Traffic from links in AI assistants usually reaches your site as referral traffic from domains like chatgpt.com, perplexity.ai, gemini.google.com, claude.ai and copilot.microsoft.com. By default, GA4 lumps it into "Referral" with every other site. Creating a custom channel group with an "AI assistants" channel makes it visible in every report. This guide walks through the setup and explains what GA4 can and can't tell you.
Step 1: Check what you already get
Before configuring anything, look at the raw data:
- Go to Reports → Acquisition → Traffic acquisition.
- Change the primary dimension to Session source.
- Search for
chatgpt,perplexity,gemini,claude,copilot.
Note which domains appear and how much traffic each sends. Also check Session source / medium for chatgpt.com / (not set) or utm_source=chatgpt.com values. ChatGPT often appends that UTM parameter to links it shows.
Step 2: Create a custom channel group
- Go to Admin → Data display → Channel groups.
- Click Create new channel group (or copy the default group so you keep the standard channels).
- Name it something like "Channels with AI assistants."
- Add a new channel named "AI assistants."
- Set the condition to Session source matches regex, for example:
(chatgpt\.com|chat\.openai\.com|perplexity\.ai|gemini\.google\.com|bard\.google\.com|claude\.ai|copilot\.microsoft\.com|grok\.com|meta\.ai|you\.com|phind\.com)
- Move the AI assistants channel above Referral in the list. GA4 assigns sessions to the first matching channel.
- Save.
Review the regex periodically. AI products add and change domains, and some, such as Copilot inside Edge, may send traffic from other Microsoft domains. Check your session source report for new AI referrers every quarter.
Step 3: Use it in reports
- In Traffic acquisition, switch the dimension to your new channel group's session channel.
- In Explorations, build a free-form report with your AI channel, landing page, sessions, engaged sessions, conversions and revenue.
- Compare AI assistants with Organic Search, Referral and Direct on engagement and conversion rate.
Step 4: Track key events and conversions
Make sure the events that matter, such as sign-ups, demo requests, trials and purchases, are marked as key events, so you can see conversions from the AI channel.
Useful views:
- AI sessions by landing page: which pages AI assistants send people to
- AI conversion rate vs other channels
- AI sessions by source: which assistant sends the most valuable traffic
Step 5: Understand what GA4 misses
GA4 shows only clicks that carry a referrer or UTM. It misses:
| Scenario | How it shows up |
|---|---|
| Link copied from an answer and pasted into a browser | Direct |
| Click from some mobile apps or in-app browsers | Direct or app referrer |
| User reads the answer and later searches your brand | Organic search (branded) |
| User types your URL after seeing your name | Direct |
| Click from Google AI Overviews or AI Mode | Google organic |
| No click at all, but your brand was recommended | Not tracked |
That's why AI referral traffic is a floor, not the full picture. See zero-click answers.
Step 6: Fill the gaps
- Self-reported attribution: add "AI assistant (ChatGPT, Perplexity, etc.)" to "How did you hear about us?" on sign-up and demo forms.
- Branded search trends in Search Console: rising branded impressions often follow increased AI visibility.
- Visibility metrics: track mention rate, share of voice and citations directly. See AI visibility metrics.
- Sales notes: ask reps to log when prospects mention AI research.
See attribution for AI search for how to combine these.
A monthly AI traffic report
- AI assistant sessions, by source, vs last month
- Top landing pages from AI assistants
- Engaged session rate and conversion rate vs other channels
- Conversions and revenue from the AI channel
- Self-reported "AI assistant" responses
- Notes on new AI referrer domains
What the numbers usually show
Teams that set this up often find a few consistent patterns:
- ChatGPT is usually the largest AI referrer, followed by Perplexity, then Gemini, Copilot and Claude, though this varies by audience.
- AI referral traffic is small relative to organic search but often engages and converts at rates comparable to or better than other channels, because visitors arrive with context.
- Landing pages differ from organic search: more pricing, comparison, documentation and specific feature pages.
- New referrer domains appear as AI products launch new surfaces, which is why the channel regex needs regular review.
Your own data may differ, which is exactly why it's worth measuring.
The bottom line
Set up an AI assistants channel in GA4, place it above Referral, track conversions from it and review new referrer domains regularly. Then remember what GA4 can't see: copied links, later branded searches and recommendations without clicks. Pair the data with self-reported attribution and visibility tracking to understand AI's full impact.
