A scanned PDF contains no text, only images. Running OCR first gives the file a real text layer, after which PDF to Word can produce an editable document instead of a picture.
In practice: (1) Step 1: run OCR PDF to add the invisible text layer. (2) Step 2: convert the OCR'd file to Word. (3) Accuracy depends on scan quality — 300 DPI is a good starting point.
Frequently asked questions
Why is my Word file one big image?
The source had no text layer. OCR first, then convert.
How accurate is OCR on handwriting?
Printed text is reliable; handwriting is hit-or-miss even with the best engines.
Do it in the full tool
Open PDF to Word for all options, or jump straight to another step below.