One page, one row
A page counts once, whatever tracking parameter,www. or trailing slash its URL carries. So https://www.example.com/pricing/?utm_source=chatgpt.com and https://example.com/pricing are one row.
A path whose first segment starts with _, such as /_next/ or /_mintlify/, is an endpoint of a site vendor. A reader never opens it, so the page lists none from any source.
Every crawler fetches robots.txt and the sitemap files, and neither is a page to improve. The page lists one only when an answer cites or reads it.
How the page counts
The answer columns count analyzed answers in the scope, one answer once per page. Every filter but the competitor filter applies to them, and the period sets the window. The crawler requests, AI sessions and key events columns follow only the period, never an answer filter. The logs and Google Analytics know no prompt, topic or persona. Their window ends on the newest day with analyzed answers in the project. A filter never moves it, and the crawled pages stay listed when no answer matches the filters. So these columns can differ from the counts on the Crawlers and Overview tabs. The windows there end on a day that their own source sets. Four tiles count the answers:
A page of a subdomain counts as a page of this site.
The columns
A session whose landing page Google Analytics does not name, or names off the project’s website, gets no row. Neither does a landing page address too long to store.
The table lists the 200 most cited pages. It adds up to 200 more from each of the logs and Google Analytics. The line under the title counts every page in scope, listed or not. A note under the table says when the table stops short.
Find a page searches the paths and the titles. A click on a column head sorts the table by that column. A search and a sort act on every page in scope, not only on the listed pages. The table then lists the first 200 pages in that order. The Pro plan and above list every page in that order.
Export CSV saves every page in scope, up to 10,000 lines, in the table’s order. A file that stops at that cap keeps the most cited pages first. A note beside the button then says so.
Pages or folders
Rows switches between Pages and Folders. The Folders view groups the pages by the first segment of their path, such as/blog/. A homepage is in /. Each folder row counts its pages. A folder counts an answer once, whatever number of its pages the answer used. The crawler requests, AI sessions and key events of a folder are the sums of its pages. The Folders view shows no Last seen column, and Export CSV then saves the folders.
What the columns do not claim
- Only some platforms report the pages they read without citing them. The page names those platforms under its answer, and it shows the “Read, not cited” column only when one of them is in scope.
- The crawler, session and key event columns show only when their source delivered data for the window. Otherwise the page shows no column, because a zero would claim a measurement. A note under the table then names the missing column and links to Settings › Project, where its source connects.
- A source that began inside the window measured none of the days before. A note over the table then names the first day of that source.
- A page that crawlers request and no answer cites still gets a row. That row is often the one to read first.