What does the site scan check, and what changed on my site?

A site scan reads every page it can reach, up to your plan's page limit, and checks each one. Since October 2026 it also checks the things the big desktop crawlers check, written for a small business: what to do, not just what is wrong.

#It obeys your robots.txt

The scan reads robots.txt first, as Google does. Addresses robots.txt asks crawlers to stay out of are not read; your homepage always is, so a site that blocks everything can still be told so.

• Pages skipped because robots.txt blocks them lists those addresses with the line that blocks each one. A note, never counted against your score: blocking the basket, account and admin pages is good practice.
• Important pages blocked by robots.txt appears only when Google is kept out of a page your site points people at (in the main menu, or linked from five or more pages).
• Sitemap lists pages blocked by robots.txt when the two files contradict each other.
• If robots.txt asks for a Crawl-delay, we wait up to two seconds between pages and read one at a time, so the scan takes longer, and say so.

#It finds your sitemap where robots.txt says it is

The scan looks at the Sitemap: lines in robots.txt first, then the usual addresses (/sitemap.xml, /sitemap_index.xml, /wp-sitemap.xml). It reads every sitemap robots.txt names, follows sitemap indexes up to 50 files, and unpacks .xml.gz files. The sitemap card on Website check reads it the same way, so the two always agree. If robots.txt points to a sitemap on another website, you are told.

#Google's own answer: is the page indexed?

With Search Console connected, the scan asks Google about your most-seen pages and the main pages Google has never shown, through Search Console's URL Inspection. You see:

• Google has not indexed this page, with Google's reason in its own words,
• Google read this page but did not index it ("Crawled - currently not indexed"),
• Google chose a different canonical, with the address Google shows instead,
• Google last found this page blocked, when the block has gone but Google has not been back.

We use at most 500 of Google's 2,000 inspections a day for your property, keep each answer for a week, and fill in the rest over the following nights. The notes under What was crawled say how many pages were checked.

#Accessibility basics on every page

Read from each page's HTML: zoom switched off, headings that skip levels, empty headings, maps and videos with no title, image buttons with no text, no main landmark, and the same link text going to several pages. They are listed to help you, and never change your score.

#Structured data, all three kinds

JSON-LD, microdata and RDFa are all read and checked against Google's rules for local businesses (plumbers, dentists, restaurants and the rest), organisations, products and offers, reviews and ratings, FAQs, articles, breadcrumbs, events and recipes. Missing required fields stop the rich result; missing recommended fields appear as Structured data could say more, a low tidy-up.

#Pages opened in a browser

Each scan opens one page of each kind (homepage, a service page, a blog post, a product...) in a browser, plus any page that looks empty, and compares it with the page as your server sends it: title, canonical, main heading, meta description, links and text. Your plan sets how many: 25, 50 or 100 pages.

#Security and images

The homepage's protective headers, pages served over HTTP, links that open a new tab without rel="noopener", images over 100 KB, images without width and height, and a main image that is lazy-loaded. Speed findings stay out of the score.

#What changed on your site

On Website check, What changed on your site compares this scan with the one before: pages added, pages no longer linked, pages newly broken or redirected, and titles, main headings, meta descriptions, canonicals and noindex that changed. Show the pages lists each one with before and after. The scan email lists the most important changes too.

Pages are counted as added or gone only when both scans read a similar number of pages, and never as gone when the scan stopped at its page limit.

#Filter by section and kind of page

On Problems, pick a section of your site (its first folder, such as /blog/) or a kind of page (blog posts, products, services, locations...) to see only their findings. The choice stays in the page address, so you can bookmark it or send it to someone.