Keep headings, links, tables, and reading order intact

Word-to-PDF conversion sounds like a one-way trip: open the document, make a PDF, send it. The problem is that a word-processing file is more than printed pages. It has heading styles, lists, table headers, image descriptions, hyperlinks, comments, tracked edits, page settings, and a reading order that may matter to a screen reader.

If you only print the document to PDF, you may get something that looks right at first glance but behaves like paper behind glass. A proper export aims to carry the structure across as well as the appearance. That difference matters for long reports, policies, submissions, contracts, and anything that must be reviewed or archived.

Structure starts in the Word document

A converter cannot reliably invent structure that the source file never had. If the title is just large bold text, it may look like a heading to a person but not to the export engine. If a table header row is formatted by hand, the PDF may not know that those cells label the rows below. If an image has no alt text, the PDF cannot provide a useful description later.

Before conversion, check the document itself:

• Use real heading styles for sections and subsections, not only larger text.
• Apply list formatting for bullets and numbers rather than typed hyphens.
• Mark header rows in tables so they can repeat and be identified.
• Add concise alt text to informative images, charts, and diagrams.
• Use descriptive hyperlink text rather than a bare pasted URL when possible.
• Remove empty paragraphs that were used only to force spacing.

Those details are not cosmetic. They are the clues the exporter uses to build a tagged PDF, a bookmark outline, and a reading sequence.

Where each export route actually takes you

"Make a PDF" hides three different mechanisms, and they do not preserve the same things.

A word processor's own Export or Save as PDF command has access to the document model. That is the only route that can map Heading 1 and Heading 2 to a bookmark outline, write the tags assistive technology relies on, carry document properties, and keep cross-references live. If tagging and bookmarks are a requirement, this is the route to use.

A print driver such as "Microsoft Print to PDF" is designed to imitate a printer. It receives rendered pages and draws them into a PDF. Text usually stays selectable, but tags, bookmarks, alt text, and the logical reading order are generally not carried through, because the driver was never given them.

A browser-based route sits in between. The browser-local Word to PDF tool renders the DOCX in the page for review and then produces the PDF through your browser's own print path, which in Chromium normally keeps text searchable and hyperlinks live. It also offers an image-only export as a fallback for documents that will not render cleanly; that fallback produces a picture of each page with no text layer at all, so use it only when the alternative is nothing. What this route does not do is build a full tag tree, so treat it as a good way to produce a readable, searchable, private PDF rather than as an accessibility-conformant publishing pipeline.

The difference is easiest to see in a 30-page report. With a tagged export from the word processor, the PDF side panel shows bookmarks for each major heading and a reviewer can jump from "Method" to "Findings" without scrolling. With a print-style PDF, the same reviewer has thumbnails and search instead.

The privacy trade-off is the reason the browser route exists at all. Most web-based converters ask you to upload the document first, which is an odd thing to do with an unsigned contract, an appraisal, or a manuscript. Rendering the DOCX in the tab keeps the file on your machine.

Headings, bookmarks, and reading order

Headings do two jobs in a PDF. Visually, they divide the page. Structurally, they tell software how the document is organised. When heading styles are used consistently, a PDF can expose them as bookmarks or an outline. That makes long documents easier to navigate and easier to audit.

Reading order is related but not identical. A two-column newsletter, a sidebar, or a floating image can look obvious on screen while confusing the sequence in which text is read aloud. If your document uses text boxes, columns, wrapped images, or complex tables, inspect the exported PDF rather than assuming the visual order is the machine order.

Tables need special care. Header rows should be marked as headers in the source document, not merely bolded. Merged cells, nested tables, and tables used for layout can make the reading order harder to preserve. For important forms or policy tables, keep the structure simple: one header row, clear column labels, and no hidden spacer cells.

Fonts, page size, and links affect trust

Fonts decide line breaks. If the PDF cannot embed the document's font, the viewer may substitute another face. That can move page breaks, change table widths, or push a signature line onto the next page. For legal, academic, or brand-sensitive documents, open the exported PDF and compare the first few pages, a dense table, and the final page against the source.

Page size is just as practical. A document authored for A4 and exported or printed as Letter may gain different margins or pagination. The reverse is also true. Set the intended page size before converting, especially when the PDF will be printed, filed, or signed.

Hyperlinks should stay live in a proper export. Test the table of contents, cross-references, footnote links, and external links in the finished PDF. If a link must work on paper, include enough visible text for the reader to identify the destination even without clicking.

If the PDF needs new running labels after conversion, use the headers and footers tool on the finished file rather than editing every page manually.

Comments and tracked changes can leak context

Tracked changes and comments are useful while drafting, but they are often not meant for the final recipient. Depending on export settings, a PDF can include markup, comment balloons, document properties, or a final-looking page that still carries metadata you did not intend to share.

Before export, decide whether the PDF should show the clean final document or a review copy with markup. Accept or reject changes when the document is final. Delete comments that were only for internal drafting. Then scan the PDF for leftover metadata or visible annotations. The privacy scanner is useful for a final pass when a file has moved through several reviewers.

This is not about paranoia; it is about matching the PDF to its audience. A lawyer reviewing redlines may need the comments. A client signing the final copy almost certainly does not.

A short conversion checklist

• Use heading styles before export so the PDF can gain bookmarks.
• Add alt text to meaningful images and mark table header rows correctly.
• Use the word processor's own Export or Save as PDF when tags and bookmarks matter; a print driver will not create them.
• Check whether the route you chose kept text selectable before you send the file; an image-only export has no text layer.
• Confirm A4 or Letter page size before you make the PDF.
• Test live links and internal cross-references in the exported file.
• Remove or intentionally include tracked changes and comments; do not leave that choice to default settings.