Why Your Compressed PDF Is No Longer Searchable (and When That Matters)

You compress a 40 MB report down to 3 MB, feel pleased with yourself, then press Ctrl+F and nothing is found. The text is right there on screen. The search box insists it does not exist.

This is not a bug. It is the mechanism, and understanding it tells you exactly when to compress and when not to.

What is inside a PDF

A PDF page can hold two very different kinds of content, and they look identical on screen.

Real text is stored as characters plus instructions: this font, this size, this position. It is compact, infinitely sharp at any zoom, selectable, copyable and searchable. A hundred pages of it might be 300 KB.

Images are stored as pixels. A scanned page is a photograph of text — your eye reads words, the file contains only coloured dots. It cannot be searched because there is nothing to search.

Most PDFs are a mixture: real text with embedded photographs and logos.

Why compression flattens it

Text is already tiny. If a PDF is large, the weight is in the images — and the only way to make images meaningfully smaller is to reduce their quality or resolution.

The most effective general-purpose approach is to render each page as a single image at a chosen resolution, compress that image, and rebuild the PDF from it. This works on absolutely any PDF and routinely produces 70 to 90 per cent savings.

The cost is that everything on the page becomes pixels, including text that used to be real text. The page looks the same. It is no longer searchable.

When the trade is fine

Most of the time, honestly.

• Emailing a report somebody will read once
• Uploading a signed form to a portal with a size cap
• Sending scanned documents that were never searchable anyway
• Sharing an invoice or receipt
• Any document a person will look at rather than query

If the recipient is going to read it and move on, searchability is worth nothing and three megabytes is worth a lot.

When it is not fine

• Reference documents — manuals, standards, legislation, anything people search rather than read
• Archives — a searchable archive is worth many times an unsearchable one
• Anything feeding a system that extracts text automatically
• Accessibility — screen readers cannot read an image of text, which makes it a real barrier on a public website
• Documents others will edit — you cannot copy text out of a picture

What to do instead

Send fewer pages. Very often the best answer. Extract the pages that matter and send a small file at full quality with the text intact. A 3-page extract beats a compressed 200-page document on every measure.

Use Light compression. Renders at higher quality, so the file is bigger but the page stays sharp. Still not searchable, but far better for printing.

Fix it at the source. If you are creating the PDF, the problem is usually oversized embedded images. Compress the photos before placing them, export at web quality rather than print, and the PDF comes out small with its text layer completely intact.

Rescan sensibly. 200 to 300 DPI in greyscale produces a fraction of the data of 600 DPI colour, with no loss of readability on a text document.

Send a link. For a large reference document, a cloud link preserves everything and sidesteps attachment limits entirely.

Always keep the original

Compression is one-directional. There is no way to turn the pixels back into text short of running optical character recognition, which is a separate process with its own error rate — and OCR on an already-compressed page performs noticeably worse, because you are asking it to read blurred letters.

Compress a copy. Keep the original. This one habit prevents most compression regrets.

How to check what you have

Open the PDF and try to select a line of text with your cursor. If you get a neat text selection, it is real text. If you get a rectangular selection box or nothing at all, it is an image.

Ctrl+F for a word you can see works just as well as a test.

Compress with your eyes open

Compress PDF offers three levels and shows the before-and-after size so you can judge the trade rather than guess at it. Start at Balanced, check the result, and step up or down.

Everything runs in your browser — your document is never uploaded, which matters given that the files people compress most are contracts, statements and identity documents. It is free for life with no sign-up and no daily limit, so you can try all three levels and compare them without anything counting against a quota.