Your cover will be judged at eighty pixels wide

Look at where your cover is actually seen.

A phone. Four books across, in a grid, in a column that is moving. Your paperback is the width of a fingernail. The person scrolling is not looking for you, has never heard of you, and will spend somewhere under a second deciding whether their eye stops.

That's the venue. Not the mock-up on your laptop at full size, with the file open, by the person who wrote the book.

You are not bad at judging your own cover. You are judging it in conditions no reader will ever share.

You see it big. You see it alone, not surrounded by competitors. You see it after two years inside the book, which means you cannot look at the image without the whole manuscript arriving behind it. When you look at a figure walking away down a road, you see chapter eleven. A stranger sees a person walking away down a road, and it costs them nothing to keep scrolling.

That's the same blindness that stops you catching your own typos, moved to a different surface. Your eye supplies what it already knows.

Everything else follows from that one fact, and it explains design choices that look arbitrary from outside.

Legibility beats illustration. A cover that can't be read at thumbnail hasn't failed at being beautiful. It has failed at being a cover. The title is the only part that survives the size reduction intact, so the title is the design.

Contrast beats detail. Detail is the first casualty of scale. A photograph with a wide tonal range collapses into grey mud at thumbnail while a flat two-colour block stays exactly as strong as it was. This is why so many good covers look almost crude at full size — they were built for the size they'll be seen at, not the size they were made at.

Three elements beat seven. At a fingernail's width the eye resolves one shape, one word, one colour relationship. Everything past that is noise you paid a designer to add.

And it has to hold up next to any book on a shelf. That's the real test, because it is the only condition under which your cover will ever be seen — never alone, always in company, always being compared.

There's a second job running underneath the first, and it decides more sales than beauty does.

The reader is answering one question: is this for me. They answer it by comparison against every book they have already bought and liked. Which means a cover that looks unlike its category doesn't read as original. It reads as amateur — the exact opposite of what the writer was reaching for.

This is the trap that catches serious books. You have written something that isn't like other books, so you want a cover that isn't like other covers, and the result tells a stranger you don't know what shelf you're on. Recognition has to come first, in about a fifth of a second, before anything interesting can happen.

The work is originality inside the convention. Unmistakably the same shelf. Unmistakably not the same book.

A designer will make something that's theirs and not mine. Real. A cover is the one part of the book somebody else authors, and it goes on the front with your name under it.

I'll get a template three hundred other books are using. Also real, and it's what the cheap end of the market produces — a good-looking image that a reader has already seen twice this month.

Both fears have the same answer, and it isn't reassurance. You see it before it ships, and nothing goes out on a placeholder. If it isn't right, it isn't up. An unfinished thing left visibly unfinished is worth more than a filler, on a cover exactly as much as anywhere else on this site.

A serif that closes up at small sizes and turns your title into a smear.

A subtitle no human will ever read at the size it appears — set at a scale that only exists in the design file.

A beautiful photograph that becomes grey at thumbnail, so the whole cover reads as a blank rectangle in the grid.

A square social crop that beheads your own name, because the preview was never checked.

A cover that works in light mode and disappears in dark.

A spine width that doesn't match the trim, discovered at the proof, when it costs something to discover.

Every one of those looks fine full-size on a desktop monitor, which is where it was approved.

Digital: front and title layout, sized for the store page, the reader, and every social preview it will ever appear in.

Print: spine and back, matched to the printer's trim.

Checked at thumbnail, because that's the only test that counts.

$30. Thirty minutes of our team's time — a cover built, checked at the size it will be seen at, on the page.

You can do all of it yourself, here, for nothing.

What the $30 buys is somebody looking at your cover who has not read your book. That's the only person whose opinion of it matches the reader's, and it is the one perspective you gave up permanently the day you started writing.

Send the file.