Crawl your site with Site audit

In shortSite audit crawls your whole website and checks every page for technical SEO problems: status codes, indexability, canonicals, titles, meta descriptions, headings, duplicates, thin content, images, links and schema. Open SEO → Audits → Site audit, switch it to Enabled, click Start Scan, then filter the URL-by-URL results and export a CSV. You pay per page crawled, and the crawl stops when your balance runs out.

What Site audit is for

Use Site audit when you want to find problems across a whole site rather than on one page: a batch of 404s after a migration, pages accidentally set to noindex, missing canonicals, duplicate titles, or thin pages nobody remembers. It works like a desktop crawler, but runs in your account and keeps every crawl in a history list.

It's a technical SEO crawler. It doesn't score AI readiness the way the AEO and GEO checker does, and its scores can't be compared with other audits. See Which audit to use.

Before you start

• Select the domain in the top bar. Without one, the page shows Select a Domain.
• Have enough credits for the pages you expect. The panel shows your balance and roughly how many pages it covers.
• Know the limits. By default the crawl stops at 500 pages, and Max Depth is 3 link levels. These show under Settings but can't be changed on the page yet; the panel says to contact support to adjust them.

Step by step

1. Open Site auditGo to SEO → Audits → Site audit. The page heading reads SEO Auto Audit with a Beta badge.
2. Enable itIf you see SEO Auto Audit is disabled, click Enable Now or the Disabled button at the top right. It switches to Enabled and the scan controls appear. Enabling doesn't start a crawl or cost anything.
3. Check the settingsClick Settings to open Scan Settings with Max Pages, Max Depth and Respect robots.txt for this domain. Max Depth only matters when the crawler has to follow links (see the next step). Respect robots.txt shows Yes, but today the crawler still fetches, charges and lists pages your robots.txt blocks; it flags them in the results instead of skipping them.
4. Start the crawlClick Start Scan. The status moves from Starting… to Discovering URLs… to Scanning…, with a live page count. Discovery reads your sitemap first. It follows links from the home page only when the sitemap gives fewer than 5 URLs, so a page that isn't in your sitemap may not be crawled. Only one crawl can run per domain at a time.
5. Stop early if you need toClick the stop button and confirm Stop scan?. Pages already crawled stay in the results and stay charged.
6. Open a crawl from the historyScan History on the left lists every crawl with its status: Completed, Stopped, No Credits or Error. Click one to load its summary and results.
7. Filter the resultsUse Filter by URL… to search, or the quick filters 4xx Errors, Not Indexed, Missing Title, Missing H1, No Schema and Thin Content. Click a column header to sort. Use Manage Columns to show the columns you need, such as canonical, redirect chain, word count or internal links.
8. Export a CSVClick CSV above the table to download every row for that crawl, for a spreadsheet or a developer ticket.

How to read the result

When a crawl finishes, the panel shows Avg Score out of 100, Critical, High and Pages. Open the crawl from Scan History for the full summary: Medium issues plus quick counts for Missing Title, Missing H1, No Schema, Thin Content, Missing Meta, Broken Links, Indexable, Non-Indexable and Errors.

• Start with Not Indexed. A page you want found that is non-indexable (noindex, blocked by robots.txt, or canonicalised elsewhere) can't rank or be cited.
• Then 4xx Errors. Fix or redirect broken URLs, and update the internal links pointing at them.
• Then titles and H1s. Missing or duplicate titles are quick wins.
• Thin Content and No Schema last. These need writing time, so pick the pages that matter most.

What it costs

Site audit charges for each page as it's crawled, so the total depends on the size of your site and the page limit. The crawl stops on its own when your balance can't cover the next page, and the history shows it as No Credits. The pages crawled so far are kept.