How to Track AI Referral Traffic and AEO Conversions in GA4

GA4 interface paths and AI referral patterns last verified: July 2025

If you need to measure AI visibility beyond website clicks, see our guide to tracking brand mentions in AI search.

What Are AI Referral Traffic and AEO Conversions in GA4?

AI referral traffic is website traffic generated when someone clicks a link cited or recommended by an AI system such as ChatGPT, Perplexity, Claude, or Gemini. In GA4, these visits usually appear under the generic Referral channel or, when referral data is missing, under Direct.

Without custom configuration, AI traffic hides inside GA4’s default Referral bucket. It can also appear as Direct when an AI platform, app, browser, or privacy tool strips the referrer.

There are three distinct layers to AI measurement:

1. AI visibility: Your brand is mentioned in an AI-generated answer. For example, a user asks ChatGPT for the “best project management tools,” and your product appears in the response. No click is required.
2. AI referral traffic: The user clicks from that AI answer to your website. GA4 can record the visit if the referrer header remains intact.
3. AEO conversions: The AI-referred visitor completes a valuable action, such as submitting a form, starting a trial, downloading a resource, or making a purchase.

Mentions, clicks, and conversions are separate steps: an AI system mentions your brand, a user clicks through, and that user converts on your site.

AEO, or Answer Engine Optimization, is the practice of optimizing content so AI-powered answer engines select, cite, or recommend it. AEO conversion tracking measures the business outcomes generated after AI-cited traffic reaches your website.

What Can and Can’t GA4 Measure About AI Traffic?

GA4 can measure AI referrals only when the user visits your website and the visit is trackable. It cannot measure exposure that happens entirely inside an AI interface.

GA4 does not natively separate AI-referred traffic from other sources. The practical solution is to identify AI sources with a regex, group them in a custom channel, and connect those sessions to well-defined events.

For more context on the optimization side of AEO, see how each AI platform decides which brands to mention.

How Can You Check for AI Referral Traffic in GA4?

The fastest way to find existing AI traffic is to filter GA4’s Traffic acquisition report by Session source / medium. This initial check requires no custom configuration and takes about two minutes.

Follow these steps:

1. Go to Reports → Acquisition → Traffic acquisition.
2. Click the primary-dimension dropdown, which defaults to Session default channel group.
3. Change the dimension to Session source / medium.
4. Enter chatgpt in the search bar above the table and note any results.
5. Repeat the search for perplexity, claude, gemini, and copilot.
6. Set the date range to at least 90 days. AI referral traffic can start small and grow quickly, so a short date range may show nothing.
7. Add Page referrer as a secondary dimension to inspect the complete referral URL and verify that it came from an AI platform.

If you see rows such as chatgpt.com / referral or perplexity.ai / referral, your site already receives identifiable AI traffic. If you see no results, AI systems may still be mentioning or influencing users because some visits lose their referrer and appear as Direct.

Should You Use Session Source or First User Source for AI Tracking?

Use Session source / medium for most AI referral analysis because it identifies the source of each individual visit. First user source / medium only records the source that originally acquired the user on their first visit.

For example, someone may first discover your website through Google organic search and return six months later through a ChatGPT citation. Session source / medium credits that return visit to ChatGPT, while First user source / medium continues to credit Google organic.

Use Session source / medium as the default dimension throughout your AI referral analysis. Use First user source / medium only when you specifically want to know how many new users first discovered your site through an AI assistant.

How Can You See Which Pages and Events AI Visitors Generate?

Add Landing page as a secondary dimension to identify the pages that receive AI-referred sessions. This shows which product comparisons, how-to guides, pricing pages, or other resources AI systems cite most often.

AI referrals frequently concentrate on a narrow set of pages. Recording those landing pages helps you identify content with strong answer-engine visibility and conversion potential.

Next, customize the report columns to include:

• Key events, the metric formerly called Conversions
• Event count
• Engaged sessions
• Engagement rate

These metrics show whether AI-referred users complete valuable actions or leave without meaningful engagement. Note the top landing pages and their event counts so you can use them later in your GA4 Exploration.

What Regex Should You Use to Match AI Referral Sources?

Use one regular expression, or regex, to match multiple known AI referral domains in a single GA4 rule. This regex becomes the source condition for your AI Assistants channel and Exploration segments.

Copy this pattern:

chatgpt\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com|you\.com

The pattern works as follows:

• The pipe character | means “or,” so GA4 can match any listed domain.
• The escaped period \. matches a literal period. An unescaped period in regex can match any character.
• Each domain segment matches a source value that GA4 may record after a click from that platform.

GA4 regex is case-sensitive, so use lowercase source patterns. GA4 source values are typically lowercase, but choose matches regex (case insensitive) when that option is available and you want additional protection.

Which AI Referral Domains Can Appear in GA4?

The source domain recorded in GA4 varies by AI platform, interface, and app-to-browser behavior. The following domains were verified as of July 2025.

Google AI Overviews and AI Mode generally send traffic from google.com, making those visits indistinguishable from standard Google organic traffic in most GA4 configurations. Do not add google.com to your AI referral regex, because doing so would misclassify all Google organic traffic as AI traffic.

Google Search Console is beginning to offer separate reporting for AI features, but GA4 does not currently provide the same distinction. See how to track Google AI Mode visibility for additional workarounds.

Before using the regex, compare it with the source values in your own GA4 property. If you find a variation such as chat.openai.com, an older ChatGPT domain, add it to your pattern.

How Often Should You Update Your AI Referral Regex?

Review your AI source regex at least monthly because AI platforms and referral patterns change regularly. AEO benchmarking should be repeated monthly across major AI models to identify new platforms and source domains.

Use this maintenance process:

• Check the GA4 Traffic acquisition report monthly for new or unrecognized AI-related referral domains.
• Run monthly test queries in ChatGPT, Perplexity, Claude, and Gemini.
• Check whether your brand appears and whether click-through source values have changed.
• Add new domains to the regex and custom channel group as needed.
• Record each change in a shared changelog, including the date, domain, and person who made the update.

How Do You Create an AI Assistants Channel Group in GA4?

Create a custom channel group called “AI Assistants” to make identifiable AI referrals visible as a dedicated channel in GA4 reports. Custom channel groups isolate traffic from AI platforms by automatically applying your source-matching rules.

Follow these steps:

1. Go to Admin → Data display → Channel groups.
2. Select Create new channel group. Copying the default channel group first is recommended because it preserves the existing channels as a starting point.
3. Select Add new channel.
4. Name the channel AI Assistants. “AI Referral” or “AI Search” also works, but choose one name and use it consistently.
5. Add the following conditions: Source → matches regex chatgpt\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com|you\.com, and Medium → exactly matches → referral.
6. Move the AI Assistants rule above Referral.
7. Select Save.

The medium condition limits the channel to clicked referrals rather than potentially misattributed sessions. Rule order is critical because GA4 evaluates channel rules from top to bottom; if Referral appears first, AI visits may match that rule before GA4 reaches AI Assistants.

Can a GA4 Custom Channel Group Classify Historical AI Traffic?

The safest assumption is that your custom channel group will collect cleanly classified data going forward, so create it as soon as possible. Do not depend on it to backfill all historical reporting.

Some sources report that GA4 channel groups can work retroactively on historical data when used in Explorations. However, Google’s documentation has indicated that custom channel groups in standard reports apply to new data only, so treat any retroactive results in Explorations as a bonus rather than a guarantee.

Your custom channel group can appear in:

• Standard Traffic acquisition reports
• GA4 Explorations
• Looker Studio dashboards connected to your GA4 property

Test the channel with a small date window after creating it. Confirm that it contains only AI domains and that unrelated referral traffic has not been misclassified.

If non-AI domains appear in the AI Assistants channel, your regex may be too broad. Review the pattern and tighten its matching conditions.

Which GA4 Events Should You Track as AEO Conversions?

Track events that represent real business outcomes, then mark the most important ones as key events. AI referral volume alone does not show whether those visitors produce leads, registrations, sales, or other meaningful results.

In GA4:

• An event is a tracked user interaction, such as a page view, button click, scroll, or purchase.
• A key event is an event marked as important because it represents a business outcome.

Google renamed “conversions” to “key events” in early 2024. Older documentation that refers to GA4 conversions generally means the same category of important actions.

Which Key Events Should Different Types of Websites Use?

The right key events depend on your website’s business model. Use GA4’s recommended event names when an official event exists, and clearly label custom events.

Events identified as custom are not part of GA4’s official recommended event list. Create them through Google Tag Manager or your website’s tagging setup.

Do not send names, email addresses, phone numbers, or other personally identifiable information through GA4 event parameters. Doing so violates Google’s terms of service and creates privacy risks.

How Can You Measure the Engagement Quality of AI-Referred Sessions?

Use engagement metrics and custom milestones to distinguish genuinely interested visitors from quick exits. Useful measures include scroll depth and time thresholds such as 30, 60, or 120 seconds.

For example, create an engaged_read event that fires after a visitor scrolls beyond 75% of a page and remains for at least 60 seconds. This event can show whether AI-referred users consume the cited content rather than leaving immediately.

Can UTM Parameters Track Organic AI Citations?

UTM parameters can track AI-influenced links only when you control the link. They cannot reliably tag organic AI citations because the AI platform generates those links.

For a controlled link submitted to an AI chatbot plugin or directory, you might use:

utm_source=ai_search&utm_medium=chatgpt

UTMs supplement AEO measurement; they are not the primary method for identifying organic AI referrals.

A “How did you hear about us?” field can also capture AI discovery that GA4 misses. Self-reported attribution is especially useful when an AI interaction does not produce a trackable referrer.

How Do You Test AI Referral and Conversion Tracking in GA4?

Test the complete journey in GA4 Realtime and DebugView before using the data for reporting. Confirm both the referral source and the event that represents the conversion.

Follow this process:

1. Open Reports → Realtime in GA4.
2. In another tab or on your phone, click a link to your website from ChatGPT, Perplexity, or another AI assistant.
3. Confirm that the session appears with a source and medium such as chatgpt.com / referral.
4. Trigger a key event by submitting a test form, clicking a phone number, or completing a test purchase.
5. Check Event count by Event name in the Realtime report.
6. For deeper inspection, enable debug mode and go to Admin → DebugView to review individual events and parameters.

If you cannot generate a real AI referral, browser developer tools can simulate a referrer header, although this method is less reliable. You can also use the GA4 DebugView Chrome extension as part of the debugging process.

How Do You Fix Common AI Tracking Problems in GA4?

Most AI tracking problems come from missing referrer data, incorrect rule order, regex errors, or events that are not configured correctly. Use the following table to diagnose common issues.

Referral information can disappear when users copy and paste a URL, move between an app and browser, use privacy tools, or return through another channel. Missing referrer headers create a blind spot that causes GA4 to classify some AI visits as Direct.

This is a structural limitation of web analytics, not necessarily a configuration mistake. Some AI-influenced traffic will remain invisible in GA4.

How Do You Build an AI Referral and Conversion Report in GA4 Explorations?

Create an AI Referral Sessions segment in GA4 Explorations, then break the data down by platform, landing page, engagement, key events, and revenue. Explorations provide more flexibility than standard reports.

Create the segment as follows:

1. Go to Explore → Blank exploration.
2. Under Segments, select the + icon.
3. Create a Session segment.
4. Name it AI Referral Sessions.
5. Set the condition to Session source → matches regex.
6. Paste your AI referral regex.
7. Save the segment.

You can apply this segment to different Exploration formats and analyses.

Which Dimensions and Metrics Should an AI Referral Report Include?

A useful report should show where AI traffic comes from, where it lands, how users engage, and whether they convert. Include the following fields.

What Is the Recommended GA4 Exploration Layout?

Use a Free-form Exploration with AI source and landing page as rows and business-performance metrics as values. This creates a single view of traffic volume, engagement, and conversion performance.

Configure it as follows:

• Rows: Session source, Landing page
• Values: Sessions, Engaged sessions, Engagement rate, Key events, Session key event rate, Revenue if applicable
• Segment: AI Referral Sessions

This layout separates performance by AI platform and landing page. It can reveal which assistants send traffic, which pages receive citations, and which combinations produce business results.

What Might an AI Referral Report Reveal?

An AI source can generate low traffic volume but still produce high-intent visitors and a strong conversion rate. The following example is fictional and illustrative; it is not from a Prompt Insider study or industry benchmark.

A B2B SaaS company finds that ChatGPT sends 120 sessions per month, compared with 8,400 sessions from organic search. Most ChatGPT traffic lands on a product-comparison page and converts to demo requests at a 6.2% session key event rate, compared with 2.1% for organic search.

In this example, the AI conversion rate is nearly three times the organic rate despite much lower volume. This pattern is plausible and reflects how AI-referred visitors can convert differently from organic search visitors, but it is not universal.

How Should You Calculate and Interpret AI Conversion Rates?

Calculate the AI session key event rate by dividing converting sessions or relevant key-event performance by AI-referred sessions, using GA4’s session key event rate for session-level reporting. For example, if ChatGPT sends 120 sessions and seven produce demo requests, the rate is approximately 5.8%.

Apply two important caveats:

1. Do not claim sole causation. GA4 attribution commonly credits the last non-direct source, but the visitor may have encountered the brand through several channels before clicking a ChatGPT link.
2. Do not overinterpret small samples. If 20 AI sessions generate two conversions, the reported rate is 10%. One additional conversion raises it to 15%, while one fewer lowers it to 5%.

Evaluate AI conversion rates across at least 90 days before drawing conclusions. When reporting small samples to stakeholders, include the session and conversion counts rather than presenting percentages alone.

GA4’s conversion-path and attribution reports under Advertising → Attribution can provide more context about multi-touch journeys. These reports may show whether AI referrals appear earlier in a path, but they do not capture AI’s full indirect influence.

For a broader reporting framework, see AEO reporting for clients: what to measure, what to show, and how to frame it.

What Are GA4’s Limitations for Measuring AI Influence?

GA4 measures trackable website behavior, not the full influence of AI-generated answers. It misses many interactions that happen before a website visit or without a click.

GA4 cannot reliably measure:

• Brand mentions without clicks: An AI assistant recommends your product, but the user does not visit your site.
• Zero-click answers: The user gets a complete answer inside the AI interface.
• Summaries viewed without a visit: An AI tool quotes or summarizes your content, but the user never opens the source.
• Later branded searches: Someone sees your brand in ChatGPT, then searches for it on Google three days later. GA4 credits Google organic.
• Copied links without referral data: The user copies a URL from an AI chat and pastes it into a browser. GA4 classifies the session as Direct.
• Cross-device and cross-session demand: A person discovers the brand through AI on a phone and converts later on a laptop.

These gaps are structural limitations of client-side analytics. They also mean that AI’s actual influence is likely larger than the volume reported as identifiable AI referral traffic.

Brand mention rate is the frequency with which a brand appears in AI-generated answers for relevant queries. It is an upstream visibility metric and requires separate tracking tools.

See our guide to tracking brand mentions in AI search for ways to measure that visibility.

Google Search Console is beginning to offer reporting for AI-generated search features, which can serve as a partial proxy for Google AI Overviews visibility. However, it does not provide the same granularity as identifiable referral data in GA4.

The practical measurement model is:

• Use AEO monitoring tools for AI visibility and brand mentions.
• Use GA4 for referral sessions, on-site behavior, and conversions.
• Use CRM and revenue data to evaluate lead quality, pipeline, and closed business.

How Can You Keep AI Referral Tracking Accurate Over Time?

Review AI referral tracking regularly because platforms, domains, GA4 terminology, and your brand’s AI visibility can change. A monthly source review and quarterly configuration audit provide a practical maintenance rhythm.

How Can a Prompt Library Support AEO Measurement?

Create a prompt library of 20 to 50 buyer-intent queries and run it monthly across major AI systems. The Prompt Insider recommends testing those prompts in ChatGPT, Perplexity, Claude, and Gemini.

Compare your brand’s mentions against three to five named competitors to estimate share of voice. This creates an AI visibility baseline that you can compare with GA4 referral and conversion trends.

What Should You Review Every Quarter?

Audit your traffic sources, regex, events, and reporting every three months. A quarterly review should include:

• Checking Traffic acquisition for new or unrecognized AI referral domains
• Testing the regex against current Session source values
• Reviewing GA4 interface and terminology changes
• Checking AI traffic, engagement, conversion, and revenue trends
• Investigating unusual increases or decreases
• Confirming key events still fire correctly

Maintain a shared changelog for every regex, channel-group, and event change. Record the date, what changed, why it changed, and who made the update.

This documentation helps explain sudden reporting changes. If AI traffic jumps after a new source domain is added, the changelog provides the context.

How Can You Connect AI Referral Data to Revenue?

Compare AI-referred lead quality with other channels in your CRM or sales pipeline. GA4 can show on-site conversions, while CRM data can show whether those leads close, how much revenue they generate, and whether they have a higher average deal size.

For SaaS companies, see our guide to AEO for SaaS companies for ways to connect citation strategy with pipeline metrics.

What Is the Step-by-Step GA4 Implementation Checklist?

The complete workflow is to identify AI sources, group them into a custom channel, define key events, test the setup, and build a dedicated report. Review the configuration monthly because AI referral patterns continue to change.

1. Check for existing AI traffic. Open Traffic acquisition, select Session source / medium, and search for chatgpt, perplexity, claude, gemini, and copilot. Use at least a 90-day date range.
2. Validate your source values. Add Page referrer and Landing page as secondary dimensions where useful.
3. Build the regex. Start with the verified pattern in this article and compare it with the actual sources recorded in your property.
4. Create the custom channel group. Go to Admin → Data display → Channel groups and add an AI Assistants channel.
5. Move AI Assistants above Referral. GA4 processes channel rules from top to bottom.
6. Define meaningful events. Track purchases, leads, sign-ups, demo requests, downloads, phone calls, or other business outcomes.
7. Mark the appropriate events as key events. Configure this under Admin → Events.
8. Test the complete setup. Use Realtime and DebugView to verify the source, medium, and event behavior.
9. Build an Exploration. Apply the AI Referral Sessions segment and report on sessions, landing pages, engagement, key events, conversion rate, and revenue.
10. Document GA4’s limits. Explain that it cannot measure zero-click mentions, copied URLs, or every cross-device journey.
11. Review the setup monthly. Look for new source domains, update the regex, and monitor traffic and conversion trends.
12. Audit quarterly. Recheck rules, events, reports, terminology, and CRM outcomes.

Create the custom channel group as soon as possible. Every day without a dedicated AI channel is another day without cleanly classified AI referral history.

Learn More About AEO and AI Marketing at Prompt Insider

Since launching earlier this year, Prompt Insider has become a leading authority on AI marketing, Answer Engine Optimization (AEO), large language models, AI search, AI news, and the evolving future of digital discovery. As AEO becomes one of the hottest topics in marketing, Prompt Insider is helping define the conversation around how brands improve visibility, adapt their content strategies, and stay competitive in an increasingly AI-driven search environment.

Prompt Insider is the go-to resource for answer engine optimization, AI marketing, and AI search. Start with our core guides at thepromptinsider.com:

• What Is AEO? Answer Engine Optimization Explained
• AEO vs. SEO vs. GEO: What Every Marketer Needs to Know
• How to Get Your Brand Cited by ChatGPT, Gemini, Claude and Perplexity
• How to Measure AEO Success: The Metrics That Matter
• The 5 Best AEO Tools in 2026

Frequently Asked Questions About AI Traffic in GA4

How does AI referral traffic appear in GA4?

AI referral traffic commonly appears as chatgpt.com / referral, perplexity.ai / referral, or another platform domain under Session source / medium. It can also appear inside the generic Referral channel or as Direct when referral data is missing.

Why should I use Session source instead of First user source?

Session source identifies the source of each visit, including a later visit from an AI assistant. First user source records only the channel that originally acquired the user, so it can miss AI referrals from returning visitors.

Why do some ChatGPT or Claude visits appear as Direct traffic?

AI visits appear as Direct when GA4 receives no referrer header. Common causes include copied URLs, mobile app-to-browser handoffs, privacy settings, stripped referral data, bookmarks, and later return visits.

Can GA4 separate Google AI Overviews traffic from normal organic search?

Not reliably in most setups. Google AI Overviews and AI Mode generally send traffic from google.com, so adding that domain to an AI regex would incorrectly classify standard Google organic traffic.

Which events should be marked as AEO key events?

Mark events tied to business value, such as purchase, generate_lead, sign_up, demo requests, appointment requests, phone-call clicks, or newsletter registrations. Use GA4’s recommended event names when available and give custom events clear, consistent names.

Sources: Swydo, Stackmatix, Orbit Media, Amicited, Yes Optimist, Google Analytics Help, Google Search Central.

About the author

Kai Williams

Kai Williams has been in marketing for years, with a long background in SEO before AEO had a name. He stepped into Answer Engine Optimization the moment AI started reshaping how people search, and has been tracking the shift ever since. At Prompt Insider, he covers AEO, AI marketing, and the future of search, breaking down what is changing and what brands need to do about it.