Organise old documents retroactively

The archive has grown over ten years, across two computer changes and three cloud services. There is a folder called Old, a folder called Old_2, and inside it doc_2019_misc.pdf. You need the 2019 amendment to your car insurance and you know it is in there. Collections like this can be cleared up, but not in the order people usually begin.

Download for free

Before & After

doc_2019_misc.pdf 2019-07-15_Versicherung_HUK_KFZ-Nachtrag.pdf

Why grown archives eventually tip over

Each year things were filed according to whatever logic made sense at the time. In 2014 by area of life, in 2018 by year, and since the last computer change not at all. Add the catch-all folders created during a migration and never dissolved again. The result is not chaos in the sense of untidiness but several orderings sitting side by side — and that is exactly what makes searching impossible, because you first have to guess which logic a document was filed under back then.

First settle what actually has to stay

Before renaming anything, separate three groups. First: material with a retention obligation or lasting practical use — tax papers, contracts, insurance policies, tradesman invoices because of warranty periods, property documents. Second: documents that are closed and whose retention period has expired. Third: material that should never have been archived. Delete the third group and, where in doubt, check the second with your accountant. What remains after that is the collection worth the effort.

Folder structure or filename?

The classic decision: deep folder trees or a flat store with descriptive names. One practical argument favours the filename — it travels with the file. Moving to another cloud service, importing into a document management system, sending by mail: the name survives, while the folder structure is lost or changes. A mixed solution usually lasts best: a few flat folders by area of life, with the actual information in the filename and the date at the front.

Old scans often have no text layer

The older the collection, the higher the share of pure image PDFs. Devices from before roughly 2010 often shipped no text recognition at all, and many users never turned the setting on. Full-text search finds nothing in those files, and a content-based tool cannot evaluate anything either at first. Filery has an optional OCR step for this. Check the share of such files on a sample before you estimate the time needed for the whole collection.

How to work through a legacy collection

Make a copy first, then settle the naming pattern, for example Date_Sender_Type with the date in ISO form. Start with the folder you actually search most often, not the oldest one. Filery reads the PDFs, recognises sender, document date, document type and subject, and shows each suggestion next to the old name. Expect several stages: the Starter plan covers 50 documents per month, Business 500. Renamed files then appear in the app's history.

Sensitive old material

Old archives contain things you do not casually send through a service: medical letters, expert opinions, papers from a divorce, payslips. Sort those out beforehand and name them by hand. For everything else: Filery reads the PDF on your machine and sends the content encrypted for AI analysis, the recognised metadata comes back, the renaming runs again on your own device, and the files themselves are not stored permanently. Anyone preferring their own provider account uses the Custom plan with their own API keys.

FAQ

The age of the file plays no role; the text layer does. If text is embedded, it is read just as with a new document. With pure image PDFs, the optional OCR and the quality of the original at the time decide the outcome.

No. Filery changes the filename, not the storage location. If you want to simplify the structure anyway, move things first and rename afterwards, so you only have to go through the collection once.

Take thirty random files from the collection, not the tidiest ones. How many of those yield usable suggestions, scaled up to the total, gives a far more dependable estimate than any rule of thumb.

That depends on how often you search the old material. If it is touched once a year for tax, a roughly sorted yearly folder is often enough. If you look things up in it regularly, the work pays for itself quickly.

Related topics

Fix messy PDF folder chaos
Rename thousands of PDFs retroactively
Enforce file naming conventions automatically

Download Now

Choose your platform and start with automatic document renaming

macOS

Universal (Apple Silicon & Intel)

Download (DMG)

Windows

Windows 10 & 11

Download (EXE)

We are currently working on official code signing for Windows. The download is already available – Windows Defender or SmartScreen may show a warning because the app is not yet certified. Choose “More info” → “Run anyway” to install.

System requirements: macOS 10.14+, Windows 10/11

All downloads are transmitted securely via HTTPS.