Site audit

Use the Site page to check how your website looks to AI crawlers, and to see every page you have in one tree.

Site audit overview

The Site page has two parts. Your tracked domains sit at the top — the primary domain is the one that gets crawled, and it’s the input to everything below. Under that is the audit itself: the domain-level result, then the full site schema.

The audit runs automatically on a weekly schedule, and you can run it yourself at any time. It crawls your site using your sitemap, plus following links when there isn’t one, then runs a set of checks on each page it finds.

There are two levels. The standard audit runs on every audit, is quick, and doesn’t use any quota — it reads your pages’ HTML directly. The deep audit runs on request, on the pages you choose, and uses your monthly allowance — it renders the page like a browser would, measures speed, and analyses the content. Only the deep audit produces SRPS+ and PERF.

How to run an audit for the whole site

Click Run audit for site in the top right of the domain card. While it runs, the button becomes Updating…, and the header shows how many pages have been found and how many are done.

A full run replaces everything: every score, every verdict and the page tree. That’s why the old scores disappear while it works rather than sitting there looking current.

The header also tells you when the last audit ran and when the next scheduled one is due.

Good to know Your plan sets how many pages one crawl covers. If your site has more, the audit says so — “Audited the first N pages” — and what you’re looking at is a partial view, not your whole site.

How to audit specific pages

You don’t have to re-run the whole site to re-check one page.

1. Find the pages in the site schema tree and tick them.
2. A bar appears at the bottom of the screen showing how many you’ve selected.
3. Click Run audit for a standard re-check, or Run deep audit for the deep one.

Clear deselects everything, and so does pressing Escape.

You can also run either level from inside a single page’s audit — see Page audit.

Pro tip Deep audits use your monthly quota one page at a time, so spend them where they’ll tell you something: your homepage, your main product or service pages, and anything you’ve published from a content recommendation. Re-auditing a page refreshes its deep result.

How to find a page in the site schema

The tree groups your site by language first — read from your hreflang tags, with the default language at the domain root — and then by URL section, using the real path segments like /blog or /glossary.

Four filters sit above it: All, Non-200, AI-cited and Deep-audited. Use Non-200 to find broken pages, and AI-cited to see which of your pages assistants have actually quoted. Expand all and Collapse all open and close the whole tree.

Click any page to open its full audit.

How to read the domain result

The domain card carries the site-wide checks — the ones that don’t belong to any single page. They’re grouped in the order a crawler meets them: reaching your site at all, letting AI bots in, being indexable, being discoverable, having one canonical address, and stating who you are. The social card sits on its own below.

Each check shows one of five states: OK, Needs work, Failing, No data, or Manual check. A manual check is one only you can confirm — verify it yourself, then click Mark done.

Below the tree you’ll also find any broken links across the site, and links the audit couldn’t verify.

For what the score itself means, see SRPS.

Good to know Changing your primary domain wipes this brand’s site audit and starts a fresh crawl, and the number of changes is limited per month. It isn’t a setting to experiment with.