2026-07-03 · all guides

The Thumbnail Test: Does Your Book Cover Work at 100 Pixels?

Your cover is a thumbnail first and a cover second

Every path a shopper takes to your book — search results, category pages, also-bought carousels, ads — renders your cover at thumbnail size, often around 100 to 160 pixels tall on desktop and barely larger on a phone. The full-size cover only appears after a shopper has already clicked, which means the thumbnail did the selling. Designing a cover on a full monitor and never checking it small is designing for a context your buyers will see last, if at all.

KDP recommends uploading ebook covers at 2560 by 1600 pixels with a 1.6 to 1 height-to-width ratio, and that resolution matters for the product page zoom. But the strategic image is the same file scaled down thirtyfold. The two views reward opposite instincts: detail and texture reward the zoom, while bold shapes and contrast win the thumbnail. When forced to choose, choose the thumbnail.

How to actually run the test

The test takes two minutes. Scale your cover to about 100 pixels tall and place it in a screenshot of real Amazon search results for your target keyword, surrounded by the actual competition. Then ask three questions. Is the title readable, or at least is the genre unmistakable? Does the cover visually separate from its neighbors, or does it blend into the row? And would a stranger know what kind of book this is with no other information?

Harden the test further: view the composite at arm length on your phone, and squint until the image blurs. Covers that survive squinting share traits: one dominant focal element rather than a collage, title type occupying a third or more of the height, and a strong value contrast between text and background rather than a mere hue difference. If your title disappears when the image blurs, it disappears in search results too.

What thumbnail-proof covers have in common

Genre-appropriate simplicity wins. Nonfiction thumbnails that convert tend toward big type on high-contrast backgrounds with a single supporting graphic; think of the bestselling habit, finance, and productivity covers, which are practically typographic posters. Fiction relies on instantly recognizable genre cues: mood, palette, and imagery that say thriller or cozy romance before a single word is read. In both cases, subtitle text is expected to be unreadable at thumbnail size, which is fine; it just cannot be structurally important to the design.

The most common thumbnail killers are thin or script fonts at small weights, low-contrast palettes like dark gray text on navy, busy full-scene artwork with no focal point, and white backgrounds with no border that dissolve into the white Amazon page. If you generate covers with a tool — ebookdone produces four cover options with every 9 dollar book — run each candidate through the same shrink test and pick the one that wins at 100 pixels, not the one that looks best full-screen.

Testing against the market, not your taste

The final arbiter is not whether you like the cover; it is whether it wins its row. Build the comparison screenshot against the actual first page of results for your main keyword, because that is the exact company your thumbnail will keep. A cover that would look striking in a bookstore can still lose on a results page where every competitor uses the same palette, and a plain cover can dominate a row of cluttered ones.

For a cheap external signal, show two candidate thumbnails to a handful of people in your target audience for five seconds each and ask which book they would click and what they think it is about. Wrong guesses about the topic are cover failures, not audience failures. And after launch, watch your click-through data if you run any ads: the thumbnail is the single biggest lever on ad click-through rate, and swapping a weak cover is one of the few changes that can double it overnight.

FAQ