Page Experience
Two checks in one place: how fast your pages are for real visitors, and whether Google can index them at all.
In appInstrumentsPage Experience
Speed & Vitals
Core Web Vitals for a page or a whole section of the site: how quickly the main content appears, how soon the page responds to a tap, and how much it shifts around while loading.
The audit names the specific cause rather than the score. Images with no width and height set, so the page jumps as they arrive. An animated GIF doing the job of a video. The main image at the top set to lazy-load, so it waits its turn when it should be first.
Two AI actions sit on the results: Explain the vitals in plain words, and What should my developer do first? The second is the useful one if you're handing this to someone else.
The pattern worth looking for is a template, not a page. Twenty slow pages sharing a layout are one fix, not twenty.
Index Coverage
The other half, and the one people don't expect. It asks Google directly what it thinks of your pages, then groups the answers into problems you can act on.
Roughly eighteen groups, in the words the industry already uses: Redirect loops, Broken redirects, Redirect chains, Broken internal links, 404 pages with traffic, Redirects in the sitemap, Noindex in the sitemap, Canonical loops, Broken canonical targets, Duplicate content, Hreflang return tags, Excluded pages in search, Few internal links, Slow server response and Crawl coverage.
404 pages with traffic and Excluded pages in search are the two to look at first on most sites. Both mean a page that search already values isn't reachable or isn't eligible, which is the cheapest kind of problem to fix.
It describes what the issue is rather than guessing why: "excluded by noindex", not "hidden by mistake". The data can't see intent, so plenty of these will be deliberate, and the ones that are a judgement call are worded as a review rather than a fix.
Who fixes what, and how long it takes
Every finding carries two chips: who it needs and how much effort it is. Who is developer, content team or anyone. Effort is copy-paste, quick edit or dev work.
That's what makes this list sortable by something other than severity. Filtering to copy-paste and anyone gives you the things you can clear this afternoon without asking anybody.
Items that are a judgement rather than a task don't get either chip. They read your call instead, because calling a deliberate noindex a task would be wrong.
When the checks run, and the daily limit
The first index check starts on its own once the crawl finishes, and then runs every night.
Google allows 2,000 page checks a day for a site, and Searchscope keeps 200 back so you can still check a page yourself after a nightly run has used its share. On a large site that means coverage builds up over several nights rather than arriving all at once.
Handing it to a developer
Redirect findings can be exported as rules for nginx, .htaccess or Vercel, and any list exports as CSV.
So the output of a session here is usually a file someone else can apply, rather than a list they have to re-type.