Scanning a document: JPG, PNG or PDF
Each format is good at something different. Picking wrongly costs you either quality or a file nobody can open properly.
JPG is built for photographs. It throws away detail the eye is unlikely to miss, which keeps files small and is exactly right for a scanned page, a photographed form or anything with shading and texture. The loss is permanent, so avoid saving the same JPG repeatedly.
PNG keeps every pixel. That makes it the right choice for screenshots, diagrams, logos and anything with sharp edges or flat colour, where JPG leaves visible smudging around the lines. The cost is size: a PNG of a photographed page can be several times larger than the JPG of the same page, with no visible benefit.
PDF is not really an image format at all, it is a container for pages. It is what you want when the thing being sent is a document: several pages in order, in one file, that opens the same way on every device and prints at a fixed size. Almost every form that asks for an upload asks for a PDF for exactly this reason.
In practice the useful path is to photograph or scan as JPG, combine the images into one PDF in page order, and compress if there is an upload limit. Going the other way, PDF to JPG gives you each page as an image when a site insists on a picture rather than a document.
Tools for this