Skip to main content
Use AI Referrals to review website visits attributed to AI assistants through your connected analytics data. This measures observed referral traffic, which is different from how often a brand appears in monitored AI answers.

​
Connect the right analytics property

Follow the analytics connection flow for the website you want to analyse. Confirm the selected property and date range, especially when one Google account has access to several client properties. Review source, landing-page, and engagement information together. A visit count describes volume; landing pages show where visitors arrived; engagement metrics help you understand what happened afterwards.

​
Understand attribution limits

Not every visit from an AI experience carries an identifiable referrer. Visits can be grouped differently by analytics settings, browser behaviour, and the way links are opened. Treat the view as traffic attributed to recognised AI sources, rather than a census of every AI-influenced visit. Do not equate an increase in monitored mentions with an increase in site traffic. An answer can recommend a brand without generating a click, and a referral can arrive from a question outside your monitored set.

​
Combine the signals

Compare referral movement with visibility and citations over equivalent periods. Use the combination to choose an investigation, then inspect the relevant pages and answers before assigning a cause.