Measuring your proof with a ruler
Reading a proof finds typos. It does not find a trim that came out a sixteenth short, a text block sitting higher on left-hand pages than right, or a gutter too tight for the book's thickness. Those are dimensions, and you find them with a ruler.
Get one with both inches and millimeters on it, ideally steel. Sit down with the proof and a pen. This takes about ten minutes, and it is the last ten minutes before you approve a print run.
The trim
Measure the width and the height of the closed book, across the front cover, not over the spine.
It should match the trim size you chose. It will not match it exactly, because the book was printed and then cut, and cutting has a tolerance — a small one, and it varies by printer. A fraction of a millimeter out is the blade.
A difference you can see without measuring is not the blade. A book that is visibly the wrong size was built to the wrong trim, and everything inside it — margins, text block, cover geometry — is built on that same wrong number.
The spine
Measure the thickness of the closed book across the spine, then work out what it should be.
On Amazon KDP, white stock bulks 0.002252 in per page and cream 0.0025 in. So 300 pages of white comes to about 0.676 in, and the same 300 pages on cream to 0.75 in. A KDP hardcover adds 0.188976 in — 4.8 mm — on top of the paperback bulk. Lulu is different again: the page count divided by 444, plus 0.06 in for glue. IngramSpark is a curve read off their own templates rather than a formula.
Differences of a few hundredths are not worth chasing. A gap of a tenth of an inch or more means the book in your hand is not the book the cover was built for: the page count moved, or you switched paper. Forty pages of white is about 0.09 in of spine on its own.
In the Cover Maker. It draws live trim, bleed and safe-area guides over your artwork, reserves the barcode area, and computes the spine from the real page count, paper color and printer — so a measurement that does not agree almost always means the interior changed after the cover was built. Start a book → · about the Cover Maker
The outside margin
Open to any full page of body text and measure from the outside edge of the page to the outside edge of the text.
You are not checking it against a rule here — you are checking it against itself. Measure the same margin on four or five pages spread through the book. They should all be the same. If they are not, the pages were cut unevenly, which is the trim tolerance again and normal within a small range.
Then ask whether the number you get is comfortable to hold. An outside margin thin enough that your thumb covers text is a setting to change, not a fault to report.
The gutter
Hold the book open, flat, without forcing it, and measure from the inside edge of the text to the point where the page disappears into the binding. Do this near the front, in the middle, and near the back. The middle is the worst case.
Then do the check that matters more than the number: with the book open and unforced, can you read the first characters of every line? If you have to press the spine down to reach them, the inside margin is too narrow for the thickness of the book.
The gutter is the total inside margin, and it scales with page count because a thicker book pulls more paper into the curve of the binding. A 300-page book builds a 0.5 in gutter; a 700-page book builds 0.75 in. KDP's own floor is 0.375 in and IngramSpark's is 0.5 in, so 0.5 is valid at both.
This is the measurement most likely to send you back to the manuscript, and widening it moves the page count, which moves the spine. Margins and gutters covers the setting itself.
The text block, left page against right
Open to a spread. Measure from the top of the page to the first line of text, and from the bottom of the page to the last line, on the left-hand page and then on the right-hand page.
The two should agree. A text block that sits higher on versos than on rectos is a layout error, not a cutting one, and it shows to a reader as a page that feels subtly restless without their being able to say why. Check two or three spreads, because one uneven cut can make a correct layout look wrong.
The barcode area
Measure the clear block on the back cover where the barcode sits or will sit. It is 2 × 1.2 in.
Confirm that nothing you care about — a line of the blurb, part of the artwork, a logo — falls inside it, and that the block sits where your printer publishes it. If the printer prints its own symbol there, that space has to be genuinely empty, whatever your design was doing in it.
Reading the results
Two rules cover almost every number you wrote down.
Small differences are the cut. A book is printed and then trimmed, and the blade lands close to the line rather than on it. Variation between copies, and between the top and the bottom of one copy, is the process working.
A large, consistent difference is the file. It repeats on every copy because it was built in. Find the number that is wrong — usually the page count or the paper — fix it at the source, rebuild both files, and proof again.
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: The proof checklist, front to back · More: Checking the spine and the wrap on a real copy · How to calculate spine width · Barcodes, and the price code beside them </content>