Compress images (reduce resolution to 150–300 DPI for screen use)
Convert images to efficient formats (JPEG for photos, PNG only when needed)
Reduce image quality where acceptable
Downsample embedded images
Remove unused embedded objects, fonts, and color profiles
Use “Optimize PDF” / “Reduce File Size” in your PDF tool
Flatten layers and transparency (or remove unnecessary layers)
Eliminate transparency effects when possible
Remove hidden content, annotations, and bookmarks not needed
Delete metadata (author, creation tools, timestamps) if not required
Avoid embedding full-page backgrounds or large raster images
Use vector graphics instead of raster when appropriate (and simplify complex vectors)
Simplify paths in vector artwork (reduce node counts)
Convert scanned PDFs with OCR efficiently (avoid embedding full-resolution scans if lower resolution suffices)
Use OCR text layers instead of embedding large page images when feasible
Re-export from the source at lower settings (preferably from the original document)
Disable “Embed all fonts” and instead subset fonts
Use standard fonts where possible
Merge PDFs carefully (recompress after merging if your tool doesn’t recompress)
Remove duplicate images and repeated resources if your tool supports it
Use PDF/A only if required (it can increase size)
Check for large attachments embedded in the PDF and remove them
Use a dedicated PDF optimizer tool and apply the strongest safe compression settings
Run a final size check and iterate on image/DPI settings to meet the target size