The UK has spent the last decade understanding the need for data catalogues. Thankfully, almost every public-sector organisation that holds significant data now uses catalogues, making their data easier to find, track and use.
And yet, ask a researcher to find the right dataset for their question and the answer is often a chain of emails, phone calls, or a half-remembered conversation at a conference. The data exists…somewhere. But the discovery layer, the connective tissue that lets people see what’s out there and decide whether it’s relevant, is often still missing.
The instinct, when faced with this, is to build another catalogue. A bigger one. A national one. A definitive one. The one that finally brings everyone onto the same platform, retires the smaller catalogues, and replaces the fragmented landscape with a single source of truth.
Underneath sits a familiar set of assumptions. That fragmentation is a failure of coordination. That everyone else’s catalogues are part of the problem, not part of the answer. That if data owners would just migrate into the central system and accept its model, the discovery problem would solve itself. And that centralisation — properly resourced and properly governed this time — will finally work.
We don’t think that’s always the answer. Public sector organisations in many cases don’t need one more catalogue at the centre. They need a way to look across the catalogues that already exist.
A single national catalogue requires every data owner to migrate their metadata into it, accept its model, and keep it updated. In practice, that’s a slow, sometimes expensive ask — one that competes with every other priority on a CDO’s desk. It also raises legitimate questions of governance: who decides what’s listed, what’s removed, and how access requests are routed?
And this assumes the existing catalogues are interchangeable. They’re not. Some are specialists in a particular tech stack. Some are custom-built for a specific geographic or domain use case. Some have been refined over a decade by the people who actually use them, with training programmes that took years to bed in. Asking those data owners to migrate isn’t asking them to move a list of datasets — it’s asking them to throw out the systems, conventions and institutional knowledge that make their catalogues work in the first place.
The result is predictable. Centralisation projects launch with ambition, capture some early adopters, and then stall.
Worse, they don’t actually solve the discovery problem. Researchers don’t just need to know about data in one place — they need to know about it across every place. Across NHS trusts, ICBs, universities, councils, government departments, and the research data services those organisations already work with.
This is the idea behind the SDR UK Metadata Catalogue, which MetadataWorks has been working with Smart Data Research UK to build.
A catalogue of catalogues is exactly what it sounds like: a federated layer that makes the metadata held in existing catalogues discoverable in one place, without requiring data — or even metadata governance — to move. Data owners keep their catalogue. Researchers get a single search surface. Standards bind the pieces together.
Three design principles make this work:
The point of all this isn’t elegance for its own sake. It’s that a federated approach is the only one that scales: it lets new data owners join when they’re ready, on their terms, without an all-or-nothing commitment to a central platform.
On Wednesday 3 June, MetadataWorks and SDR UK are running a free, hour-long instructional webinar, Catalogue of Catalogues: Discoverability, Interoperability and Smart Data Research UK, for public-sector data leaders who want to see how this is being done in practice.
You’ll see a live walkthrough of a researcher moving from a research question to candidate datasets in real time. You’ll hear what the team learned building the catalogue, what we’d do differently, and what’s coming next. And we’ll cover practical paths in for both data owners (how to list your datasets) and researchers (how to feed back and join the working groups shaping future releases).
If you’re a chief data officer, head of data, information-governance lead, or part of a data-strategy team responsible for making your organisation’s data findable, governed and re-usable at scale, this is an hour well spent.
Wednesday 10 June 2026, 11:00–12:00 BST. The session will be recorded for those who can’t make the live slot.