Clean a document before you send it SCRUB

What is actually in an Office file

A .docx, .xlsx or .pptx is a zip archive, and several of the parts inside it are about you rather than about the document. SCRUB opens the archive, rewrites those parts and zips it back, so the file you send is the file you meant to send.

• docProps/core.xml: the author, the last person to save it, the creation and modification times, and any category or keywords. The author is usually whoever created the template, which on a document that has been passed around a firm is frequently not the person sending it.
• docProps/app.xml: the company name, the total editing time in minutes, the revision count, and the application that wrote it. Total editing time is the field that surprises people: a document billed as a day's work carries the number of minutes it was actually open.
• The template path. Often a path on a network share, which names the organisation, the department and sometimes the client matter.
• Comments, and tracked changes. A tracked deletion stores the deleted text. Accepting changes is not the same as removing the record of them, and a document that looks clean in Word can still carry every sentence somebody took out of it.
• Custom document properties, which document management systems use to stamp a matter number, a client code or a classification onto every file they touch.

What it does not touch, on purpose

• Anything visible. If a name is in a paragraph, it stays in the paragraph. This removes what travels alongside the document, not what is written in it; for that you want SEAL on a PDF or your own eyes on the text.
• Data in a spreadsheet you cannot see. Hidden rows, hidden sheets, filtered-out values and the cached results of a pivot table are content rather than metadata. CLEANSHEET is the tool for those, and what a spreadsheet still contains explains why the pivot cache is the one that catches people out.
• Photographs. An image placed in a document keeps its own EXIF, including where it was taken. Run it through CLOAK before it goes in.
• A PDF's text. On a PDF this removes the author and producer fields and nothing else. A PDF is not a zip of properties and a black rectangle over a paragraph is not a redaction, which is a different job with a different tool.

Checking it worked

You do not have to take this tool's word for it, and you should not take any tool's word for it. Copy the cleaned file, rename the copy to .zip, open it, and read docProps/core.xml and docProps/app.xml in any text editor. They are short and they are in plain English. X-RAY will list the same fields without the renaming if you would rather, and what is hidden in a Word document goes through the whole set.