How to Create and Organise a Collection

A Collection groups related Sources and gives them shared catalogue context. Use one when several datasets belong to the same programme, producer, geography, subject or release family.

Every Source has one required primary Collection, and the Source's public visibility derives from that Collection. Collection design therefore affects both organisation and access.

Before you create the Collection

Choose the account that should own it and make sure you can create resources there, then decide what makes the Sources a coherent family. Good Collection boundaries are things another person could recognise later: a national census release, a conservation programme, a family of administrative boundaries, or a regularly updated transport dataset.

Avoid creating a new Collection simply because a new file arrived. If an existing Collection already owns the same subject and stewardship, using it usually produces a clearer catalogue.

Create the Collection

Switch to the individual or organisation account that should own the Collection.
Open Collections and start the create flow.
Enter a concise title and stable slug.
Write a short description of what belongs in the Collection.
Add the catalogue metadata you can support from evidence, such as subject tags, countries, geographic scope, collection type, update cadence, version label, original creator, original URL and licence information.
Review the Collection's visibility before treating it as a public catalogue resource.
Create the Collection and reopen its detail page to verify the result.

Collections can also carry review-state information such as whether the metadata is current, needs review or has become stale. Use those fields when they help a team manage a maintained catalogue rather than relying on memory.

Add Sources to the right primary Collection

When you upload, import, extract or otherwise create a Source, choose the Collection that should own it. The primary Collection is not just a folder label: it supplies the visibility relationship and shared catalogue context for the Source.

Keep Collection-level and Source-level information separate: a Collection can state that a family comes from a national statistics office, while an individual Source records the exact reference date, retrieval date, processing notes, limitations and recommended citation for one release.

If two Sources have different licences, vintages or derivations, do not hide those differences behind a broad Collection description; record the shared context at Collection level and the specific evidence on each Source.

Review organisation before publishing

Ask these questions before making the Collection broadly discoverable:

Would a new reader understand why these Sources belong together?
Does the geographic scope describe the family accurately?
Is the original creator distinct from the Geobble account maintaining the Collection?
Are licence details supported by an authoritative source?
Could changing the Collection's visibility expose Sources that should remain restricted?
Is the update cadence or version label still current?

Because Source visibility follows its primary Collection, changing Collection visibility deserves the same care as changing access on the datasets themselves.

Expected result

You should end with a Collection that has a clear identity, useful shared metadata and Sources whose individual records explain their own dates, transformations and limitations. A well-organised Collection reduces duplicated metadata without flattening important differences between datasets.