How to Track ChatGPT, Perplexity and AI Traffic in GA4
To track AI traffic in GA4, look at your referral sources: ChatGPT, Perplexity and the other AI tools send visitors that show up under their own domains, like chatgpt.com and perplexity.ai. GA4 does not give you an “AI” channel out of the box, so the traffic is there but scattered, and easy to miss unless you go looking for it.
The good news is that finding it takes a few minutes, and once you build a simple segment or channel group, you can watch AI referral traffic grow over time and tie it to conversions. This guide shows you exactly how to track AI traffic in GA4, what source each tool shows up as, how to segment it, and why some of it hides inside “Direct”.
What This Covers
- The step-by-step way to track AI traffic in GA4
- What referral source each AI tool shows up as
- How to see ChatGPT traffic specifically
- How to segment AI traffic into its own group
- Why some AI traffic hides in “Direct”, and what to do
The Short Answer
AI tools appear in GA4 as ordinary referral traffic, each under its own domain. There is no built-in “AI” bucket, so your job is to identify those domains, group them, and track them together. It is manual the first time and automatic after that.
Two things make it trickier than normal referral tracking. Some AI tools do not always pass a referrer, so a slice of their traffic lands in “Direct”. And the source names change as tools rename their domains. So treat your AI segment as something you review and update every few months, not a set-and-forget report.
It is worth doing even while the numbers are small. AI referral traffic is early for most businesses, but the visitors tend to arrive warm, because an AI just recommended you by name. Setting up the tracking now means you have a clean baseline to compare against later, so when the volume does climb you can prove it grew rather than guessing. A blind spot you fixed early is worth far more than a dashboard you scramble to build once the traffic already matters.
How to Track AI Traffic in GA4
Here is the process end to end. Five steps get you a clean, repeatable view.
- Open the Traffic acquisition report. In GA4, go to Reports, then Acquisition, then Traffic acquisition. This is where session-level source data lives (Google Analytics Help).
- Switch the dimension to Session source / medium. Use the dropdown at the top of the table to change the primary dimension, so you can see exact referring domains rather than grouped channels.
- Search for the AI sources. In the table’s search box, look for the AI domains one at a time: chatgpt.com, perplexity.ai, gemini.google.com, and copilot or bing for Microsoft. Note the sessions, engagement and conversions for each.
- Build a segment or custom channel group for AI. In an Exploration, create a segment where Session source matches any of those AI domains, or set up a custom channel group named “AI / LLM” so the traffic reports together automatically going forward.
- Track it over time and against conversions. Add your AI segment to a saved Exploration or a Looker Studio report, and watch whether AI referral sessions and their conversions climb month over month as your AI visibility grows.
Once that segment exists, tracking AI traffic in GA4 becomes a glance rather than a hunt. For the bigger question of which AI engine actually sends the most, see ChatGPT vs Perplexity vs Google AI Overviews.
What Source Does Each AI Tool Show As?
Each AI tool refers traffic under its own domain, so recognising the names is half the job. The exact strings can change as tools update, so confirm them against your own data, but these are the common ones.
| AI tool | Typical GA4 source |
|---|---|
| ChatGPT | chatgpt.com (older: chat.openai.com) |
| Perplexity | perplexity.ai |
| Google Gemini | gemini.google.com |
| Microsoft Copilot | copilot.microsoft.com or via bing.com |
| Google AI Overviews | Usually blended into google / organic, hard to isolate |
Note the last row. AI Overviews traffic appears inside normal Google Search rather than as a separate referrer, so you cannot cleanly split it out in GA4. That is a real limitation, and it is why a manual citation check matters alongside analytics, as we cover in AEO vs SEO vs GEO.
How Do I See ChatGPT Traffic in GA4?
Filter for its domain. In the Traffic acquisition report with Session source / medium as the dimension, type “chatgpt” into the table search, and you will see sessions referred from chatgpt.com (and chat.openai.com from older links). That row is your ChatGPT referral traffic.
To go deeper, open an Exploration, add Session source / medium as a row, and pull in metrics like sessions, engaged sessions, conversions and revenue. Now you can see not just how many people arrive from ChatGPT, but whether they convert. If you are earning citations there, this is where the payoff shows up, and how to get cited by ChatGPT covers the earning side.
How to Segment AI Traffic in GA4
Group all the AI domains into one segment so they report together. In an Exploration, create a new segment with a condition like “Session source matches regex” and list the AI domains, for example chatgpt.com|perplexity.ai|gemini.google.com|copilot. Apply it, and every AI referral rolls up into one clean number you can trend over time.
For a more permanent solution, build a custom channel group in your GA4 admin that routes those same sources into an “AI / LLM referral” channel. Then AI traffic appears as its own channel in standard reports, no manual filtering required. The same segment logic works for Perplexity specifically, which pairs with how to appear in Perplexity AI answers.
What Should You Watch Once It’s Set Up?
Once your AI segment exists, three numbers are worth watching each month. First, AI referral sessions over time. You are looking for a trend, not a single figure, so compare month to month rather than obsessing over any one week. A steady climb means your AI-search work is landing.
Second, engagement and conversions from AI traffic. AI visitors often arrive with high intent because a tool just recommended you, so check whether they engage and convert better than your site average. If they do, that is a strong argument for investing more in AI visibility. Third, the split across tools. Watch which AI source sends the most useful traffic for you, because it tells you where to focus, and for most sites the mix shifts over the year as the tools grow at different rates.
Why Some AI Traffic Hides in “Direct”
Some AI tools do not reliably pass a referrer, especially inside apps or when links open in certain ways. When the referrer is missing, GA4 has nothing to attribute the visit to, so it files the session under “Direct / none”. That means a portion of your real AI traffic is invisible as AI traffic, sitting in your Direct bucket instead.
You cannot fully fix this, but you can allow for it. Watch whether Direct traffic rises in step with your AI visibility, use UTM tags on any links you control, and treat your measured AI numbers as a floor rather than the full picture. This under-counting is one more reason to measure citations and brand mentions, not just sessions, a theme we keep coming back to across the AEO and GEO work.
Common Mistakes When Tracking AI Traffic
A few mistakes make AI tracking less useful than it should be. The first is treating the measured number as complete. It never is, because of the Direct and AI Overviews gaps above, so read it as a floor and a trend, not a total.
The second is forgetting to update the source list. Tools rename domains and new ones launch, so a segment built once and never touched slowly drifts out of date; review it every few months. The third is measuring sessions but ignoring conversions. Raw AI sessions are interesting, but AI-driven leads and sales are what actually justify the work, so always connect your AI segment to a conversion, even a soft one like a form view, so you can see value, not just volume.
The Nobody Cares Take
Most businesses have AI traffic in GA4 right now and have never looked for it, because it is scattered across a handful of domains with no label on it. Ten minutes and one segment turns that blind spot into a trend line, which is a genuinely useful thing to have as AI search grows.
But keep the limitation in mind: GA4 will always under-count AI, because AI Overviews blend into Google and app referrers go missing into Direct. So the number you build is directional, not exact. It tells you the shape and the trend, not the whole truth.
That is fine, as long as you use it that way. Track the trend, tie it to conversions where you can, and pair it with a monthly manual check of whether the AI tools actually name you. Analytics tells you who clicked; the manual check tells you who got cited. You want both.
Frequently Asked Questions
How do I see ChatGPT traffic in GA4?
Open Reports, then Acquisition, then Traffic acquisition, and switch the dimension to Session source / medium. Search the table for “chatgpt” and you will see sessions from chatgpt.com (and the older chat.openai.com). For depth, build an Exploration with Session source and metrics like conversions to see whether ChatGPT visitors actually convert.
What referral source does Perplexity show as in GA4?
Perplexity typically appears as perplexity.ai in your Session source / medium data. Search for “perplexity” in the Traffic acquisition table to find it. As with all AI tools, confirm the exact string against your own data, since domains can change, and remember some Perplexity visits may land in Direct if no referrer is passed.
How do I segment AI traffic in GA4?
Create a segment in an Exploration where Session source matches a regex listing the AI domains, such as chatgpt.com|perplexity.ai|gemini.google.com. Apply it to roll all AI referrals into one number. For a permanent version, set up a custom channel group in GA4 admin that routes those sources into an “AI / LLM referral” channel.
Why does GA4 show AI traffic as Direct?
Because some AI tools, especially inside apps, do not pass a referrer. When GA4 receives no referrer, it attributes the visit to “Direct / none”. So a share of your real AI traffic hides in Direct and cannot be cleanly identified. Watch whether Direct rises alongside your AI visibility, and treat measured AI numbers as a floor.
Can I track Google AI Overviews traffic in GA4?
Not cleanly. AI Overviews appear within normal Google Search, so their traffic blends into google / organic rather than showing as a separate referrer. You cannot isolate it reliably in GA4. This is why a monthly manual check of whether you appear in AI Overviews matters alongside your analytics data.
Is AI referral traffic worth tracking if it is small?
Yes. AI referral volume is still modest for most sites, but it is growing, and the visitors often arrive with high intent after an AI recommended you. Tracking it now sets a baseline so you can see the trend, prove the value of your AI-search work, and catch the moment it starts to matter.
Do I need to set up anything special to track AI traffic?
No special tracking code is required; AI tools already show up as referral traffic in standard GA4. What you set up is the organisation: a segment or custom channel group that groups the AI domains so they report together. That takes a few minutes once and then works automatically for every future visit.
How often should I check my AI traffic in GA4?
Monthly is enough for most sites. AI referral volume moves slowly, so a monthly review shows the trend without noise. Pair it with a monthly manual check of whether the AI tools name you, and update your source list every few months as tools rename domains or new ones appear.
