Changes and diffs

Every crawl is compared against the previous one, so you can see exactly what changed on a page and when.

Why this is the useful view

A list of everything wrong with a site is mostly historical: it has been that way for a long time and nobody is going to fix it this week. A list of what changed since yesterday is actionable, because something caused it and that something is recent.

Most real incidents look like this: nothing was announced, a plugin updated, and forty pages quietly changed their canonical.

What is compared

Status codes — a 200 that became a 301 or a 404.
Title, description, canonical, robots — the metadata that decides how a page is indexed.
Content — headings and body text, so a truncated or emptied page is obvious.
Structured data — schema that changed or stopped validating.
Links — links added, removed, or newly broken.

Reading a diff

Open a page’s change and you get before and after, marked inline. Sorting by change type answers “what happened to the site”, where sorting by page answers “what happened to this page”.

The pattern to look for is many pages changing the same way at the same time. One page losing its meta description is an editing mistake. Two hundred pages losing it overnight is a deploy, a plugin update, or a template change — and that is a single cause with a single fix.

Common culprits

A plugin update changing how metadata is generated.
A theme deploy altering the template.
A migration rewriting URLs, leaving canonicals pointing at the old ones.
A staging flag shipped to production — the classic being a site-wide noindex.
A CDN or optimisation layer rewriting HTML in transit.

From a change to a fix

If the change is content, the page usually has a Workover document behind it: fix and push. If it is structural, it is a theme or plugin setting, and the diff’s timestamp tells you which deploy to look at.