You convert a tidy PDF to Word and open something that looks vaguely like the original but behaves nothing like it. Type one word and the whole page shifts. Understanding why this happens tells you which documents will convert cleanly and which never will.
The two formats want opposite things
A PDF is a description of a finished page. It says: put this character at this exact coordinate, in this font, at this size. It is deliberately fixed — that is the entire point of the format, and why a PDF looks identical on every device.
A Word document is a description of structure. It says: this is a heading, this is a paragraph, this is a table with three columns. Word decides where things land when it renders them, which is what lets text reflow when you edit.
A converter sees a line of text sitting slightly above another, in a slightly larger font, and has to guess: is that a heading, or just a bold first line? Two columns of text, or a two-column table, or a table with invisible borders? Nothing in the PDF says. It has to be deduced from geometry alone.
What converts well, and what does not
| Document type | Expect |
|---|---|
| Straight prose, single column | Excellent — near perfect |
| Reports with clear headings | Good; check heading styles |
| Simple tables with visible borders | Usually good |
| Multi-column layouts | Mixed; column order can scramble |
| Borderless tables and forms | Often becomes loose text boxes |
| Magazine layouts, wrapped images | Poor |
| Scanned pages | Nothing at all — see below |
The scanned-document trap
If your PDF came from a scanner, a phone camera or a photocopier, it contains no text. It contains photographs of text. Converting it to Word produces a document with a picture on each page and not a single editable word, because there was never any text there to extract.
The test takes two seconds: open the PDF and try to select a sentence with your cursor. If you cannot highlight individual words, it is a scan.
The fix is OCR — optical character recognition reads the image and adds the recognised characters as a real text layer. Run OCR first, then convert. Doing it the other way round gives you nothing.
Why the result is full of text boxes
This is the most common complaint, and it is the converter choosing fidelity over editability. Faced with a layout it cannot confidently interpret, it can either reproduce the appearance exactly by pinning each block into a positioned frame, or guess at the structure and risk the page looking wrong.
Text boxes preserve the look but make editing miserable, because nothing reflows. If you got a box-heavy result, the practical move is usually to stop fighting it: select all, copy, and paste as plain text into a fresh document, then apply your own styles. You lose the original formatting but gain a document that actually behaves.
Getting a better conversion
- Use the original PDF, not one that has been printed and rescanned.
- Check it has real text by trying to select some. Run OCR first if not.
- Convert only the pages you need. Splitting out the relevant section reduces the number of layout puzzles the converter has to solve, and it is faster.
- Expect to fix headings and tables. Budget a few minutes of cleanup rather than expecting perfection.
- Ask whether you need Word at all. If you only want to change a date or cover a stale price, editing the PDF directly is far less work than converting, fixing and exporting back.
Common questions
Why is my converted document full of text boxes?
Because the converter could not confidently work out the structure, so it preserved the exact appearance instead. Copying everything and pasting as plain text into a new document is usually faster than untangling the boxes.
The fonts are wrong. Why?
The PDF embedded a font your computer does not have, so Word substituted the nearest match. Substituted fonts have different letter widths, which shifts line breaks and can cascade through the whole document.
My tables came out as a jumble.
Tables without visible borders are the hardest case — nothing in the file marks them as tables, only text that happens to line up. Tables with drawn borders convert far more reliably.
Can I convert back to PDF afterwards?
Yes, and Word exports PDF directly. Be aware that each round trip degrades the layout a little, so avoid converting back and forth repeatedly.