Removing the author name from office documents
Guide · 18 August 2026
Where the record lives
A .docx, .xlsx or .pptx file is a ZIP archive. Alongside the parts that hold the
actual text, sheets or slides sit small XML files dedicated to properties:
docProps/core.xml and docProps/app.xml, plus an optional
docProps/custom.xml for properties added by other software. The
OpenDocument formats — .odt, .ods, .odp — keep the equivalent in a single
file called meta.xml. None of it appears on the page itself: it is
packaging, and most of it is written by the application, not typed in by you.
What is in it
- Author and last modified by — two names, and often two different ones. A document drafted by one person and finished by another names both.
- Company and manager, when they were configured on the machine that edited the file.
- Total editing time, a counter in minutes, and a revision number that grows with each editing session.
- The application and its version, and the template the document was built on.
- Created and modified timestamps.
- Title, subject, keywords, comments, category and content status — free-text fields that keep whatever was once typed into them.
- Custom properties, arbitrary named values that some corporate
tooling writes into
docProps/custom.xml.
OpenDocument files carry the same story under different names: an initial creator, a
last editor, a generator string, dates, an editing duration and an editing-cycles
counter, all in meta.xml.
The day you press Send
These fields are often chattier than a GPS fix. A quote that started life as a copy of another client's quote can keep that client's name in its title. A report names the colleague who reviewed it last. A document sent to a competitor names your company. A CV written on someone else's computer lists that someone as the author — the name in the properties and the name on the page do not have to match. The editing-time counter tells its own story: minutes spent, recorded without comment, on a file you may have presented as quick work or as long work.
Deleting the files is the wrong fix
The obvious move — open the archive, delete docProps/core.xml and
docProps/app.xml — produces a broken document. Every part of an
OOXML package is declared in [Content_Types].xml and referenced by the
package relationships. Delete the parts and those declarations point at nothing;
Word and Excel then report the file as corrupted.
QuietMeta therefore rewrites these files instead of removing them. The cleaned
document still contains valid, declared properties files — they simply no
longer carry the fields you asked to remove. For OpenDocument there is one extra
formality the cleaner also respects: the mimetype entry must stay at
the front of the archive, stored without compression, for the file to remain valid.
Selective, category by category
Removal uses the same categories as for photos, so the interface does not change with the file type. Four of them act on documents:
- Author and copyright — creator, last modified by, initial creator, company, manager.
- Title, description and keywords — title, subject, keywords, comments, category, content status, and the whole custom-properties file, which is replaced by a valid empty one rather than examined field by field.
- Timestamps — the created and modified dates.
- Creation software — the application name, its version, the template, and the generator string in OpenDocument files.
The properties are rebuilt keeping every field whose category you did not tick. Removing your name does not cost you the document's title.
What stays, and why
Counts that say nothing personal — pages, words, characters, lines, paragraphs, slides, notes — are kept. Fields the cleaner does not recognise are also kept, deliberately: deleting a field written by software the cleaner has never seen could produce a document that software can no longer open.
That caution has one consequence worth stating plainly. The revision number and the editing-time counter, which the inspector shows under a technical heading, belong to no removal category today, so cleaning leaves them in place. They appear in the report on the cleaned file rather than being silently assumed gone.
The content itself is not touched
Everything that is not a properties file — paragraphs, worksheets, formulas, slides, embedded images — is copied into the new archive as it was, without being interpreted. The archive is recompressed on the way out, but a ZIP has nothing lossy about it, unlike a JPEG: the document content comes out identical.
The same fact has a flip side. Names that live inside the content are content, not properties: a comment in the margin, tracked changes, a name typed in a header or on the page itself all pass through unchanged. Cleaning the properties does not proofread your document; that part stays yours, in your editor.
Checked after, not assumed
After cleaning, QuietMeta reads the produced file again and builds its report from that second reading, not from the intention. You see what is gone and what is still there. And when a document simply had nothing in the categories you selected, the report says so in as many words: this document carried no properties in the categories you selected.
Reading a document without opening all of it
Inspection never unpacks the whole archive. Only the few metadata files are decompressed, none larger than 512 KB, and an archive that declares more than 2,000 entries is set aside. The fields are read by plain tag extraction rather than an XML parser, so external-entity tricks aimed at parsers do not apply.
Doing it
Drop a document on the inspector to see every property it carries, grouped by category, then choose what to remove. Reading works on DOCX, XLSX, PPTX, ODT, ODS and ODP alongside images and PDF; removal works on all six document formats, plus JPEG, PNG, WebP and PDF.
Files sent to the web version are held in memory for one request and released — never written to disk, a database or a backup. For whole folders at once, or for files you would rather not send anywhere at all, the desktop application does the same work offline.
Inspect a document Get the desktop app