Retention and archiving

Retention and archiving cover the end of the content lifecycle, a stage often represented in diagrams and neglected in calendars.

Unpublishing takes content off the public site. Archiving preserves material and its context for historical, legal or reference use, with public or restricted access as appropriate. Deleting removes the record, subject to applicable retention duties and legal holds. Retention schedules cover records whether or not they’re published. GOV.UK’s withdrawal and history mode are examples of marking historical content while keeping it public, rather than definitions every organisation must follow.

Why it matters for content

Keeping everything live can feel safe, but it has costs. Old content goes on being read and acted on: last year’s prices, a returns policy PDF from 2019, an appeal for an emergency that closed months ago. It competes with current pages in search, contradicts newer guidance and makes every audit, redesign and migration longer, because each page kept is a page someone should be checking. Where it holds personal details, such as a customer’s or volunteer’s story, keeping it can also run into privacy law: the GDPR and similar laws elsewhere limit how long personal data can be held.

The other common pattern is the retention debate that never ends. Without agreed criteria every page has a champion, every retirement becomes a negotiation and the content stays up while the meeting is rescheduled. It ends in the same place as keeping everything, with more meetings.

Removing content carelessly can have costs too: broken links, a lost public record, a breached retention duty. An old policy announcement may need to remain publicly accessible as a historical record, clearly marked and linked to its replacement. A policy agreed in advance settles most cases before anyone gets attached to a page, and saves the argument for the few that deserve one.

How AI changes the picture

Externally invisible isn’t the same as invisible to machines. An internal assistant with access to the shared drive or intranet will happily quote the archive, and a banner saying “no longer current” helps a reader but may not survive being summarised. Pages taken off the site can also linger in search indexes and in the sources an assistant was given before they came down.

So retention rules need technical consequences: whether archived material stays eligible for retrieval, how its status shows up in metadata and what replaces it as the canonical source. Otherwise yesterday’s policy can return as today’s answer with excellent grammar.