Get started
How Lookout checks a site
A scan reads your site’s pages and records what it finds, with dates. This page explains what it reads, its limits and when a problem counts as fixed.
What a scan reads
A scan starts from your home page, your robots.txt and your sitemap, reads the pages your sitemaps list, and follows the internal links it finds. It reads the HTML your server sends. It doesn’t run your page’s JavaScript, so content that only appears after scripts run isn’t checked.
It reads two pages at a time to start with, and up to four once your site has answered a run of requests without trouble. When your robots.txt sets a Crawl-delay, it reads one page at a time and waits that long, up to 10 seconds, between requests. When your site asks it to slow down, or a request times out, it tries the page again a few times before recording a failure.
What it checks
This list is the same one each project shows under Project settings, What a scan checks.
- Limits: up to 100 nested sitemaps, 7,500 sitemap locations, 7,500 crawled URLs and 7,500 internal link destinations, in durable batches every minute.
- Discovers linked pages and checks metadata, canonical targets, noindex, duplicate metadata, sitemap health, internal links, hreflang declarations, basic JSON-LD syntax and the properties Google requires for merchant listings on Product markup.
- Follows redirect chains and loops for up to 10 hops inside the property, counted in the 7,500 URL limit.
- Checks H1 headings and main-content length; the length check is advisory and flags indexable sitemap pages under 150 words.
- On-page checks cover image alt text, insecure subresources, viewport, link names, HTML size, language, nofollow and metadata length, and up to 200 in-property images get a HEAD size check.
- Advisory checks are suggestions: they never change health or send alerts.
- Limit or request failures make a partial scan and keep open findings. Pages shows dated evidence; older snapshots do not establish passes for new checks.
- Googlebot crawl permission does not confirm indexing. Lab performance and optional restricted rendered comparisons are separate evidence.
When scans run
| Schedule | What happens |
|---|---|
| Daily | About once a day. New projects start on this schedule. |
| Weekly | About once a week. |
| Manual | Only when you choose Run scan. |
When a problem counts as fixed
A problem is marked as fixed in one of two ways:
- A later scan checks the same URL again and the check passes.
- A complete scan no longer finds the URL: no sitemap lists it and no crawled page links to it. Lookout can no longer check it, so the problem closes even though no check passed. If the page still exists, check it yourself.
Each scan is compared with the last complete one, so you can see what got worse and what was fixed. A broken link is confirmed by a 404 or 410 response. Lookout checks it again on every scan, even once nothing links to it, and only a 200 response clears it: a redirect or a timeout doesn’t.
Pages counts and lists the URLs the latest complete scan recorded. A URL it has no evidence for, such as a page no sitemap or link leads to any more, or one too big to read, moves to a separate list at the foot of the table, with the date of the last scan that checked it.
What a scan doesn’t claim
- It can’t find a page that no sitemap lists and no crawled page links to, so it doesn’t claim to have seen every page on your domain.
- A passed check confirms what Lookout saw on your site, not that Google has re-crawled the page.
- It never changes your website.