top of page

How to Track AI Traffic in Google Analytics 4, and What GA4 Can't See

Writer: Harold Bell
Harold Bell
22 hours ago
5 min read
Smartphone on a wooden desk showing the Google Analytics app's loading screen with its orange bar-chart logo.

Key takeaways

  • Google Analytics 4 doesn't include an AI channel by default, so visits from ChatGPT, Perplexity, Claude, Gemini and Copilot usually land in Referral or Direct.

  • Fix it with a custom channel group: add an AI Assistants channel that matches AI referrer domains, and place it above Referral.

  • ChatGPT often adds utm_source=chatgpt.com to the links it shows, which makes attribution easier. Some AI traffic still arrives with no referrer at all.

  • GA4 only sees the people who click. Pair it with self-reported attribution and share of model in what I call the AI Traffic Triangle.


Ask most B2B teams how much traffic they get from AI assistants and you'll get a shrug. It's in there somewhere, mixed into Referral, or hiding in Direct.


That's a missed signal. AI-referred visitors tend to behave differently. They've usually read a summary already. They arrive with a specific question, often closer to a decision. And their numbers are a leading indicator of whether your AI visibility work is paying off.


Setting up GA4 to see them properly takes about ten minutes. Understanding what GA4 can't see takes a little longer, and matters more.



How do you track AI traffic in Google Analytics 4?

To track AI traffic in Google Analytics 4, create a custom channel group with an AI Assistants channel that matches session sources from AI domains such as chatgpt.com, perplexity.ai, claude.ai, gemini.google.com and copilot.microsoft.com, then place it above the Referral channel. Use that channel group in reports to see AI sessions, engagement and key events separately.



Why GA4 doesn't show AI traffic by default


GA4 replaced Universal Analytics in July 2023, and its default channel groups were designed before AI assistants sent meaningful traffic. A click from ChatGPT usually shows up as a referral from chatgpt.com, sitting next to every other referring website. Some AI traffic, especially from desktop and mobile apps, arrives with no referrer and lands in Direct.


So AI traffic is there, but scattered. You have to group it yourself.



How to set up an AI Assistants channel in GA4


  1. Open channel groups. In Admin, go to Data display, then Channel groups.

  2. Create a new group. Start from a copy of the default channel group so your other channels stay intact.

  3. Add a new channel. Name it "AI Assistants."

  4. Set the condition. Session source matches regex, using a pattern like the one below.

  5. Reorder. Drag AI Assistants above Referral. GA4 assigns each session to the first channel that matches, so order matters.

  6. Save and use it. In reports, switch the primary dimension to your new channel group.


A starting regex for the condition:


chatgpt\.com|chat\.openai\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com


Review the list every quarter and add new assistants as they appear. Custom channel groups also apply to historical data, so once it's set up you can look back at past AI traffic.



Partner Over Product: Winning deals without losing the plot, by Harold Bell. Available on Amazon.


UTM parameters and AI traffic


ChatGPT often appends utm_source=chatgpt.com to links in its answers, which makes those visits easy to attribute even when the referrer is lost. Other assistants vary, and the behavior can change.


Where you control the link, such as content you distribute or AI share buttons on your own pages, add your own UTM parameters. The trade-offs of share buttons are covered in what is CiteMET.



What to measure for AI traffic


Metric

What it tells you

Sessions by assistant

Which platforms are sending visitors

Landing pages

Which content AI is citing and sending people to

Engagement rate

Whether visitors find what the AI promised

Key events and conversions

Whether AI visitors take action: demo requests, signups, downloads

Assisted conversions

AI's role earlier in longer B2B journeys


The landing page report is the hidden gem. It shows you, page by page, which content AI assistants are actually citing and clicking through to. That's your citation list from the click side. Compare it with what you see in AI search visibility tracking.


Then improve those pages for the visitor AI sends. They've already read a summary, so they want proof, specifics and a clear next step, not another introduction. What is conversion rate optimization and landing page best practices cover how.



What GA4 will never show you


Here's the uncomfortable part. Most people who read about you in an AI answer never click. That's the zero-click reality, and it applies even more strongly inside assistants than on Google. GA4 only sees the minority who visit.


So the AI channel in GA4 is a floor, not a ceiling. A buyer who reads your framework in ChatGPT, then searches your brand a week later, shows up as branded organic or direct traffic. The AI got no credit. The influence was real.



The AI Traffic Triangle


To see AI's real impact, I combine three sources. Each covers a blind spot in the others.


Corner

What it measures

Blind spot

GA4 AI channel

Visits and conversions from AI clicks

Everyone who didn't click

Self-reported attribution

What buyers say influenced them, via a "how did you hear about us?" field

Depends on memory and honesty

Share of model

How often AI names you across a fixed prompt set

Doesn't show traffic or revenue directly


Together, they tell a credible story: AI is naming us more often (share of model), more buyers mention AI when they reach out (self-reported), and the visitors AI sends convert well (GA4). See share of model and content marketing ROI for how to present it.


Add a free-text "how did you hear about us?" field to your demo form if you don't have one. It's the cheapest, most honest attribution tool in B2B, and it's often the first place AI influence shows up.



Google Analytics in the Engine Optimization Matrix™


GA4 measures the downstream result of the Citation column in the Engine Optimization Matrix™: the visits AI citations actually send. On its own, it undercounts. Paired with citation and mention tracking, it completes the picture.




Stop guessing which engine to fix first. Mastering AI Visibility puts AI SEO, AEO, GEO and LLMO in one Demand Doctrine™ edition, built on the Engine Optimization Matrix™. Out November 18.



Frequently asked questions


Does GA4 track ChatGPT traffic?

Yes, but usually as a referral from chatgpt.com or as direct traffic. A custom channel group makes it visible as its own channel.


How do I create an AI channel in GA4?

In Admin, go to Data display and Channel groups, copy the default group, add an AI Assistants channel matching AI domains in session source and place it above Referral.


What regex should I use for AI traffic in GA4?

A starting pattern is chatgpt\.com|chat\.openai\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com. Update it as new assistants appear.


Does ChatGPT add UTM parameters?

ChatGPT often adds utm_source=chatgpt.com to links in its answers, which helps attribution.


Why does some AI traffic show as direct?

Some visits from AI apps arrive without a referrer, so GA4 classifies them as direct.


Do custom channel groups apply to past data?

Yes. Custom channel groups in GA4 apply to historical data, so you can see past AI traffic once one is set up.


Why doesn't GA4 show the full impact of AI?

Most people who read about you in an AI answer never click, so GA4 only captures a fraction of AI's influence.


What is the AI Traffic Triangle?

A measurement approach combining the GA4 AI channel, self-reported attribution and share of model to estimate AI's full impact.



About the author. Harold Bell is the founder of MQL Magnet and has spent about 16 years in enterprise technology marketing, working with brands like Cisco, Rubrik and Nutanix in different capacities. He's a member of the Forbes Communications Council, the creator of the Engine Optimization Matrix™ and the author of Partner Over Product.

bottom of page