Visits from AI assistants behave differently from ordinary search traffic. People often arrive after reading a summary, already knowing what they want. If you track AI traffic in GA4 properly, you can see how much of it you get, which pages it lands on and whether it converts.

Where AI traffic shows up by default

By default, GA4 classifies most AI assistant visits as Referral traffic, mixed with every other website that links to you. Some visits arrive with no referrer at all and appear as Direct, for example when users copy a link or when an app does not pass referrer information. To report AI as its own channel, you need a custom channel group.

Common AI referral sources

Typical AI assistant referrers
AssistantCommon referrer domains
ChatGPTchatgpt.com (older visits: chat.openai.com)
Perplexityperplexity.ai
Google Geminigemini.google.com
Microsoft Copilotcopilot.microsoft.com
Claudeclaude.ai
Othersyou.com, phind.com, poe.com and similar

Referrer domains change over time, so review your referral report every few months and add new sources. Note that Google AI Overviews and AI Mode clicks generally appear as normal Google organic traffic rather than as a separate referrer.

Step 1: Check your current LLM referrals

  1. In GA4, open Reports, then Acquisition, then Traffic acquisition.
  2. Change the primary dimension to Session source.
  3. Search for "chatgpt", "perplexity", "gemini", "copilot" and "claude".
  4. Note the sessions, engagement rate and key events for each.

This gives a quick baseline of ChatGPT referral traffic and other LLM referrals.

Step 2: Create a GA4 custom channel group

  1. Go to Admin, then Data display, then Channel groups.
  2. Create a new channel group (copy the default group so other channels keep working).
  3. Add a new channel called "AI assistants".
  4. Set the condition: Session source matches regex, using a pattern such as:

^(chatgpt\.com|chat\.openai\.com|perplexity\.ai|.*\.perplexity\.ai|gemini\.google\.com|copilot\.microsoft\.com|claude\.ai|you\.com|phind\.com|poe\.com)$

  1. Move the AI assistants channel above Referral so it is evaluated first.
  2. Save. The channel group applies to historical data in reports and explorations.

Google's help centre explains custom channel groups and key events in more detail.

Step 3: Track AI traffic in GA4 reports and conversions

  • In Traffic acquisition, switch the channel dimension to your custom group to compare AI assistants with organic search, paid and referral.
  • Add key events such as form submissions or demo requests to see conversion rates by channel. Our GA4 lead generation tracking guide explains how to set those up.
  • Build an exploration showing landing pages for the AI assistants channel. These are the pages AI tools find most useful.

Step 4: Capture what software misses

Many AI-influenced journeys are invisible to analytics: a buyer asks ChatGPT for recommendations, then later searches your brand name or types your URL. To measure these:

  • Add an open "How did you hear about us?" field to forms and code answers monthly.
  • Ask sales to record mentions of AI tools during discovery calls.
  • Watch branded search trends alongside AI referral growth.

Our marketing attribution models guide explains how to combine software and self-reported data.

Measuring AI visibility beyond traffic

Traffic is only one indicator. A complete view includes:

Ways to measure AI visibility
MethodWhat it shows
AI assistants channel in GA4Visits and conversions from AI tools
Prompt testingWhether your brand appears for priority questions
Self-reported attributionAI influence without clicks
Branded search trendsRising awareness, partly from AI mentions
Search ConsoleOverall Google performance, including AI features

Interpreting AI traffic

AI referral volumes are usually small compared with search, but engagement and conversion rates can be higher because visitors arrive with context. Look at:

  • Which pages AI tools send people to; strengthen and update those pages.
  • Engagement rate and conversions compared with organic search.
  • Growth trend month over month.

If AI tools mostly send visitors to an outdated article, refresh it first; that page is effectively your front door for AI users. Our guide to updating old blog posts can help.

Common tracking issues

  • Regex too narrow: subdomains or new referrers are missed. Review sources quarterly.
  • Channel order: if Referral is above AI assistants, AI visits are captured as referrals.
  • Consent settings: visitors who decline analytics cookies may be modelled or missing, depending on your consent mode setup; see our Google Consent Mode v2 guide.
  • UTM overrides: some AI tools append utm_source parameters, which set the source directly. Include those values in your regex.

Turning data into action

Use AI traffic data to guide content: pages that attract AI visitors deserve regular updates, stronger calls to action and clear next steps. Questions visitors ask after arriving, in chat or forms, show which topics to cover next. Combine this with our generative engine optimization approach to improve visibility over time.

If your GA4 setup is not producing trustworthy channel and conversion data, our conversion and analytics team can audit and fix it.

Frequently asked questions

Does GA4 have a built-in AI traffic channel?
Not by default. Most AI assistant visits appear as Referral or Direct, so you need a custom channel group to report them separately.
Why does ChatGPT traffic sometimes show as direct?
Some apps and copied links do not pass referrer information, so GA4 cannot see where the visit came from and records it as direct.
Are Google AI Overviews clicks tracked as AI traffic?
Clicks from Google's AI features generally appear as Google organic search traffic, not as a separate AI referrer.
How much traffic do AI assistants send?
It varies widely by industry and audience. For most sites it is still small compared with search, but often growing and frequently well engaged.