No formulas, no formatting, no second sheet: just rows of text separated by commas. Click any cell to see exactly what it becomes, and choose the format the receiving system expects.
We converted one source document into every major output format and measured the file that came out. The source was a 37.4 KB DOCX with headings, bold and italic text, a table, numbered and bulleted lists, and one embedded image.
Sizes are from one representative document converted with Pandoc and LibreOffice. Your own files will differ, especially with more images, but the relative ordering holds.
A PDF has no paragraphs. It stores “draw these glyphs at x, y in this font”, often one word or one line at a time, and a scanned page stores only a picture. Drop any PDF to see which kind you have and what its page instructions look like.
The comma is the column separator, so “Harbour St, Leeds” would split into two columns. The rule is to wrap such values in double quotes, and double any quotes inside. Type your own value to see how it's written.
Harbour St, Leeds"Harbour St, Leeds"Excel shows one thing and stores another. A CSV can only hold one of them, so each kind of cell follows a rule.
Double-click a CSV and Excel guesses each column's type. Long ID numbers lose digits, codes with leading zeros lose them, and anything that looks like a date becomes one. Save it again and the damage is written to the file.
400321887762error4.00321E+1100417error4173-4error03-AprMARCH1error01-MarCSV is for moving data between systems. For anything a person will read, these fit better.