How to compare two versions of a PDF and find every change

Someone sends you the updated contract and says only the dates changed. Maybe they did. Reading two twelve-page documents side by side is the slowest and least reliable way to find out. Here's how to compare them properly.

Why reading side by side doesn't work

When you compare two documents by reading them in parallel, you check the things you expect to have changed. Your eyes skip over familiar paragraphs. And because one added sentence pushes every line after it down, the two versions stop lining up after the first edit, so you lose your place exactly where it matters.

A missing "not" or a "three" that became "two" is easy to read past. A comparison tool doesn't get tired.

Two kinds of comparison

• Text comparison reads the words in both files and lists what was added, removed or changed, regardless of layout. It's the right choice for contracts, policies, terms and anything where the wording is what counts.
• Visual comparison lays the pages over each other and highlights pixels that differ. It catches changes to logos, tables, charts and formatting, but gets noisy as soon as text reflows.

For most documents, start with a text comparison. Add a visual one only when the look of the page is part of what you need to check.

Before you compare

1. Make sure both files contain real text. Try selecting a sentence in each. If one is a scan, run text recognition (OCR) on it first, or the comparison will have nothing to compare.
2. Know which file is old and which is new. Rename them if you have to. "Contract final v2 (1).pdf" helps nobody.
3. Compare like with like. If one version has a cover page or signature page the other doesn't, expect those to show up as changes and set them aside.

The traps that hide real changes

• Reflowed text. One added word can shift every line break after it. A good tool compares word by word, not line by line, so a moved line break isn't reported as a change.
• Headers, footers and page numbers. These create noise on every page. Learn to recognize them and look past them.
• Formatting-only edits. A text comparison won't tell you that a clause is now in bold or struck through. If formatting matters, check it visually.
• Images of text. A scanned signature block or a pasted image of a table isn't read as text.

Go through every change

Work through the list of changes in order, with the page number for each. For every one, ask: what changed, and does it matter? Slow down for:

• amounts, percentages, dates and deadlines,
• notice periods and durations,
• names of people, companies and places,
• small words that flip meaning: not, no, may, must, shall, unless, within.

A worked example

Nadia's landlord sends a renewed tenancy agreement and writes that only the dates have been updated. She runs a word-by-word comparison of the old and new PDFs. It finds five changes.

Two are the new start and end dates, as promised. The third is on page 2: the monthly rent went from 950 to 980. The fourth is on page 7, in the termination clause, where "three months' notice" became "two months' notice". The fifth is a new sentence in the section on pets.

Reading side by side, she might have caught the rent. The notice period, eight paragraphs into a page of dense clauses, she almost certainly wouldn't have. Now she can ask about all three before she signs.

Ask for tracked changes, then check them

If the other side sends a marked-up version showing their edits, that's helpful. But they made it, and an edit can be missed, deliberately or not. Run your own comparison anyway. It takes a minute.

Contracts are private documents

Contracts, offers and agreements tend to contain names, addresses, salaries and account numbers. Uploading two versions to a comparison website means sending all of that twice. ClearRead's compare tool, one of the PDF tools in the full version, runs on your own device and shows each change word by word, with the page it's on.

Common questions

How do I compare two PDF files for differences?

Use a text comparison tool that reads both files and lists what was added, removed or changed, word by word, with the page of each change. Check first that both PDFs contain selectable text.

Can I compare a scanned PDF with a digital one?

Yes, but run text recognition (OCR) on the scan first. Without it, the scan is only an image and there is no text to compare.

Why does the comparison show changes on every line?

Usually because the text reflowed, so line breaks moved. Use a tool that compares word by word rather than line by line, and ignore headers, footers and page numbers.