What changed?
Microsoft Clarity announced AI Scrape-to-Referral Insights on August 13, 2026. The announcement describes bringing Bot Analytics and Clarity behavioral analytics together with an AI Scrape-to-Referral Ratio card and links into session recordings.
The stated use is to compare AI crawler activity with referral visits and inspect the behavior of recorded AI-referred sessions after they reach a site.
Why it matters
Crawler requests and referral visits describe different steps. A combined view can help teams investigate whether bot activity and observed AI-platform referral sessions appear together in their reporting, then inspect recorded sessions where available.
A ratio is a directional comparison, not proof that a particular crawler caused a particular visit. It cannot establish that all AI crawling is represented, that all AI referrals are observed, or that the observed sessions represent every AI-assisted journey.
Who is affected
Clarity users who review AI crawler activity and website sessions may use the insight to form questions about content discovery and referral behavior.
Measurement impact
Bot activity is evidence of requests attributed to crawlers under the product’s detection and reporting rules. Referral sessions are a separate signal based on visits that reach the site and are recognized in analytics. The combined view does not by itself create a deterministic crawler-to-user identity link.
Session recordings can provide context for a recorded visit, subject to Clarity collection, consent, and privacy configuration. They do not reveal which unobserved AI interactions took place before the visit.
What to check
Treat the card as an investigation aid and validate the separate signals behind it.
- Review how Clarity defines the AI crawler activity and referral measures shown in the view.
- Compare equivalent date ranges and confirm that site, filters, and measurement scope are consistent.
- Inspect available session recordings as individual examples, not as a complete or representative census.
- Compare recognized referral sources with GA4 acquisition data while keeping each system’s definitions explicit.
- Do not interpret crawler volume as human traffic, conversion intent, citation, or causal impact.
Recommended action
Use scrape-to-referral reporting to prioritize investigation—for example, identifying content sections where observed bot activity and recognized referral visits differ. Validate each layer independently, retain unknown or unattributed journeys as unknown, and report the result as directional rather than causal.
Original source
This update is based on the original announcement from Microsoft Clarity. The summary and measurement recommendations above are editorial analysis.