Ask an archivist whether a collection is “catalogued” and a card index or a typed finding aid usually counts as yes. Fair enough — someone did the intellectual work. They surveyed the material, worked out the arrangement, wrote the descriptions, and produced something a researcher can use.
Ask whether that same collection is discoverable, and the answer changes.
Why a finding aid and a discoverable collection are not the same thing
A finding aid answers the question “what is in this collection, and how is it arranged?” Discoverability answers a different question: “can someone who doesn’t already know this collection exists find it?”
A paper finding aid, however well made, only answers the first question. It sits in a filing cabinet or a card drawer, in one building, readable by one person at a time, findable only by someone who already knows to ask for it. No search engine indexes it. No AI answer engine can cite it. No online catalogue portal can surface it alongside anything else you hold. It is real, professional cataloguing work — and it is invisible to almost everyone who might want it.
This is a different starting condition from a digitisation backlog. A backlog of scanned images that nobody has described yet is one problem — we’ve written about why digitisation on its own isn’t the finish line. This is the opposite shape of problem: the description already exists, and it’s good. The barrier is that it’s only on paper. The source material might not even be digitised yet — that’s a separate task. Here, the finding aid itself is what’s locked away.
What it actually costs a researcher to rely on a paper finding aid
Picture the researcher’s route into your collection today. They have to know, or guess, that your institution holds relevant material — usually from a footnote, a colleague, or luck. They have to contact you, or turn up in person. They have to be in the room, at the right time, with someone available to pull the drawer or the folder.
Once they’re there, they can search your index of your collection. They cannot search it alongside the twelve other archives that might hold related material, because none of those indexes talk to each other, let alone to the open web. A researcher working from home, or from another country, or simply searching Google for a name, a place or an event, will never land on your holdings at all. The finding aid did its job for the one person who found the drawer. It does nothing for everyone else.
Why funders read a paper index differently to an online one
Funders such as the National Lottery Heritage Fund increasingly ask for evidence of public engagement and access, not just evidence that cataloguing happened. That distinction matters more than it might seem.
A card index in a drawer is easy to describe in a funding report and hard to evidence as public benefit. Nobody outside your reading room can see it, search it, or point to a moment where it helped someone. An online catalogue entry, by contrast, generates its own evidence: page views, search terms, enquiries from people who found you unprompted. If a report needs to demonstrate that cataloguing translated into public access, a paper finding aid is the wrong kind of proof, no matter how good the underlying work is.
The National Archives’ own surveys of the UK sector have consistently pointed to a substantial share of holdings remaining uncatalogued to any meaningful standard. A finding aid that exists only on paper sits in an odd middle ground here — catalogued in the archivist’s sense, uncatalogued in the funder’s and the public’s.
How structured data connects a finding aid to search, a portal and Explore
None of this is really about digitisation for its own sake. It’s about what structured, digital description makes possible once it exists.
A public catalogue portal needs records it can index and search — a scanned photograph of a card is not searchable text. A staff or public research feature that answers questions across your holdings and cites the items behind its answer needs the same thing: structured records and transcripts it can actually query, not a photograph of a drawer. Cross-referencing between related items, filtering by fonds or accession, surfacing a record to someone who didn’t know to ask for it by name — all of it depends on the description existing as data, not as paper.
That’s the through-line from retroconversion to public benefit: it isn’t an efficiency exercise that happens to make cataloguing tidier. It’s the precondition for everything downstream of the catalogue.
Where Archivers.ai fits
Archivers.ai’s managed retroconversion service takes existing finding aids and card indexes — good ones, the kind your predecessors compiled with real care — and turns them into structured, ISAD(G)-aligned records. We handle the onboarding and field mapping, and every converted record carries a draft/review status so nothing reaches a public page without a member of your team approving it first.
Once that description is structured, it can do things a card can’t. It becomes searchable across your whole catalogue, browsable by fonds and accession, and part of the underlying data behind public access and Explore — the research feature that answers questions across your reviewed records and shows its working.
If you want to understand what the conversion process itself involves, read what is retroconversion? It walks through the practical steps. This post is about why it’s worth doing: the archive is bigger than the catalogue, and a finding aid in a drawer, however well written, doesn’t make more of it discoverable.