Converting a PDF to Word looks like a one-click operation, but the underlying mechanics matter — especially when the PDF is scanned, multi-column, or heavy on tables.
Text-PDFs vs scanned PDFs
A 'text PDF' has selectable text — its words are stored as Unicode strings with positions. A 'scanned PDF' is just images of pages and has no extractable text until OCR runs. The pdf-to-word tool handles the first; pdf-ocr handles the second.
What gets preserved
Plain paragraphs, lists, and headings transfer cleanly. Multi-column layouts often collapse into single-column flow. Tables drawn with vector lines may convert as plain text rather than Word tables.
A reliable two-step process
If the PDF is scanned, run pdf-ocr first to embed a text layer, then run pdf-to-word. This recovers structure that a single-pass converter would miss.