The file is re-saved with object streams enabled, which repacks the PDF’s internal structure more tightly. That is compression of the file format, not of your images: text-heavy documents, forms and exports with redundant object data usually shrink noticeably, while files that were already tightly encoded may barely move. The tool is upfront about the outcome either way. Scanned PDFs are essentially photographs of paper — the images inside them are already compressed, so meaningful reduction there means downsampling the images, which this structural pass deliberately does not attempt. No upload, no signup; encrypted files are reported rather than silently mangled.
When a portal enforces a hard KB limit and the structural saving is not enough, combine tools: convert the pages with PDF to JPG at a lower quality, then rebuild a smaller document from those images with Image to PDF. For mostly-digital PDFs, though, this one-click re-save is usually the right first move — try it before anything more involved. The original file is never modified either way: the optimized version always arrives as a fresh download, so you can compare results freely.
Also useful: PDF to JPG, Merge PDF, and Image Compressor.