Is a Word file parsed differently from a PDF?

Yes, in some ways. A Word (.docx) file stores text as paragraphs with styles, in the order you typed it, so its main text usually has a clear order. Problems tend to come from content outside the main text, such as text boxes, shapes, and sometimes the page header and footer.

A PDF stores text with exact positions on the page. Its order depends on how it was exported, so simple documents read well, while columns or freely placed elements can come out in an unexpected order.

So each format has its own weak points. In both cases, a simple, single-column layout with everything in the main body is what makes parsing reliable. Test a PDF with the copy-and-paste test and a Word file by saving it as plain text.