The Tracking page gives you two ways to look beyond crawler totals: a weekly activity grid and a detail view for each page. You can see when AI requests cluster, then open a page to check which bots fetched it and whether your server let them through.
See the pattern across the week
The “When AI is active” grid shows the last four weeks by weekday and hour, in your local time. Switch between recorded visits from AI assistants and live retrieval requests. The latter are fetches made while an assistant answers a question; training and search-indexing traffic stay out of this view.
Look at a particular page
Click a page path in Crawler activity to open its details. The drawer shows requests over your selected period, the bots behind them, and any access errors. A chart puts crawler activity alongside recorded AI referral visits, so you can inspect both for the same page.
The purpose labels separate training, search indexing and live retrieval. Verification shows which requests could be matched to a vendor’s published addresses; not every bot can be verified. A request using an AI bot’s name is not, by itself, proof of its identity.
These views show recorded requests and attributable visits. They cannot prove that a fetched page was used in an answer, and a flat referral line does not prove that no one arrived from AI. For pages with repeated retrieval activity, prompt suggestions based on crawler activity offer a way to investigate gaps in what you track.
Live tracking is included on paid plans and during a trial. Crawler data needs the WordPress plugin, Cloudflare edge tracker or NGINX server tracker; the browser snippet records visitors only. See the tracking setup guide to connect your site.
