How data pack versioning works

Version codes

Each dataset has a version code made of the country, the dataset family and the release month, for example UK-BIZ-2026-09. When the data is re-processed, a new version such as UK-BIZ-2026-10 is created and listed in the dataset changelog.

What is recorded when you buy

For every pack delivered to you we store:

• pack ID (e.g. UK-0017)
• dataset version (e.g. UK-BIZ-2026-09)
• your order number and allocation date
• the SHA-256 checksum of the exact file you received
• the record count of that file

Files are never silently replaced

If a file that has already been delivered changes on the server, its checksum no longer matches the recorded one. The system freezes that pack, pauses downloads and alerts an administrator instead of quietly handing out different data. New data is always added as new packs or a new version.

Why this matters to you

• You can prove exactly which data you received and when.
• If you report a problem with a pack, support can see the precise file and version.
• When a new version is released, previous buyers can be notified and decide whether to buy the new packs.