Every conversion needs some clean-up. Done in the right order it takes about ten minutes; done in the wrong order you redo work, because each step changes what the next one has to deal with.
First: decide whether to repair or reconvert
Click into the middle of a paragraph and press End. If the cursor stops at the end of the visual line rather than the paragraph, you have one-paragraph-per-line, or worse, text boxes.
Text boxes are not worth repairing. There is no reliable way to turn a canvas of boxes back into flowing text; you would be retyping with extra steps. Reconvert with something that produces real paragraphs and start from there. Why converters do this at all is covered in the text box guide.
1. Remove the stray line breaks
If every line is its own paragraph, fix that before anything else, styles applied first would have to be reapplied afterwards.
Use Find and Replace with wildcards off. Replace ^p^p with a placeholder such as @@, then replace ^p with a single space, then replace @@ back with ^p. That protects the genuine paragraph breaks while removing the ones the PDF invented.
Then search for - to catch words hyphenated across a line break and rejoined with a stray hyphen in the middle.
2. Apply heading styles
Not bold and a bigger font, the actual Heading 1, Heading 2 styles. This is what makes the navigation pane work, makes a table of contents possible, and keeps the document consistent when you change anything later.
Fastest route: open the navigation pane (Ctrl+F) so you can see the outline appear as you go, then click each heading and apply the style. On a long document, select one heading, use Select All Text With Similar Formatting from the right-click menu, and style them all at once.
3. Fix the fonts
Conversions often carry substituted fonts, sometimes several per paragraph, which is why documents look subtly wrong in a way that is hard to place.
Select all (Ctrl+A) and set one body font. If you have applied heading styles first, modify the styles rather than the text, so future edits stay consistent. This is why step 2 comes before step 3.
4. Rebuild the lists
Bullets frequently arrive as literal characters typed at the start of a line rather than as a list. They look the same and behave completely differently, no automatic renumbering, no proper indentation, no continuation.
Delete the typed bullet characters, select the lines, and apply a real list style. For numbered lists this matters more: a typed "3." stays 3 forever when you insert an item above it.
5. Check the tables
Tables are the least reliable part of any conversion, because a PDF does not record that cells belong to a table, it records text at coordinates and lines drawn on a page. Expect merged cells to be wrong and column widths to be arbitrary.
For a small table, fixing it is quicker than anything clever. For a large one, or a document that is mostly tables, you probably want a spreadsheet rather than a Word file, converting the tables separately and pasting them in usually beats repairing them in place.
Two things to check at the end
Read the first and last page against the original. Conversions sometimes drop a page, or read a header as body text. That is invisible in the middle of a long document and obvious at the edges.
Turn on formatting marks (Ctrl+Shift+8) and scroll through quickly. Stray paragraph marks, tabs used as indentation and manual line breaks all show up immediately, and all cause trouble later.
Saving the work next time
Most of steps 1, 2 and 4 exist because the converter did not do them. A converter that rejoins paragraphs, applies real heading styles and emits proper lists removes most of this list before you start.
That is what the converter here does, and it reports how many lines it joined so you can see whether step 1 is still needed.