Top pages shows how each indexed page appears across four different measurement sources. Open Dashboard → Analytics → Top pages, or use Top pages from Knowledge Base → Site pages.
What's at the top
- The briefing: a dark card with the headline counts, a short summary of what they mean (which page leads, where to start), and a few key facts.
- Page counts: indexed pages, pages cited by AI, pages with AI visits, pages with Google clicks, and pages read by AI crawlers. Each count names its own source window. A source that isn't connected or measured says so instead of showing a number.
- Most cited in AI answers and Most AI-referred visits: the top five pages for each, across every page that matches your search.
Search and the 30- or 90-day crawler window sit above the briefing and apply to everything on the page.
What each column means
| Reading | Meaning | Source |
|---|---|---|
| Google search | Clicks, impressions, click-through rate, and average position for the page | Google Search Console |
| AI citations | Measured AI-answer rows in the latest usable full scan that cited the exact owned page | Stride scans |
| AI sessions | Sessions from the 15 AI assistants that AI traffic counts, that landed on the page path | Google Analytics 4 |
| AI crawler requests | Verified, claimed, and failed autonomous crawler requests for the page path; user-triggered fetches are excluded | Your connected agent-log sources |
These readings are shown together for comparison, but they are not added into one total or score. A citation, session, search click, and crawler request are different events.
Bars beneath clicks, citations, and AI sessions make relative strength easier to scan. Each metric scales independently against the strongest value in the complete filtered result set, and the leading page in each column is highlighted; the number beside the bar remains the exact total.
Source windows and unavailable values
Each source keeps its own observed window. Search Console page readings are stored monthly, AI sessions use the last 28 complete days of the daily Google Analytics read (an older monthly page reading until that daily read is ready), citations use the latest usable full scan, and crawler requests use the selected 30- or 90-day range. The page counts at the top show the exact window being used.
A dash (—) means the reading is unavailable for that page. It does not mean a
measured zero. This can happen before the first source sync, when a connected
source needs attention, when a sparse provider result did not include that
page, or when a path could match more than one indexed URL. Near-complete
crawler sources show 0 when their measured window contains no autonomous
crawler request for an unambiguous page path. Customer-reported and sampled
sources remain — when that page has no reading. A citation value of zero is
shown only when a usable scan measured answers and none cited that exact page.
Why crawler sources stay separate
Cloudflare Worker, manual API, and Vercel Log Drain readings appear on separate lines. The same request can reach Stride through more than one transport, so adding them together could double-count activity. Coverage labels remain source-specific: near-complete, customer-reported, or sampled.
Crawler data is path-only. If two indexed URLs differ only by query parameters, Stride withholds path-only GA4 and crawler readings rather than assigning the same activity to both pages.
How URL matching works
Stride matches protocol and www variants, trailing slashes, fragments, and
known tracking parameters such as utm_source. Meaningful query parameters,
path and query casing, and non-default ports remain part of page identity, so
genuinely different page variants are not collapsed.
Do not read the table as attribution
The page view shows signals observed for the same page. It does not claim that an AI citation caused a search click, session, or crawler request. Use it to find pages worth investigating, then judge each source in its own window and context.