Reading AI Referral Traffic
When someone clicks a citation in an AI answer, the visit often arrives with a recognisable referrer. Separating those visits shows which assistants send you people and what those people do.
Recognise the referrers
Common sources include chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and claude.ai. Some assistants add a utm_source parameter to links. Visits from Google AI Overviews arrive as ordinary Google organic traffic and cannot be separated from other organic clicks.
Set up a channel in GA4
In GA4, create a custom channel group with an "AI assistants" channel that matches those referrer domains with a regular expression, placed above the default Referral channel. Then compare engagement and conversions with organic search.
Read server logs too
Logs show AI crawlers fetching your pages (for indexing or user requests) even when nobody clicks. A rise in ChatGPT-User or Perplexity-User fetches on a page means it is being used in answers.
chatgpt\.com|perplexity\.ai|gemini\.google\.com|copilot\.microsoft\.com|claude\.ai|chat\.openai\.com Match on session source in a GA4 custom channel group and place it above Referral.
Pro tips
- Report AI referral sessions and their conversion rate monthly.
- Check which landing pages AI visitors arrive on; those pages are being cited.
- Expect undercounting: many people read the answer and never click.
Common mistakes
- Assuming AI Overviews clicks can be isolated in GA4.
- Leaving AI referrals mixed into the generic Referral channel.
- Judging GEO only by clicks.
Group AI assistant referrers into their own GA4 channel, watch which pages they land on, and use server logs to see usage that never becomes a click.