A Cover Thumbnail Is Its Own Reading Size

Make a small, separate proof

A cover approved at full size can fail when a retailer or reading app reduces it to a small rectangle. Create a thumbnail proof from the delivered master, not from an enlarged design mock-up. Put it beside plausible neighboring covers in a neutral grid. The question is not whether every decorative detail survives; it is whether the title, author name, and any necessary series distinction can still be found without guessing.

Set a reading order

Ask three ordinary questions in order: What is this book called? Who is it by? Which volume or edition is it? If the eye catches an illustration before the title, that is not automatically a fault. If the title becomes a texture, or a series number is confused with a price badge, flag it. A thumbnail check is a reader-retrieval check, not a demand for larger type everywhere.

Use a worked comparison

Imagine a dark blue cover with a fine silver title and a large pale bird. At full size the contrast feels quiet and deliberate. At thumbnail size, the bird remains but the silver letters merge with the background. First test a stronger title treatment or a less busy field; do not immediately add a sales claim. Re-export, compare at the same reduced size, and retain the rejected proof with a one-line reason.

Check the package as well

For EPUB, confirm that the intended cover image is identified in the package manifest and that any HTML cover description carries useful alternative text when the image conveys more than title and author. DAISY distinguishes the package cover from the HTML presentation: the former is commonly used in bookshelves, while the latter is where a description can serve the reader. Both deserve inspection.

Record exceptions honestly

A jacket designed around an intentionally unreadable mark may be an art decision, but record who accepted its retail consequence. Do not describe a cover as accessible merely because a package property exists, and do not claim retailer rendering was tested unless it was. Preserve the exported dimensions, filename, date, and comparison image with the release record. That makes a later replacement traceable rather than aesthetic amnesia.

Check whether the exported file has unexpectedly flattened or blurred lettering, especially after an automated size reduction. Compare the final file, not only the working artboard, against the approved proof. If a cover is revised, ask whether its alternative text, package reference, product image, and thumbnail asset all need replacement. A visible new image joined to stale metadata is still a mismatched edition record.