Measuring visitors coming to your site from AI tools is one way to see the concrete return of your visibility work. Some AI platforms appear as a referrer and can be separated in Google Analytics 4 (GA4).
How does AI traffic appear?
When a user clicks a link in a tool like ChatGPT or Perplexity and comes to your site, this visit is often recorded as “referral” traffic. You can distinguish these visits by looking at the referring domain (for example, chatgpt.com, perplexity.ai).
How do you separate it in GA4?
- Go to the traffic acquisition report. Look at referring sources (session source / medium).
- Filter AI sources. Create a segment or filter containing the relevant domains.
- Define a channel group. If you like, create a custom channel group like “AI Referral” to collect these sources.
- Attribute conversions. Track how much this traffic converts to forms, calls or sales.
The limits of this data
GA4 traffic is not the whole picture. The reasons:
- Zero-click. Since the AI answer is seen directly, the user may not click at all; in that case no trace remains in GA4, but brand visibility still happened.
- Source obfuscation. Some platforms don’t pass the referrer or show it differently.
- Measurement gaps. A mention without a click doesn’t mean there’s no value.
The right approach
So AI traffic should be evaluated not on its own but together with visibility metrics like mention rate and share of answer. Traffic is a concrete signal; but you can’t measure the whole of visibility with traffic alone. This holistic approach is the foundation of our visibility measurement service.
Summary
You can separate AI traffic as a referrer in GA4; it’s a concrete indicator of your work. But because of the zero-click effect, traffic doesn’t show the whole of visibility. Reading traffic together with mention and share-of-answer metrics gives the most accurate picture.