Why tables do not come out perfect
A PDF does not contain a table. It contains text placed at coordinates and lines drawn on the page, so every converter has to guess where rows and columns begin and end.
That guess is reliable when the table has visible borders and real text. It becomes unreliable with merged cells, hidden grid lines, or pages that are scans.
Steps for a clean conversion
- Confirm the PDF is text-based: try selecting a number in the table with your cursor. If it highlights, the converter can read it.
- Convert only the pages that hold the table if the document mixes prose and tables.
- Open the .xlsx and compare row and column counts with the PDF before editing anything.
- Fix merged header cells and wrapped text by hand; these are the usual failure points.
Keep the PDF open side by side while you check the sheet.
What still breaks, and the fix
Merged cells split into several cells, and a cell with two lines of text can arrive as two rows. Numbers may lose leading zeros (007 becomes 7) or arrive as text.
Set the affected columns to Text format before pasting values back, and merge header cells manually. Formulas are never preserved, because a PDF does not store them.
Scanned PDFs need OCR first
If the page is an image, a scan or a photo, there is no text layer to extract, so conversion produces an empty or nearly empty sheet. Run OCR on the file first, then convert the searchable PDF.
The same applies to pages exported from some design tools where text was converted to outlines.