By the AEO GEO Marketing team · Updated
- GA4 lumps AI assistant visits into “Referral” by default.
- A custom channel group with a regex of AI domains separates them cleanly.
- Some AI traffic arrives without a referrer and appears as direct.
- Compare AI referrals with other channels on engagement and conversions.
- Combine analytics with prompt tracking for a full picture.
Why you need to track AI traffic separately
Visits from AI assistants are growing for many sites, and they often behave differently from other traffic: visitors arrive with context from the AI answer, view specific pages and convert at different rates. If these visits are hidden among generic referrals, you cannot judge whether your AI search work is paying off.
Google Analytics 4 does not have a built-in AI channel. With a few minutes of configuration, you can create one and report on it alongside organic search, paid ads and social.
Which AI sources to include
Common referrer domains for AI assistants include:
- chatgpt.com and chat.openai.com (ChatGPT)
- perplexity.ai (Perplexity)
- gemini.google.com and bard.google.com (Gemini)
- copilot.microsoft.com and bing.com/chat (Microsoft Copilot)
- claude.ai (Claude)
- you.com, phind.com and other AI search tools
- meta.ai (Meta AI)
New assistants appear regularly, so review your referral reports every few months and add new domains.
Step 1: Find existing AI referrals
Before configuring anything, see what is already arriving. In GA4, open Reports, then Acquisition, then Traffic acquisition. Change the primary dimension to “Session source / medium” and search for the domains above. This shows how much AI traffic you already receive and which assistants send it. Note the exact source values, because some appear with or without “www” or with different subdomains.
Step 2: Create a custom channel group
- Go to Admin, then Data display, then Channel groups.
- Copy the default channel group so you keep all existing rules.
- Add a new channel called “AI Assistants”.
- Set its condition to: Session source matches regex, with a pattern listing the AI domains, for example chatgpt|openai|perplexity|gemini|bard|copilot|claude|you\.com|phind|meta\.ai.
- Move the new channel above “Referral” in the order, because GA4 assigns sessions to the first matching channel.
- Save.
Custom channel groups apply to historical data in reports, so you can see past AI traffic immediately.
Step 3: Build an AI traffic report
Create an Exploration (Explore, then Free form) with:
- Rows: Session source, Landing page.
- Columns: your new channel group.
- Values: Sessions, Engaged sessions, Average engagement time, Key events (conversions), Revenue if relevant.
Filter to the AI Assistants channel. This shows which assistants send traffic, which pages they land on and how well those visits convert. Landing pages reveal which content AI engines are citing.
Step 4: Compare with other channels
Add the AI Assistants channel to your standard traffic acquisition report and compare it with Organic Search, Paid Search and Direct. Look at engagement rate, conversion rate and revenue per session. Many businesses find that AI referrals, while smaller in volume, show strong engagement because visitors arrive pre-informed. These comparisons help you decide how much to invest in AI search relative to other channels.
Dealing with missing referrers
Not all AI traffic carries a referrer. Some apps and browsers strip referrer information, and users sometimes copy a URL from an answer and paste it into a new tab. These visits appear as direct traffic.
Ways to estimate the gap:
- Watch for unexplained growth in direct traffic to deep content pages, which people rarely type in manually.
- Add “AI assistant (ChatGPT, Perplexity, etc.)” as an option in your “How did you hear about us?” field.
- Track branded search growth in Search Console.
None of these is perfect, but together they give a more realistic picture.
Tracking conversions from AI visitors
Make sure your key events (formerly conversions) are configured: form submissions, bookings, calls, WhatsApp clicks, purchases and sign-ups. Then check conversion rates for the AI Assistants channel. If AI visitors engage but rarely convert, the landing pages they arrive on may need stronger calls to action or a clearer path to contact.
For businesses with long sales cycles, connect GA4 to your CRM so leads from AI referrals can be followed through to revenue.
Using UTM parameters where you can
You cannot control the links AI engines use to cite you, but you can tag links you place in content that AI tools may surface, such as links in your llms.txt file or in documentation. Use utm_source and utm_medium values that make the origin clear. Do not add UTM parameters to internal links on your own site, which would break session attribution.
Server logs: the other half of the picture
GA4 measures human visits. It does not show crawler activity. To see whether AI crawlers such as OAI-SearchBot, PerplexityBot or ClaudeBot are visiting your site, check server logs or your CDN’s bot analytics. Crawler visits to a page often precede citations of that page, so log data is a useful early indicator and helps diagnose access problems.
A monthly reporting routine
- Review AI Assistants sessions, engagement and conversions month on month.
- Note the top landing pages from AI referrals.
- Compare with your prompt tracking results.
- Check for new AI referrer domains and update the regex.
- Report findings alongside organic and paid performance.
Fifteen minutes a month is enough to keep AI traffic visible to decision-makers.
Common mistakes
- Leaving the AI channel below “Referral” in channel order, so it never matches.
- Using a regex that is too broad and catches unrelated domains.
- Forgetting new assistants and subdomains.
- Judging AI value on sessions alone, without conversions.
- Assuming all AI influence is captured by referrals.
Example regex patterns
A starting regex for the AI Assistants channel condition might look like this: chatgpt\.com|chat\.openai\.com|perplexity\.ai|gemini\.google\.com|bard\.google\.com|copilot\.microsoft\.com|claude\.ai|you\.com|phind\.com|meta\.ai. Escape dots with a backslash so they match literally, and separate domains with the pipe character. Test the pattern against your source list in an Exploration before saving the channel group, and keep a copy of the pattern in your documentation so it can be updated consistently.
Looker Studio dashboards
If your team uses Looker Studio, connect GA4 and add the custom channel group as a dimension. A simple dashboard page can show AI Assistants sessions over time, top AI sources, top landing pages and conversions, next to organic search for comparison. Add Search Console data for branded queries and your prompt-tracking results from a spreadsheet. Sharing one dashboard keeps marketing, sales and leadership looking at the same numbers.
What AI landing pages tell you
The landing pages that receive AI referrals reveal which content engines cite. Often they are not the pages you expect: detailed guides, pricing pages, comparison pages or documentation may outperform the home page. Use this insight to strengthen those pages with clearer calls to action, update them regularly and create more content of the same type for related questions.
Segments and audiences for AI visitors
Once AI referrals are identified, create a GA4 segment for them and an audience of users who arrived from AI assistants. The segment lets you analyse their paths through the site, which pages they view after landing and where they drop off. The audience can be used for remarketing in Google Ads, showing tailored messages to people who discovered you through AI answers. Because these visitors often arrive already informed, remarketing with a direct offer, such as a free audit or consultation, can perform well.
Attribution across the journey
AI assistants often appear early in the buying journey: a prospect asks ChatGPT for options, visits two or three sites, then returns days later through a branded Google search or directly to convert. Last-click attribution gives all the credit to the final visit. Use GA4’s path exploration and data-driven attribution, plus CRM data where you have it, to see how often AI referrals start journeys that convert later. This usually shows AI search is worth more than last-click reports suggest.
Consent and privacy considerations
Analytics data depends on consent in many regions. If visitors decline analytics cookies, their sessions may not be recorded fully, which affects AI referral counts as much as any other channel. Use Google Consent Mode where appropriate, make sure your consent banner is compliant and clear, and remember that analytics figures represent consenting visitors. Trends remain useful even when absolute numbers are incomplete.
Key terms in this guide
- AI Overviews: AI-generated summaries that appear at the top of some Google search results, with links to the web pages used as sources.
- LLM (Large Language Model): An AI model trained on large amounts of text to understand and generate language. LLMs power ChatGPT, Gemini, Claude, Copilot and many AI search features.
- RAG (Retrieval-Augmented Generation): A technique where an AI system retrieves relevant documents, usually from a search index, and uses them to write a grounded answer. Most AI search engines use a form of RAG, which is why being retrievable matters.
- Citation: A link or reference to a source shown inside an AI-generated answer. Citations are the main unit of success in GEO.
- Prompt tracking: Regularly running a fixed set of questions through AI engines and recording which sources and brands they cite, to measure GEO progress over time.
- llms.txt: A proposed plain-text file at a website’s root that gives AI systems a concise, structured overview of the site and links to its most important content.
- ClaudeBot: Anthropic’s web crawler, used to gather publicly available content for its Claude models and related search features.
- PerplexityBot: Perplexity’s crawler, which indexes pages so they can appear as sources in Perplexity answers.
How AEO GEO Marketing can help
If you would rather have specialists handle this, these services cover the work described in this guide:
Paid Ads & Analytics
Google & Meta Ads managed for maximum ROI with full attribution reporting. It includes google ads & pmax, meta & linkedin ads, tracking & attribution and more, from $1,000/mo.
AI Citation Monitoring
Monthly tracking of where AI engines cite you and your competitors — from $249/mo. It includes prompt tracking, competitor comparison, brand description check and more, from $249/mo.
Not sure where to start? The free 48-hour AI snapshot shows where AI engines cite you today.
Frequently asked questions
Does GA4 track Google AI Overviews separately?
No. Clicks from AI Overviews appear as organic Google search traffic. Use Search Console and dedicated tracking to understand AI Overview performance.
Will the custom channel affect historical data?
Custom channel groups are applied retroactively in standard reports and explorations, so you can analyse past traffic.
Can I track AI traffic in other analytics tools?
Yes. The same approach works in Matomo, Plausible, Adobe Analytics and others: group sessions by referrer domain.
How accurate is AI referral tracking?
It accurately captures visits that carry a referrer. Visits without a referrer are hidden in direct traffic, so treat the numbers as a minimum.
Related guides and services
Continue with these guides:
- How to measure AI search visibility: A practical framework for tracking citations, mentions and revenue from AI search, with the metrics that actually matter.
- What is GEO? The 2026 guide to Generative Engine Optimization: GEO is how brands get cited inside AI answers. Here is how it works, why it matters and how to start.
- What is AEO? Answer Engine Optimization explained: Answer engines extract a single best answer. AEO makes sure that answer is yours.
- GEO vs SEO: what changes, what stays the same: Rankings still matter — but being cited now matters more. A practical comparison.
Or see how we put this into practice:
- Paid Ads & Analytics: Google & Meta Ads managed for maximum ROI with full attribution reporting.
- AI Citation Monitoring: Monthly tracking of where AI engines cite you and your competitors — from $249/mo.
Get a free 48-hour GEO audit that shows exactly where AI engines cite you — and where they cite your competitors.
