AI engines like ChatGPT, Perplexity, and Claude are starting to influence how users discover brands. The challenge for marketers is that this traffic doesn’t look neat and tidy in Google Analytics 4 (GA4). Let’s walk through exactly how you can track it.
Step 1: Check Referral Traffic for AI Domains
-
In GA4, go to
Reports → Acquisition → Traffic acquisition. -
Change the Primary Dimension to Session source / medium.
-
Look for referrals from domains such as:
-
perplexity.ai -
you.com -
phind.com -
chat.openai.com
-
Step 2: Isolate Direct Traffic Spikes
A lot of AI traffic shows up as “direct” (no referrer). To find AI-like traffic inside “direct”:
-
Create a new exploration in GA4.
-
Use Landing page as the dimension.
-
Filter for first visits with Session default channel grouping = Direct.
-
Look for users landing directly on blog posts or deep pages (not just the homepage).
Step 3: Add UTM Parameters for AI Testing
If you’re feeding links into AI tools (via partnerships, experiments, or prompt injections), make sure you tag them. Example:
In GA4, these show up under Session source / medium just like other campaigns.
Step 4: Monitor Branded Search as an Indirect Signal
Sometimes AI mentions trigger curiosity, and users search your brand on Google or Bing instead of clicking.
To track this:
-
Connect GA4 with Google Search Console.
-
Check Search Console → Queries for spikes in branded keywords after your content starts appearing in AI answers.
Step 5: Set Up a Custom Segment for AI Traffic
You can group all AI sources together into one segment for easier monitoring:
-
In GA4 Explorations, click + Segment → User segment.
-
Add conditions like:
-
Session source contains perplexity -
Session source contains you.com -
OR
Session medium = ai_recommendation.
-
-
Save as AI Traffic Segment.
What This Gives You
Once you’ve set this up, you’ll have:
-
Referral attribution for known AI domains.
-
A filterable segment for analyzing AI traffic behavior.
-
A way to spot indirect impact (branded search spikes).
-
A UTM framework to track AI experiments cleanly.
Over time, this will evolve into a dedicated channel in your reporting — just like “Organic Search” or “Paid Social.”
👉 At Optizent, we’re helping clients adapt to these changes and build frameworks that won’t break when AI becomes a primary traffic driver. Contact us if you’d like a walkthrough tailored to your setup.