Converting PDF to Word without wrecking the layout

PDF and Word are solving different problems. A PDF records exactly where every glyph sits on the page. A Word document records a flow of paragraphs that reflows as you type. Converting between them means guessing at a structure the PDF never stored, which is why results vary so much between documents.

What converts cleanly

• Single-column text documents — reports, letters, articles.
• Simple tables with clear ruled lines.
• Documents generated from Word or Google Docs in the first place.

What usually needs fixing afterwards

• Multi-column magazine layouts, where reading order has to be inferred.
• Tables without borders, which often arrive as loose text boxes.
• Heavy use of custom fonts, which get substituted.
• Scanned pages, which contain an image rather than text.

Scanned documents

A scan is a photograph of a page. Without text recognition there is no text to extract, so the conversion produces a Word file containing an image. Layers does not run text recognition, so for scans the honest advice is to use a dedicated recognition tool or retype the section you need.

A practical workflow

1. Convert to Word and open the result before you do anything else.
2. Fix the headings first — correct heading styles make everything after them easier.
3. Rebuild tables that arrived as text boxes rather than fighting the imported ones.
4. Convert back to PDF from the same workspace when you are finished editing.