It wasn’t long ago that “Search” meant Google. But in 2026, the landscape has shifted. Users aren’t just typing keywords into a search bar; they are having full-blown conversations with ChatGPT, asking Gemini for product comparisons, and getting travel advice from Perplexity.
But if you look at your standard Google Analytics 4 (GA4) reports, this traffic is nowhere to be found.
Or rather, it is there, but it’s hiding. It’s likely buried in your “Referral” traffic or masked as “Direct,” making it impossible to see if your brand is actually being recommended by these AI models.
At Loud Digital, we believe you can’t improve what you don’t measure. So, we’re going to show you exactly how to build a custom GA4 report that isolates traffic from AI sites like ChatGPT, Gemini, Claude, and Copilot.
The Problem: GA4 Doesn’t Have an “AI” Category
By default, GA4 lumps traffic into broad buckets like “Organic Search” or “Referral.” It doesn’t natively know that chatgpt.com is an AI tool. To fix this, we need to manually tell GA4 what to look for using a “Regular Expression” (Regex).
Step 1: Define Your “AI Chatbots” Channel
The best way to do this is to create a permanent Channel Group. This ensures that “AI Chatbots” appears alongside “Organic Social” and “Email” in all your future data.
-
Go to Admin (the gear icon).
-
Under Data display, select Channel groups.
-
Click Create new channel group and name it something like “AI Custom Group”.
-
Click Add new channel and name it AI Chatbots.
Now, we need to set the rule. Set the condition to Source > matches regex.
Use this specific code:
Code snippet
.*(chatgpt|openai|gemini|bard|perplexity|claude|anthropic|copilot|edgeservices).*
⚠️ A Warning About Bing & Copilot
You might be tempted to just add bing to that list. Don’t do it.
If you include generic “bing” in your AI filter, you will accidentally steal all your SEO traffic from Bing Search and misclassify it as AI.
In the code above, we have used copilot and edgeservices. The latter (edgeservices) is the specific footprint often left by the sidebar chat in the Microsoft Edge browser. This ensures we catch the AI users without messing up your organic search data.
Step 2: Create the Dedicated Report
Now that the data is categorised, let’s build a view where you can actually analyse it.
-
Go to Reports > Acquisition > Traffic acquisition.
-
Click the Pencil Icon (Customize report) in the top right.
-
Set the Default View: We want to see exactly which bot is sending traffic. Go to Dimensions, click the three dots next to Session source / medium, and select Set as default.
Step 3: The “Filter” Fix (Crucial!)
If you just save the report now, you’ll see all your traffic. We want this report to show only the AI bots.
-
In the customised report menu, click Report Filter.
-
Select Add filter.
-
Set the Dimension to Session source.
-
Set the Match Type to matches regex.
-
Paste the same code we used earlier:
Code snippet
.*(chatgpt|openai|gemini|bard|perplexity|claude|anthropic|copilot|edgeservices).* -
Click Apply.
Why the .* matters: You’ll notice we added .* at the start and end of the code. This is a “wildcard.” It tells GA4 to match the text even if there is extra junk around it (e.g., l.openai.com or m.chatgpt.com). Without these wildcards, your report might show up empty!
Step 4: Save and Publish
-
Click Save > Save as a new report.
-
Name it “Traffic Acquisition (AI View)”.
-
To make it easily accessible, go to your Library (folder icon at the bottom of the reports menu), find your “Life Cycle” collection, and drag your new report into the menu list.
FAQ: Why does it say “(not set)”?
Once your report is running, you might see rows that look like this:
chatgpt.com / (not set)
Don’t panic. This is normal.
This happens because of privacy settings (Referrer Policy) on sites like OpenAI. They tell GA4 “This user came from ChatGPT,” but they don’t explicitly send the “Referral” tag. Because your filter is looking for the Source (chatgpt), the data is still accurate, even if the Medium is a bit messy.
Summary
You now have a dedicated view of how the new wave of AI search is impacting your site. You might find that while Google brings volume, users coming from AI are asking highly specific questions and may have higher engagement rates.
