Your website content is a Commonwealth record. What that means before you delete a page.

An Australian Government website is a Commonwealth record. Deleting a page on one is a disposal decision, not housekeeping. That single fact moves your website into records management, and most content teams never hear it.

Whether you owe the Archives a copy before you switch a site off depends on whether the National Library of Australia harvested it. Nothing in your content management system decides it.

I have taken government websites off the internet. The records question is never the one a content team expects. It also tends to arrive far too late.

Who the Archives Act binds

The Archives Act 1983 covers Commonwealth records. The National Archives lists it as a law for Australian Government agencies. Under it, an agency needs Archives authorisation to destroy a Commonwealth record. If your records are state or territory records, your own law applies instead. The ACT, for one, has the Territory Records Act 2002, with its own disposal schedules made under it.

Find out which records law your agency sits under before you apply any of this. The rest of this guide is written for the Commonwealth instruments.

The National Archives sets out the lawful routes to destroying a Commonwealth record. A specific law calls for it, the National Archives permits it, or a normal administrative practice covers it. You cannot use that last route for content a records authority already covers. Your website is content a records authority covers.

Step 1: Check whether the National Library of Australia harvested your website

The National Library captures Australian Government websites. When the harvest works, the National Archives guidance says you have nothing about the website left to send the Archives.

Check the harvest before you plan anything else, because it sets which class you sentence the site to.

Is your site missing from the Library's list of archived websites? The guidance says to tell their web archive team, so it goes into a future harvest. Do that now, not in the week you switch the site off.

A third path opens if the Library will not get to it in time. The guidance says to capture a copy with open source software, and to manage it as material a records authority covers.

Step 2: Find what the harvest missed

The capture is not complete. The National Archives says the harvest does not always pick up videos, PDF files and Word documents. Those may need managing outside the Library process.

Attachments are the gap, and attachments are where a lot of government service content lives.

Read that against your own site. Think of the forms, the fact sheets, and the guidelines nobody has opened since the last redesign. Those files are the least likely to survive a harvest. They are often the clearest record of what your agency told people.

One more thing worth knowing early. The Library does not hand over copies of the sites it preserves. If you need a copy for your own business use, take it yourself.

Step 3: Treat the page as a container, not a single record

A page is rarely one record. The National Archives says published content covered by a records authority has to sit in an approved system. It also says a class for specific content should take precedence over website content management.

A policy or a speech on your site should keep its own disposal rules, whatever happens to the page around it.

So the retirement question splits. Ask what happens to the site, then ask what happens to the things published on it. A clean harvest can settle the first and leave the second open.

Step 4: Sentence the site against the right records class

Sentencing means matching a record to a class in a records authority. The class sets what happens to the record, and when. Agency public websites sit under the Publication function of one records authority. Its name is the Administrative Functions Disposal Authority, Express Version 2.

Class 62616 covers agency public websites the National Library has harvested. You can destroy those once you decommission the site and confirm the capture.

Class 62615 covers agency public websites it has not harvested. Those count as national archives first. You send a copy to the Archives, then you can destroy the site once you confirm the transfer.

Both classes can end in destruction, so the harvest sets the work in between rather than the outcome.

Neither class is a free pass. A record under a disposal freeze must not go until the freeze lifts. Nor can one wanted for a court matter or an access request.

Step 5: Get the decision made before the site comes down

The National Archives is direct about the order. An agency has to confirm the Library captured the site before it decommissions that site or changes it heavily. Capture first. Decommission second.

Getting that order wrong is the failure, because a switch-off you cannot reverse is a disposal you cannot evidence.

In my experience the sign-off is not the content team's to give. Find whoever holds records decisions in your agency. The records authority points you to your agency's information manager for advice on applying it. They will want the harvest evidence, the class, and the date you plan to turn the site off. Book that talk into the project plan, not into its final week.

What good records management looks like on a website

The team knows whether the Library harvests their site, and checks rather than assumes. The team treats attachments as records in their own right. Every retirement carries a class, evidence of capture, and a named person who signed it off.

A retirement that cannot show its capture evidence is a disposal nobody authorised.

When your team last retired a section of the site, who confirmed the capture, and where is that confirmation now?

Photo by Salomé Guruli on Unsplash