The Structure tab draws your site as the folders its pages sit in: how many pages each folder holds, which of them have errors, and how many clicks away from the start page they are. It reads what your latest scans already found, so it costs no credits. The view is on every plan; the download is on Starter and up.
Open Site Crawler, then Structure, and pick a site from My Sites. You need at least one scan of that site. When a scan ends, its pages are first added to Site Pages, and the structure is built after that. Until the pages are in, a site with no earlier scan can still show "Scan this site first". You do not need to scan again: open the tab again a little later. While the folders are counted, the tab says "Building the structure". For a big site this takes a minute or two.
What is counted
Every number counts the same pages as the Pages view of Site Pages: the site's own HTML pages that the latest scan found. Images, scripts, PDFs, links to other sites and pages no scan found any more are not in it, so the total at the top matches the Pages count in Site Pages.
A folder is a part of the address. A page at /blog/news/post counts in the whole site, in /blog, in /blog/news and in itself. A query string (?page=2) never makes a new folder: the page counts in the folder of its path. /shop/index.html is the page of /shop.
Folders
The tree lists each folder with its pages, its share of the site, how many pages work, redirect, are broken or were not checked, how many are not indexable, how many have errors or warnings, how many are weakly linked (one link in, or none) and how many are deep. Open a folder to see its subfolders; a folder name opens its pages in Site Pages. The tree works with the keyboard: arrow keys move, Right and Left open and close a folder, Enter opens its pages.
- The path only label marks a folder with no page of its own: nothing lives at the folder's address, it only groups the pages below it.
- The ids grouped label marks a folder whose name stands for many: numbers, years, dates and long ids in an address are grouped into one folder, so
/product/101and/product/202do not make hundreds of folders. It opens the nearest real folder in Site Pages. - A green +12 or red −4 next to a count is the change in pages since the previous result.
Errors and warnings need a detailed scan
Errors and warnings are known only for pages a detailed scan read. A folder none of whose pages was read shows a dash and "Needs a detailed scan": that is not the same as "No errors found", and it is never drawn as clean. A fast scan still fills in every other column, with one limit. Not indexable counts pages that redirect or are broken. A page with noindex, or with a canonical pointing to another address, is counted only after a detailed scan has read it.
Depth
The Depth view shows how many clicks it takes to reach each page from the start page, as a chart (pages per depth, split by status) and as a grid of the biggest folders by depth. The OK part of every bar deeper than four clicks, and the grid cells at those depths, are drawn in the warning colour: search engines tend to visit deep pages less often. Not linked holds the pages no crawled page links to, and the pages whose depth was never measured. Every bar, and every cell in the row of a folder, opens those pages in Site Pages. The Other row of the grid holds the pages outside the biggest folders; its cells only show numbers. Show as a table gives the same numbers as text.
Depth and inlink counts come from the newest scan that crawled the whole site. A scan that covered only part of it leaves them as they were, and the line above the view says which scan they come from. Until one complete crawl exists, depth is not measured, so the Depth view shows every page as Not linked.
Treemap
The Treemap draws folders as rectangles: the area is the number of pages, the colour is the share of measured pages that have errors. A hatched rectangle means nothing was measured. Select a rectangle to open its pages; the arrow keys move between rectangles. Show as a table lists the same folders as text, with their change since the previous result.
Big sites
The structure is built after each scan and kept, so the tab opens at once however big the site is. It lists up to five levels of folders, at most 200 entries under one folder and about 2,000 folders in all. Folders beyond that are grouped as Other (1,204 pages in 310 folders), and the totals still add up to the whole site. Pages deeper than five levels count in their fifth-level folder.
Is it up to date?
A line above the view says which scan the pages come from. While a new scan is being added, the tab says it is updating and refreshes by itself; if a rebuild ever fails, the tab keeps the previous result and says so.
Download
Export gives the folder tree as CSV or Excel, one row per folder: path, level, pages, share of the site, status counts, issue counts, depth and the pages the folder had in the previous result. Paths that start with an equals sign are written as text, so a spreadsheet never runs them. Exports come with the Starter plan and up.