Decide what you actually need
Plain text is the right target when the content will be searched, fed into another program, or rewritten. It is the wrong target when the visual layout carries meaning, as in forms and invoices.
If you need structure preserved, exporting to a word processor format or to structured data keeps headings and tables that plain text discards.
Run the conversion
Upload the PDF, confirm whether it is text-layer or scanned, and run the conversion. Save the raw output before editing so you can return to it.
For long documents, convert in sections that match the document structure, such as by chapter. Errors are easier to locate that way.
Clean the predictable artefacts
Rejoin hyphenated line breaks: a search for a hyphen followed by a line break catches most of them.
Delete repeated running headers and page numbers with a single find-and-replace per pattern.
Collapse the double line breaks that appear where the original had column or page boundaries, then reinstate real paragraph breaks by reading.
Preserve the structure you care about
Mark headings with a consistent prefix as you clean, so the outline survives even in a flat file.
Keep tables in a delimited form such as comma or tab separated values rather than trying to preserve visual alignment with spaces.