How To Track Traffic From ChatGPT, Perplexity And Other AI Tools In GA4
AI tools like ChatGPT, Perplexity, and Claude aren't just answering questions anymore—they're becoming discovery engines. People ask them for product recommendations, research help, and content suggestions, and these tools are sending traffic to websites.
But here's the problem: this traffic often shows up in GA4 as "Referral" or "Direct", mixed in with everything else. If you want to understand how much traffic AI tools are actually driving to your site, you need to track it separately.
How AI And LLM Traffic Usually Shows Up In GA4
By default, GA4 doesn't have a special category for AI traffic. Instead, it gets lumped into existing channels based on how the visitor arrived:
Most Common: Referral
If someone clicks a link from ChatGPT, Perplexity, or another AI tool, it usually arrives as a referral with the source set to the AI tool's domain.
Examples: chatgpt.com, perplexity.ai, claude.ai
Sometimes: Direct
If a user copies a URL from an AI tool and pastes it directly into their browser, GA4 sees it as direct traffic with no referrer information.
This is impossible to track without UTM parameters, so some AI traffic will always be hidden in your direct channel.
Rarely: UTM-Tagged
Some savvy users or platforms add UTM parameters to links they generate. If you're lucky, you might see utm_source=chatgpt or similar.
Example: utm_source=perplexity&utm_medium=ai
The main issue: AI traffic gets mixed in with regular referral traffic from blogs, news sites, and other websites, making it hard to see how much visibility you're getting from AI tools specifically.
How To Identify Traffic From AI Tools
The best way to identify AI traffic is by looking at the referrer domains. Here are the most common sources you'll see:
| AI Tool | Referrer Domain(s) |
|---|---|
| ChatGPT | chatgpt.com, chat.openai.com |
| Perplexity | perplexity.ai |
| Claude | claude.ai, anthropic.com |
| Google Gemini | gemini.google.com, bard.google.com |
| Microsoft Copilot | copilot.microsoft.com, bing.com/chat |
| You.com | you.com |
| Meta AI | meta.ai |
| Poe (by Quora) | poe.com |
| Pi | pi.ai |
| Phind | phind.com |
| Character.AI | character.ai |
| HuggingChat | huggingface.co |
Pro tip: This list will grow as new AI tools launch. Make sure to check your referral sources regularly to catch new AI platforms sending traffic your way.
Why You Should Group AI Traffic Into Its Own Channel
You might be thinking: "It's just referral traffic, why does it need its own channel?" Here's why:
- AI tools behave differently than blogs or news sites. They're not linking to you editorially—they're recommending you based on user queries. The intent and context are completely different.
- You want to measure AI visibility separately. If Perplexity starts sending you 1,000 visits a month, that's a signal that you're being cited in AI responses. That's valuable data worth tracking on its own.
- AI traffic can have different conversion behavior. Users coming from AI assistants might be deeper in their research or ready to buy. Separating this traffic lets you measure its true value.
- It makes reporting cleaner. Instead of scrolling through hundreds of referral sources, you can see "AI Traffic" as a single line item in your reports.
Think of AI tools as a new category of discovery, similar to how you wouldn't mix organic search with display ads. They deserve their own bucket.
Step By Step: Create An "AI And LLM" Channel Group In GA4
Here's how to set up a custom channel for AI traffic in GA4. This will catch all the major AI tools and keep them separate from your regular referral traffic.
Open GA4 Admin Panel
Go to Admin (bottom left corner) → Under the Property column, click Data display → Click Channel groups.
Create A New Channel Group
Click "Create new channel group". Give it a name like "Traffic with AI Channel" and add a description if you want.
You can also copy the Default channel group and modify it—this way you keep all the existing channels and just add AI traffic as a new one.
Add A New Channel Called "AI Traffic"
Click "Add new channel" and name it "AI Traffic" or "AI & LLM".
Under Conditions, set the following:
- Dimension: Session source
- Match type: matches regex
- Value: (see the regex pattern below)
Regex pattern to match AI tools:
^(chatgpt\.com|chat\.openai\.com|perplexity\.ai|claude\.ai|anthropic\.com|gemini\.google\.com|bard\.google\.com|copilot\.microsoft\.com|bing\.com/chat|you\.com|meta\.ai|poe\.com|pi\.ai|phind\.com|huggingface\.co|character\.ai|writesonic\.com|jasper\.ai|notion\.so/ai|exa\.ai)$What this regex does: It matches any session where the source is one of the AI tool domains listed. The ^ and $ ensure exact matches, and the | separates each domain.
Reorder Your Channels (Important!)
GA4 checks channel rules from top to bottom and assigns traffic to the first match. You need to put "AI Traffic" above the "Referral" channel.
Click "Reorder" and drag "AI Traffic" so it appears before "Referral" in the list. This ensures AI tool referrals get caught by your AI channel instead of falling into the generic Referral bucket.
Save Your Channel Group
Click "Save" at the top right to apply your new channel group.
Note: GA4 custom channel groups are retroactive, so this will apply to your historical data too. You'll immediately see how much AI traffic you've been getting.
How To Measure AI Traffic: Reports And Insights
Once your AI channel is set up, here's how to analyze it:
Traffic Acquisition Report
Go to Reports → Acquisition → Traffic acquisition. Use the dropdown at the top to switch from "Default channel group" to your new custom channel group.
You'll see "AI Traffic" as its own row showing sessions, users, engagement rate, and conversions. Compare it to other channels to see how it performs.
Landing Pages
Go to Reports → Engagement → Landing page. Add a secondary dimension of "Session source" and filter for your AI tool domains.
This shows you which pages AI tools are linking to most often. If certain content gets a lot of AI traffic, that's a signal it's being cited in responses.
Conversions
Go to Reports → Engagement → Conversions and add "Session source" as a dimension. Filter for AI sources.
This tells you if AI traffic is converting at a higher or lower rate than other channels. Some sites find AI traffic converts really well because users are further along in their research.
Assisted Conversions
In Advertising → Attribution, you can see if AI traffic is assisting conversions even when it's not the final click. For example, someone might discover you through Perplexity, then come back via Google search to convert.
Tips For Getting The Most Out Of AI Traffic Tracking
Update Your Regex Regularly
New AI tools launch frequently. Every few months, check if there are new platforms you should add to your regex pattern. Keep an eye on your referral sources for domains you don't recognize.
Don't Expect 100% Coverage
Some AI traffic will always show up as "Direct" because users copy-paste URLs without clicking them. There's no way around this—just know that your AI channel is showing you the minimum amount of AI traffic, not the total.
Compare Engagement Metrics
Look at bounce rate, average session duration, and pages per session for AI traffic vs. other channels. This tells you if AI visitors are engaged or just quickly bouncing. If they're bouncing, maybe your content doesn't match what the AI tool said about you.
Share With Your Team
This is still a pretty new thing, so if you set this up, make sure your team knows about it. Show them the "AI Traffic" channel in reports and explain what it represents. It's a cool data point that not many companies are tracking yet.
Track AI Traffic Across All Your Properties
If you want to track AI and LLM traffic consistently across many properties, GA4 Channel Manager lets you create a dedicated AI traffic channel once, then apply it everywhere.
It can also review your source and medium values and suggest which ones should fall under this new AI channel, so you don't have to manually hunt through your referral data.
Try GA4 Channel Manager Free →Final Thoughts
AI tools are changing how people discover content. If you're not tracking this traffic separately, you're missing out on valuable insights about how your content is being surfaced in AI responses.
Setting up an AI traffic channel takes about 10 minutes, and once it's done, you'll have clean data showing exactly how much visibility you're getting from ChatGPT, Perplexity, and other AI platforms.
Plus, it's a pretty cool thing to show your team or clients—not many people are tracking this yet, so you'll be ahead of the curve.