Some files shrink dramatically and others not at all. Where the difference comes from.
Compression reduces size by representing repeated material briefly. Files where the same content recurs often shrink the most. Plain text documents, spreadsheets and program source hold a lot of repetition and compress well, while photos and video are already compressed and have little repetition left, so they stay the same size.
There are two kinds. Lossless returns exactly the original when unpacked; lossy discards information people rarely notice to reach far smaller sizes. Documents and programs cannot tolerate a single changed character, so they always use lossless.
The other purpose of compression is bundling rather than size. Sending hundreds of files individually is slow because each is handled separately, while one bundle transfers faster and preserves the folder structure. With very many small files, a size reduction appears as well.
Archives can be password protected, but the strength varies widely by format and settings. Older schemes are easily broken, and file names sometimes remain visible even when contents are protected. For genuinely sensitive material, relying on an archive password alone is unwise.
With archives from unknown sources, the danger is running what is inside rather than unpacking itself. Files that look like documents can be executables, so check the full extension. Unpacking also spills the original folder structure and can overwrite existing files, so extracting into an empty folder is the safer habit.
Each button runs this keyword on that search engine