Move beyond aggregate bot counts. Group crawler events into sessions, inspect page sequences and inferred intent, and see which Brand Intelligence version each request actually received.
Human traffic stays unchanged · Bot verification is labeled by integration
Traditional server logs scatter a visit across rows. Aggregate dashboards compress it into requests per day. Neither view shows the path through your site, which artifact was returned, or whether the traffic can be treated as verified. AI Sessions keeps those distinctions visible.
Crawler analytics is most useful when it shows both the behavior you observed and the conclusions the evidence cannot support on its own.
Record detected AI bot requests at the point where they reach your integration. Keep the crawler identity, requested URL, response, and delivery context together as first-party evidence.
Turn isolated log lines into a sequence of related activity. See the pages visited, the order of requests, session duration, and an explicitly labeled intent inference.
Connect a request to the route and Brand Intelligence artifact served at that moment. Delivery evidence helps distinguish no crawl, wrong route, stale version, and successful receipt.
The data plane serves from the edge and records the event asynchronously. Session building happens later, so logging does not add a control-plane dependency to the live crawler request.
Identify a supported AI crawler from request context and apply the verification rules available for the integration method.
Resolve the requested route and return the appropriate origin response or published Brand Intelligence artifact.
Group related events into a visit sequence and calculate the pages, timing, and observable behavior of the crawl.
Review the session alongside Tracker results and published versions without treating access as proof of citation.
A claimed bot identity, a successfully delivered artifact, and a later model answer are related but different observations. Ooky preserves each layer so teams can reason about where a problem occurred.
Intent is inferred from observed request behavior. It is labeled as inference, not presented as a fact about the crawler provider's internal purpose.
See the detected crawler and whether the integration can support verified classification.
Review related URLs, timing, response codes, and the inferred purpose of the session.
Inspect whether the response came from the origin or a particular published intelligence version.
Use Brand Tracker for answer evidence rather than assuming a successful crawl guarantees a citation.
Need to discuss your stack or evidence requirements? Book a product demo.
An AI crawler session is a group of related bot request events that represents one observed visit sequence. It can include the crawler identity, pages requested, timing, responses, and the content artifacts delivered during that sequence.
No integration can make the same verification claim. Ooky's managed Full DNS path can mark traffic verified under its stricter ingestion path. SDK, WordPress, and customer-managed Worker events remain unverified claimed user-agent traffic, which is shown as a different evidence class.
No. A visit proves that a crawler requested content and, when delivery records are available, what it received. Training, indexing, retrieval, citation, and model-answer behavior happen later and are controlled by the AI provider. Brand Tracker measures the answer separately.
AI Sessions is designed around detected bot traffic. Human traffic continues to the origin unchanged. The product view focuses on crawler events and the sessions built from those events rather than ordinary human web analytics.
Managed Full DNS provides the fullest request-path context and can support verified-bot classification. SDK, WordPress, and customer-managed Worker paths are still useful for observing requests and delivery, but their authentication model intentionally does not assert that a claimed user agent is verified.