Advertising a book, honestly
Nobody writing this down for you has anything to sell you. That matters, because most advertising advice authors read is written by people who run ads for a living, and it is shaped by that.
Every ad system works the same way underneath. You pay to put your book in front of people who were looking at something else. A few of them look at it. A smaller number click. A fraction of those buy.
That last step is the one the seller does not control, and the one that decides everything. The platform sells you the arrival. What happens after the click is decided entirely by your page: the cover at the size it actually appears, the blurb, the sample, the price, the reviews, and whether the person who arrived wanted this kind of book in the first place.
So an ad cannot fix a bad cover. It cannot rescue a blurb that summarizes the plot instead of selling the book, and it cannot place you in front of the right readers if the book is filed in the wrong category. An ad buys more people arriving at whatever is already there. If your page converts badly, advertising buys the same bad conversion in greater volume and charges you for the privilege.
The order of operations
The cover. It is the ad, in most formats. If it disappears at thumbnail size, you are paying to show people a small gray rectangle.
The blurb. The click is a question and the blurb is the answer. A reader who arrives interested and leaves unconvinced cost you money and told you nothing you can act on.
Categories and keywords. These decide who the system thinks the book is for, which decides who your ads are shown to and what the store recommends afterwards. Advertising a miscategorized book means buying the wrong audience at full price.
The price. A price that sits wrong for the format and length of your book turns interested readers into people who bookmark it and forget.
Then advertise. Only then. Each of the four above is free to change and affects every sale you will ever make. Advertising is the one lever that costs money every time you pull it, and it multiplies whatever the other four are already doing — in both directions.
In the Cover Maker. The thumbnail check shows your cover at the size an ad or a search result actually displays it, which is the size that decides whether a paid impression turns into a click. The ebook cover exports at 1600 × 2560 for the listing itself. Start a book → · about the Cover Maker
Measurement is less reliable than the dashboard suggests
Attribution is approximate. A reader sees your ad on a phone at lunch, thinks about it, and buys that evening on a tablet — or in a different store, or in paperback instead of ebook, or a month later after a library hold. Some of that gets credited to the ad. Much of it never does. Reporting undercounts, and it does not tell you by how much.
The first sale is usually not where the money is. For a series, what matters is read-through: how many people who buy book one go on to buy books two and three. A campaign that loses money on the first book and pulls a steady share of those readers through the rest of the series can be profitable while its own report shows a loss.
Work out your read-through before you spend. Compare units sold of book two against book one over the same period, with no advertising running. That ratio is the only thing that tells you what a first-book sale is actually worth to you. Without it, you are guessing in whichever direction suits your mood.
Where the money goes
A standalone with nothing after it. One book, no series, no next release. The only value of each sale is that sale. This is the hardest case to make pay, and it is the one most beginners start with.
Bidding on famous author names. It feels clever and it attracts the wrong reader — someone hunting for a specific writer, who sees a book that is not by them. You pay for that click every time.
Advertising a book with no reviews. You are buying traffic to a page that gives the reader a reason to hesitate. Get reviews first; they are cheaper than clicks.
Continuing because stopping feels like giving up. The most expensive mistake, and the least discussed. A campaign running for four months because turning it off would be an admission is not a marketing strategy.
A sane way to start
A small budget. Small enough that losing all of it changes nothing about your year.
A fixed period. Two weeks, four weeks — a date on which the test ends whatever it is doing.
One variable at a time. Change the image or the copy or the audience, never two together. Two changes give you a result you cannot read.
A decision rule, written down first. Before you spend anything, write the sentence: "If the result is X, I will do Y." X can be units, or a cost per sale you are willing to pay, or read-through into book two. The rule exists so that the decision is made by the person you were before you had money in the game, not the one watching a dashboard at midnight.
Advertising rewards a catalog and punishes a single book. If you have one book out, the cheapest and most reliable campaign available to you is writing the second one.
Make yours
Everything here is built into the tools, and they are free to use: name your book, bring your manuscript or your artwork, and work as long as you like. You only pay when you want the files without the preview watermark.
Next: Promotion after launch week · More: Categories and keywords, without the folklore · How to write a blurb that sells the book · The thumbnail test