Page timeline

The page timeline reads your most important pages once a day and keeps a snapshot whenever one changes, like a commit history for your live website. Open Page timeline in the organisation menu to see each page, when it last changed and exactly what changed. It uses no AI and no analysis units.

UX Robot chooses the pages for you, up to 25 per organisation:

• the home page from the organisation profile;
• every page with an open issue or one shipped and waiting for a check;
• the starting page of each journey, newest first.

Only pages on the website address in the organisation profile, or its subdomains, are read. A page that stops being chosen, for example because its issue was verified fixed, keeps its history but is no longer checked. Pages are read while the organisation has a website address and an active subscription.

Select a page to see its timeline, newest first. The first entry is the first snapshot; each later entry says between which two daily checks the page changed and shows:

• the old and new title and description, and a new status such as a page that now returns 404;
• headings added or removed;
• wording added or removed, one line per paragraph, list item, button or field label, including placeholders and the accessible names of icon buttons.

An edited line shows as its old version removed and its new version added. Moving a block around the page is not counted as a change, and hidden elements are ignored.

When GA4 is connected, each change is put next to the page’s figures from the 14 days before it and up to 14 days after. Days on which the change may have happened are left out of both sides, and the latest two days wait until GA4 has finished processing them, so the first figures appear once three full days have passed. The comparison fills in daily until the 14 days after are complete.

• Views per day: GA4 page views divided by the number of days.
• Engagement rate: engaged sessions that included the page, out of all sessions that included it.
• Key events per session: key events on the page, per session that included it.

Countries you exclude from evidence in the organisation’s data settings are excluded here too.

Many issues come with an after picture that proposes exact wording, such as changing “Call” to “Call 03 9650 5525”. After each daily read, UX Robot looks on the issue’s page for that wording. When all of the proposed wording is there, and any text it replaces is gone, the issue’s history records New wording is live, with a link back to the page.

• It applies to open issues and issues shipped and waiting for a check, so it also catches a fix that went live without anyone marking it shipped.
• It never marks an issue fixed. Run a spot check, or run the journey again, to verify the fix with a screenshot.
• It is recorded once. If a later check finds the issue still there, partly fixed or back again, the wording is looked for afresh.
• Proposals that only change the layout have no wording to look for.

Select Check now on the Page timeline to read every chosen page straight away, for example after a release. It is free and can be used once every ten minutes; changes appear on the timeline within a few minutes. Coding agents can follow up with request-ux-issue-verification once the change is live.

• UX Robot reads public pages as a visitor, never signed in, and identifies itself as UXRobot/1.0. If your firewall blocks unknown bots, allow that user agent.
• It reads only the organisation’s website and its subdomains, on the standard web ports. Every redirect is checked, and an address that points to a private network is refused.
• It stores each snapshot’s title, description, headings and readable text, never screenshots, cookies or form submissions.
• A page that cannot be read shows why on the timeline and keeps its last snapshot.