Why your Word document changes when you save it as a PDF

You finish a report, export it to PDF, open the result, and something has moved. A heading that sat neatly at the top of page four is now stranded at the bottom of page three. A table has split. The typeface is subtly wrong. Occasionally something worse appears: a colleague's comment, or tracked changes you thought you had accepted.

None of this is random, and all of it is preventable once you know what conversion actually does.

Word lays out for a printer, not for a page

This is the root of most surprises. Word calculates line and page breaks partly from the capabilities of your currently selected printer, including its margins and paper size. Change the printer and the layout can shift, and exporting to PDF is effectively selecting a different output device.

The practical result is that a document composed against one printer driver may repaginate when exported, moving everything after the point where the difference first bites. It is why a document can look correct to you and different to a colleague opening the same file.

Setting the page size explicitly, and keeping margins comfortably inside what any printer can manage, removes most of this. Documents laid out with almost no margin are the ones that shift most.

Fonts, again

If a typeface is not embedded in the exported file, readers substitute. Substitutes have different letter widths, which changes where lines wrap, which changes where pages break. One missing font can cascade through an entire document.

Most export dialogs have an option to embed fonts, and it is not always switched on. It is worth finding once. Some licensed fonts refuse embedding, which is worth knowing before you build a hundred-page report around one.

The things that should not have been exported

This is the category that causes actual harm rather than inconvenience.

• Tracked changes still present, showing every edit and who made it
• Comments in the margin, sometimes including frank remarks about the recipient
• Hidden text that was never intended for the reader
• Document properties carrying the author name, or the name of whoever owned the template
• Fields such as dates that update to today rather than staying as written

Tracked changes and comments deserve particular care, because Word can display a document as if changes were accepted while still storing them. Viewing the final version is not the same as accepting the edits. Accept or reject them properly, delete the comments, and only then export.

A checklist before exporting

1. Accept or reject all tracked changes, and delete every comment.
2. Check document properties and clear the author field if it should not travel.
3. Set page size and margins explicitly rather than relying on defaults.
4. Turn on font embedding in the export options.
5. Export, then read the PDF itself rather than assuming it matches.

That last step is the one people skip under time pressure, and it is the only one that actually catches problems. Page through the PDF, look at where the breaks fell, and check the first and last pages carefully.

When the layout still will not behave

If a table keeps splitting awkwardly, the fix belongs in Word rather than in the PDF: keep-with-next settings on headings, and repeat-header-row on tables, control this properly and survive conversion.

And if you are working from a PDF someone else produced and need to edit it, converting back to Word will reconstruct the layout rather than preserve it. Expect to spend time tidying, particularly with multi-column pages and complex tables, because the conversion is inferring structure that the PDF never really stored.