How to Organise a Large PDF Library

A folder holding four thousand PDFs named "scan_0043.pdf" and "document (2).pdf" is not an archive. It is a place documents go to become unfindable.

Naming is most of the work

A good filename answers three questions: when, what, and who. Put the date first in ISO format so files sort chronologically on their own.

Pick one and use it everywhere. An inconsistent convention is only marginally better than none, because you cannot predict what to search for.

OCR everything

This is the difference between a library you can search and one you can only browse. A scanned document without a text layer cannot be found by content — not by your operating system's search, not by your document manager, not by anything. Running OCR over an archive once, as a batch, transforms it permanently.

Folders: shallow and obvious

• Two or three levels at most — deep hierarchies get abandoned.
• Organise by the thing you look for first: client, year, or document type.
• Keep an inbox folder for unfiled documents, and empty it weekly.
• Do not create a folder for a single file.
• Be consistent about singular and plural, and about capitalisation.

Habits that keep it working

1. 1Rename a file when it arrives, not later. Later never comes.
2. 2OCR scans as part of filing them, not as a project you will do one day.
3. 3Merge related documents into one file — an invoice and its receipt belong together.
4. 4Compress large scans at filing time.
5. 5Review the inbox folder weekly.

Backups

An organised archive that exists in one place is one failure away from nothing. Three copies, two kinds of storage, one of them off-site is the conventional rule and it is conventional because it works. Test a restore occasionally — a backup nobody has ever restored from is a hypothesis.