How AI Referral Traffic Works
AI referral traffic is the set of visits that begin when a person clicks a link inside an AI assistant's answer. The assistant cites or recommends a page, the person clicks, and the visit arrives on your site. It is a distinct source from paid ads, organic search and social.
To see how to measure it step by step, read how to track AI assistant traffic to your website.
Why It Is Often Missed
Not every assistant passes a clean referrer. Some links may carry an identifying referrer or a tag in the URL, while others may arrive with no referrer at all and land in the "direct" bucket of a standard analytics report. A site that only looks at default channel groups can therefore undercount visits from AI assistants.
For example, a visitor asks an assistant for the best tools in a category, clicks your listing, and signs up. Without a rule that recognizes that referrer, the sign-up may be credited to direct traffic.
How to Identify It
- Match known assistant referrer domains and URL tags in a custom traffic rule
- Group them into one "AI assistants" class next to paid, search, social, referral and direct
- Compare visitors, sign-ups and revenue from that class against other sources
This is a core part of web attribution, and the same tag-reading logic applies as for UTM parameters. Because credit usually goes to the final visit, the gap is most visible under last-touch attribution.
Linkrunner's Web SDK (beta) classifies visits from ChatGPT, Perplexity, Claude, Gemini, Copilot and Grok as their own traffic class.