A three-page PDF has no business being 20 MB. But scan a few pages, export from a design tool, or save a slide deck as PDF, and that's exactly what you get — a file too big to email. The reason is almost always the images inside, and once you know that, shrinking it is easy.
where the weight actually is
A PDF is a container. Text and vector shapes are tiny. What blows up the size is embedded images, saved at far higher resolution than anyone will ever view. A scanner drops in 300 or 600 DPI pages. A design export embeds full-resolution photos. The PDF keeps every pixel, and the file balloons.
Fonts add a little. Duplicate images that never got deduplicated add a little. But 90% of the time, it's the pictures.
the fix: downsample the images
Compressing a PDF properly means re-sampling those embedded images down to a sane resolution and re-encoding them. For on-screen reading, 150 DPI is plenty and cuts the size hard. For print, stay nearer 300. Good compressors give you a level — screen, ebook, print — which is really just picking that target.
In practice a scanned document often drops 70–90%. A lean text PDF won't shrink much, and that's fine — there was nothing to cut.
a few things to watch
- Don't over-compress a file you'll print. Screen-level compression looks soft on paper. Match the level to how it'll be used.
- A compressor shouldn't make it bigger. If the “compressed” file comes out larger, the tool should hand back the original. Ours does.
- Watch for embedded junk. Some PDFs carry JavaScript or attachments you don't want to pass on. Re-writing the PDF cleanly strips that out — a nice side effect of compressing.
the quick way
Drop the file into a PDF compressor, pick Balanced for most cases (Strong if you need it as small as possible, High quality to keep sharp images), and download. A scanned contract that came in at 18 MB usually lands around 2–3 MB, still perfectly readable.
That's the whole trick: it was never really about the PDF, it was about the pictures inside it.