How to make a PDF smaller (and when you can't)

You have a 24 MB PDF and a 10 MB attachment limit. Before reaching for a compressor, spend thirty seconds working out why the file is large — it decides whether compression can help at all.

What makes a PDF large

Almost always: images.

A page of typed text is a few kilobytes — some font data and character positions. A scanned page is a photograph, and at 600 dpi that’s several megabytes. So a 40-page text document might be 400 KB, while a 10-page scan is 30 MB.

In rough order of impact:

1. Scanned pages. Every page is an image. The usual culprit behind an enormous file.
2. Photographs at full resolution, placed on a page that displays them 8 cm wide.
3. Embedded fonts, especially when many typefaces are each embedded in full.
4. Structural bloat — revision history, orphaned objects and uncompressed cross-reference tables left behind by whatever exported the file.

Only that last category can be removed without losing anything.

Try lossless first

Lossless mode rewrites the document’s internal structure using object streams and drops objects nothing references. Text, fonts and images are untouched, and quality is identical.

On a file exported by careless software this can cut a real chunk. On an already-efficient file it saves almost nothing — and a compressor that reports a big saving on such a file is either lying or has quietly degraded something. Ours reports the real number, including when that number is zero.

When to go stronger

Strong mode draws each page as an image and rebuilds the PDF from those images. On a scanned document the saving can be substantial, because your scanner’s inefficient output gets replaced with well-compressed JPEG. The tool reports the actual before-and-after sizes so you can judge the result rather than trust a claim.

What you give up is real:

• Text stops being text. No selecting, searching or copying.
• Screen readers can’t read it, which matters if anyone you send it to relies on one.
• Fine detail softens, most visibly on thin lines and small type.
• The file can no longer be usefully converted to Word.

For a scan you’re emailing so someone can look at it, that’s a fair trade. For a contract someone needs to search, or anything that must stay accessible, it isn’t.

Three things that work better than compression

Remove pages. If you only need 10 pages of a 100-page report, splitting them out discards the other 90 entirely, with zero quality loss on the pages you keep. Often the best answer, and often overlooked.

Re-export from the source. If the PDF came from Word, re-export with “minimum size” selected. The original application knows which images are decorative and which matter — a compressor can only guess.

Rescan at a sensible resolution. 600 dpi is for archival reproduction of artwork. For a text document that someone will read, 200–300 dpi is plenty, and scanning in greyscale or black-and-white rather than colour reduces the data further before you even start.

Don’t send the PDF at all

If the file is large because it contains 40 photographs, a PDF is the wrong container — you’re wrapping already-compressed images in a document format that adds overhead. Send the images, or a shared link.

For genuinely large files, a link from Drive, Dropbox or WeTransfer beats fighting an attachment limit. It’s also kinder to the recipient’s inbox.

A note on “compress by 90%” claims

Any tool can produce a 90% smaller PDF by rendering it to low-quality images. The number alone means nothing without knowing what was destroyed to get it.

The useful questions are: is the text still text, is it still readable at 100%, and does it still print acceptably? A tool that shows you the before-and-after sizes and tells you what mode did what is being straight with you. One that shows a big percentage and nothing else is selling the number, not the result.