
Almost always because a cold reader was sent straight to a buy button. Send them to a free book and an email sign-up instead, and sell to the list.
Sources, read 16 September 2026
- Amazon Ads, Amazon Attribution launches for KDP authors, 30 September 2022.
- BookBub Insights, BookBub vs Facebook vs Amazon ads: authors weigh in, updated 2 March 2026.
- Written Word Media, 2025 indie author survey, 1,346 authors.
- Lanop Accountants, VAT on Facebook ads for UK businesses, the reverse-charge position.
- Meta, Best practices for minimum budgets: the page exists but its body would not load for me; figures above are marked unverified.
