Remove Excel metadata

Workbooks get forwarded. A price list goes to a customer who sends it to a colleague who sends it to a competitor, and the properties travel the whole way — naming the person who built it and the company they built it at, long after anybody remembers where the file came from.

The fields are the same as any Office document: author, last saved by, company, manager, template, creation and modification dates, and the cumulative editing time. Custom properties are cleared too, and in corporate templates those often carry a client name, a matter number or a classification label.

Everything that makes it a spreadsheet: values, formulas, formatting, charts, pivot tables, named ranges and defined links. This rewrites the file's description of itself, not its contents.

Which means the things worth checking are in the content. Hidden sheets and hidden columns are still there and still readable. Cell comments name their authors. A pivot table keeps a cached copy of its whole source range, so a pivot built from data you later deleted can still contain that data. None of those are metadata, and none of them are removed here.

A workbook that once pulled a figure from another file keeps the path to it, and those paths are startlingly descriptive: a drive letter, a server name, a department folder, sometimes a client. The formula may have been replaced by a value years ago and the link definition can still be sitting in the file.

These are stored as document-level definitions rather than as properties, and clearing the properties does not touch them. Excel’s own Data → Edit Links dialog is where they are broken, and doing that before cleaning the file here is the order that leaves nothing behind.

An XLSX is a ZIP of XML parts. Removing the properties removes some of those parts and rewrites others, and because the archive is repacked in the process, the result is rarely byte-identical in size to the original even though every cell is unchanged.

A small increase is as normal as a small decrease — it depends on how the original happened to be compressed. What matters is that the workbook opens with the same values, formulas and formatting, which is the thing to check rather than the byte count.

In Excel, File → Info shows the properties panel; after cleaning, the author and company fields are empty rather than showing your name. That is the check most people want and it takes ten seconds.

The more thorough check is to open the cleaned file’s `docProps` folder — an XLSX renamed to `.zip` opens in any archive tool — and read `core.xml` and `app.xml` directly. That reads what is actually in the file rather than what one program chooses to display, which is the difference between believing the properties are gone and knowing it.

Hiding a worksheet removes it from the tabs along the bottom and from nothing else. The same is true of a hidden column or a collapsed group of rows: the values are in the file, they are read by every formula that references them, and unhiding is a right-click away.

It is the most common way a workbook says more than it was meant to — the sheet with the cost basis behind the price list, the column with the individual salaries behind the total. Before a workbook goes outside, right-clicking a sheet tab and looking for an enabled Unhide is a five-second check that no metadata tool can do for you.

A pivot table keeps its own copy of the source data inside the file so it can be re-sliced without the original. That cache survives deleting the sheet the pivot was built from, which means a workbook can carry a full copy of data that appears nowhere in it.

Anyone can reach it by double-clicking a total, which spills the underlying rows onto a new sheet. If the source data was the sensitive part, removing the sheet is not enough — the pivot has to be converted to values, or the cache disabled in its options and the file saved again.