How to Track ChatGPT Traffic in GA4:
AI Referrals, Shopify Reports and the Traffic You Cannot See

To track ChatGPT traffic in GA4, start with the AI Assistant channel in the default channel group, which catches referrers like ChatGPT, Gemini and Copilot. ChatGPT also adds utm_source=chatgpt.com to links. Build a custom channel group as a backup, check Shopify's Agentic reports, and ask buyers how they found you. Referrers alone undercount AI.

This guide covers what each tool shows as of October 2026, how to set it up step by step, and where AI influence hides in your reports. If you are still deciding whether AI search deserves budget, our AI search optimization guide for Shopify covers the full picture first.

Where does AI traffic show up in your reports today?

Each AI surface lands in a different place. This is the map.

The last two rows are the reason GA4 alone will undercount AI's role.

Does GA4 track ChatGPT traffic automatically?

Mostly, yes. GA4's default channel group now includes an AI Assistant channel. Google describes it as traffic "from sources like ChatGPT, Gemini, Deepseek, Copilot, or Grok." When the referrer matches Google's list of AI assistants, GA4 sets the medium to "ai-assistant" and the campaign to "(ai-assistant)."

Search Engine Journal reported the change on May 14, 2026. Three limits matter:

• The list is not published. Google names examples only. A new assistant can land in Referral until Google adds it.
• No referrer, no match. Visits from in-app browsers, mobile apps or copied links can still land in Direct.
• AI Overviews and AI Mode are excluded. Google counts those clicks as Organic Search.

Check it now: in GA4 go to Reports, then Acquisition, then Traffic acquisition, and look for an AI Assistant row. If you have segments or reports that filter on medium = referral, review them. AI sessions now carry a new medium value.

Does ChatGPT add utm_source=chatgpt.com?

Yes. OpenAI's Publishers and Developers FAQ says: "ChatGPT automatically includes the UTM parameter utm_source=chatgpt.com in referral URLs."

Two practical effects:

1. It survives a missing referrer. If the referrer header gets stripped, the UTM still tells GA4 and Shopify where the visit came from.
2. It only sets the source. No medium or campaign comes with it. Check where these sessions land in your property. If you see chatgpt.com sessions under Unassigned or Referral, a custom channel group that matches on source fixes it.

OpenAI ties this to sites that allow OAI-SearchBot. If you block that crawler in robots.txt, your pages cannot appear in ChatGPT search summaries in the first place. Our llms.txt and AI crawler guide for Shopify covers which bots to allow.

You can see the raw data in GA4 under Reports, then Acquisition, then Traffic acquisition. Switch the primary dimension to Session source and search for "chatgpt."

How do you build an AI channel group in GA4?

Build a custom group when you want control over which sources count as AI. Google's custom channel groups help page documents these steps:

1. In GA4, open Admin. Under Data display, click Channel groups.
2. Click Create new channel group. This copies the default group.
3. Name it, for example "Channels with AI."
4. Add a new channel and name it "AI assistants."
5. Set the condition to Source, matches regex, and paste: .*(chatgpt\.com|chat\.openai\.com|perplexity\.ai|gemini\.google\.com|copilot\.microsoft\.com|claude\.ai|deepseek\.com).*
6. Save the channel.
7. Reorder the channels and move "AI assistants" above Referral. Google says traffic goes to "the first channel whose definition it matches."
8. Click Apply, then Save group.

Facts to know before you start:

• Standard properties get 2 custom channel groups. GA4 360 gets 5.
• Google says custom groups "can be applied to your reports retroactively," so you see history too.
• Update the regex when a new assistant matters to you. Google's own example on that page says the same.

Then, in Traffic acquisition, switch the channel dimension to your new group. Our D2C GA4 setup guide covers the ecommerce events that make this data useful.

How do you measure Google AI Overviews and AI Mode traffic?

You cannot split AI Overviews clicks from regular organic clicks in GA4. Google's AI features page says that traffic is reported under the "Web" search type in Search Console, inside overall Search data.

What you can see is impressions. Search Console now has a Generative AI performance report. It counts how often links to your site were shown in AI Overviews and AI Mode, with page, country, device and date. Key facts:

• It rolled out to all websites worldwide as of August 31, 2026.
• It shows impressions only. Clicks stay in the standard Performance report.
• It may not appear if your site has too few AI impressions.

For a Shopify brand, use it this way: export the pages table monthly. If product and collection pages appear, your feed and schema work is reaching AI Mode. Our guide on Google AI Mode shopping covers the Merchant Center side.

How do you see AI referrals in Shopify analytics?

Shopify gives you two views.

Agentic performance reports. Go to Sales channels, then Agentic. Shopify's help page lists four metrics: Sales, Orders, Online store sessions and Online store conversions. Click a channel name, such as ChatGPT, for its own report. Sales combine referral sales and direct checkout sales. These reports are not yet available for headless stores.

Sessions by referrer. Go to Analytics, then Reports, and open Sessions by referrer. Shopify's acquisition reports page lists dimensions for referring channel, referring platform, referring medium and traffic type. Look for AI domains in the referring channel and platform columns.

One point matters for ChatGPT. Per Shopify's ChatGPT page, shoppers complete checkout on your online store, in a ChatGPT in-app browser or new tab. So ChatGPT purchases create sessions you can attribute. Copilot and Gemini direct checkout orders do not. Use the Agentic report as your source of truth for total AI channel sales.

Why does most AI influence show up as direct or branded search?

Because many shoppers never click the link in the answer.

A common path: someone asks ChatGPT for "a fragrance-free moisturizer under $40." Your brand is named. They open a new tab, search your brand on Google, and buy. GA4 credits Organic Search or Direct. ChatGPT gets nothing.

Add the visits with no referrer header, and AI's real contribution sits inside three numbers you already track:

• Direct sessions, especially to product pages, which people rarely type by hand.
• Branded search clicks in Search Console's Performance report. Filter queries containing your brand name.
• New customer orders with no clear first touch.

None of these prove AI caused the lift. Read them together. If branded search and direct sessions rise while paid spend stays flat, and your AI mentions grew in the same quarter, AI is a fair suspect. Our D2C attribution modeling guide covers how to weigh signals like this.

How do you add an AI option to "How did you hear about us?"

A post-purchase survey is the cheapest way to see AI influence that analytics misses.

Set it up in four steps:

1. Install a post-purchase survey app. Search the Shopify App Store for "post-purchase survey." Pick one that shows on the Thank you page and exports to your analytics or Klaviyo.
2. Write one question. "How did you first hear about us?"
3. Add AI as its own answer. Use "ChatGPT or another AI assistant" next to Google search, Instagram, TikTok, Friend or family, Podcast, and Other.
4. Randomize the order so the first option does not get extra clicks.

Then compare survey share against GA4 share every month. A fictional $2M apparel brand might see AI at 1% of GA4 sessions and 4% of survey answers. The gap is your untracked AI influence. Sync the answer to a Klaviyo profile property too, so you can compare repeat rate by source.

What should you expect AI traffic to look like?

Small in share, strong per visit. Adobe Digital Insights reported for US retail sites in July 2026:

• AI-driven visit share up 62% year over year.
• AI-referred visitors converted 60% better than non-AI traffic.
• They generated 53% more revenue per visit.
• They were 14% more engaged.

Adobe's data covers more than one trillion visits to US retail sites. Your numbers will differ. Track your own baseline from the week you set this up.

What should a weekly AI traffic check include?

Keep it to 10 minutes and five numbers.

If you want help wiring this into your reporting, you can book a call with our team. For the work that grows these numbers, see how reviews and mentions shape AI recommendations and how to get products into ChatGPT shopping.