Two archives, two problems, one word doing all the work.
Archive A scanned forty thousand photographs last year. Every image is a beautiful high-resolution TIFF. Not one of them has a title, a date, or a box number attached. Nobody can find anything without opening files one at a time.
Archive B has a card index that any experienced archivist could search in minutes. It has never been near a scanner. The material it describes is still in its original boxes. It doesn’t exist online, so as far as a researcher with a search engine is concerned, it doesn’t exist at all.
Both archives would probably describe what they need next as “digitisation.” They need two different things.
Digitisation converts the material
Digitisation is the process of turning physical items into digital surrogates: scanning documents, photographing objects, capturing audio or film. The input is the thing itself — a letter, a photograph, a reel of tape. The output is an image or media file that represents it.
Digitisation does not, by itself, tell you what’s in the file. A folder of ten thousand scanned pages with no titles, dates, or reference numbers is still an unsearchable archive — it has just changed format. Digitisation Is Not the Finish Line covers this trap in more depth.
Retroconversion converts the catalogue
Retroconversion (short for retrospective conversion) is a different process entirely: turning existing catalogue data — card indexes, typed finding aids, register books, spreadsheets — into structured, digital, searchable records.
The input here is description that already exists somewhere, usually on paper. The output is a proper catalogue record: consistent fields, a reference number, a place in the collection’s hierarchy, ready to sit in a database or be published online.
Crucially, retroconversion has nothing to do with whether the underlying material has been scanned. A card index for a collection still sitting untouched in boxes is a perfectly valid retroconversion project on its own. For the full picture of what that involves, see what retroconversion means and how it works.
Why the two get muddled
Because both projects use the word “digital” and both often happen around the same time, in the same budget line, sometimes even in the same sentence in a funding application. But they answer different questions.
Digitisation asks: can a person look at this item without touching the original?
Retroconversion asks: can a person find out this item exists, and where, without a member of staff walking to a filing cabinet?
An archive can answer yes to one and no to the other. That mismatch is exactly what produces Archive A and Archive B above — a pile of undiscoverable images on one hand, a perfectly good paper finding aid invisible to the internet on the other.
At a glance
| Digitisation | Retroconversion | |
|---|---|---|
| Converts | Source material (documents, photographs, audio, objects) | Catalogue data (indexes, finding aids, registers) |
| Typical starting point | Original physical items | Existing paper or legacy catalogue |
| Typical output | Digital surrogates (images, audio/video files) | Structured, searchable catalogue records |
| Usually commissioned by | Collections or conservation teams, funders backing access to specific material | Collections managers, heads of service tackling a description backlog |
| Where they overlap | A single project can do both — scan the item and catalogue it at the same time | Neither depends on the other; each can happen alone |
And migration is a third, separate thing
There’s a related term worth ruling out too: system migration. Moving a catalogue you already hold in one system into another — say, from a legacy database into ArchivesSpace or AtoM — is not retroconversion, because the data is already machine-readable. It needs mapping and cleaning, not conversion from paper. If that’s your situation, CALM migration is the more relevant starting point.
Where Archivers.ai fits
Because these are genuinely separate problems, Archivers.ai is built to handle both without forcing an institution to run two projects with two suppliers. It transcribes and structures source material where digitisation is still needed, and it takes existing catalogue data — paper, legacy exports, whatever form it’s in — and structures it into the same ISAD(G)-aligned hierarchy of fonds, series, file and item.
Every record carries draft or review status, so nothing reaches a public portal until someone with archival judgement approves it. Whichever direction a collection needs — material converted, description converted, or both — the output lands in the same standards-based catalogue, ready to export to EAD3, Dublin Core, or the system a service already runs. If retroconversion is the piece you’re missing, that’s a good place to start.