How to track referral traffic from AI platforms & LLMs

By Thierry Ngutegure, Head of Data Insights
May 19, 2025

Google may still reign over traditional search, but artificial intelligence (AI) is quickly emerging as a formidable challenger. AI-powered search delivers faster, more specific answers to users’ queries, with less effort than conventional search engines require. As a result, a new kind of website traffic has emerged: AI referral traffic.

AI referral traffic is website traffic from large language models (LLMs) such as ChatGPT, Gemini, Copilot, and DeepSeek. Monitoring your LLM traffic provides critical insights into your website’s performance, enabling you to make targeted website improvements for your business’s success. This article will walk you through how to track your AI traffic on Google Analytics 4 (GA4) and Looker Studio.

How much traffic are you getting from LLMs?

According to an AI traffic study of 3,000 sites by marketing intelligence tool Ahrefs, 63% of sites receive at least one visitor from an AI chatbot. As such, you have roughly a two-in-three chance of receiving an AI visitor. We also know from our own LLM user intent studies that users who click through from LLMs show similar engagement potential to those originating from organic search.

The study also found that of the total LLM traffic, ChatGPT traffic, Perplexity traffic, and Gemini traffic made up 98%, with ChatGPT accounting for 50%. The remaining less than 2% came from Claude and three other AI chatbots: Jasper, Copilot, and Mistral.

As for overall monthly traffic, only 0.17% of a site’s monthly visitors make up AI referrals. However, the percentage of AI views and visitors is much higher for sites with less than 1,000 visitors per month.

Why segment LLM traffic?

Referral traffic from LLMs is rapidly increasing. Gartner, a leading technology research and advisory firm, predicts that AI chatbots and other virtual agents will lead to a 25% drop in traditional search engine volume by 2026. This shift in how consumers find information and discover brands drives the need for monitoring, analysing, and leveraging AI traffic to refine search marketing strategies.

The solution? Traffic segmentation. Segmenting traffic means categorising your website visitors into different groups based on traffic sources. By segmenting your LLM traffic, you can see and understand:

What percentage of your total traffic comes from AI
Which AI platforms send you the most traffic
What type of content attracts more AI-driven traffic
Which topics bring high-quality AI referral traffic.

Two ways to segment LLM and chatbot traffic

You can track referral traffic from ChatGPT and other LLMs through GA4 exploration reports or Looker Studio reports and dashboards.

GA4 is an analytics platform that collects, processes, aggregates, and organises data from websites and apps to offer insights into the customer journey. This information includes traffic sources such as search engines and AI chatbots.

Looker Studio is a tool that connects to multiple data sources like GA4, enabling you to showcase data in interactive dashboards and engaging reports that are easy to read, share, and customise.

In a nutshell, GA4 is a data source, while Looker Studio is the data visualiser. Let’s look at how to use these tools to track your LLM traffic.

GA4 explore Reports snapshot

To segment LLM traffic with GA4 Exploration, follow these steps:

Log in to your GA4 account and go to Explore.
Click the “+” symbol to create a new Exploration report.
Set Dimensions to Session source/medium and choose Views as the primary metric. You can select more metrics, such as Key Events, Engaged sessions, and Engagement Rate.
Click the “+” symbol beside Segments and create a new session segment. Give it a name like “AI traffic.”
Define the segment using the conditions Session source and matches regex.
Paste and apply the following regular expression to filter for LLM traffic sources:

Stay on top of new LLM tools to keep the regex updated.

GA4 Explore traffic analysis

Having set up a custom “AI traffic” segment, it’s time to visualise your LLM referral traffic over time and draw impactful insights. To do this, follow these steps:

Change the visualisation to a line chart.
Set values to Sessions.
Choose your preferred date range.
Adjust granularity to daily, weekly, or monthly, depending on your analytical needs.

GA4 custom channel group

Another way to monitor site traffic from LLMs is through GA4 channel groups. A channel group is a category of website traffic sources grouped based on predefined conditions. There is currently no default channel group catering to AI traffic, but you can create a custom Google Analytics AI traffic channel group. Here’s how:

Go to Admin and select Channel Groups under Data display.
Click Create new channel group and enter a group name like “AI Chatbots.”
Click Add new channel and name the channel, for example, “LLM traffic.”
Under channel conditions, select Source, then matches regex.
Copy/paste the same regex you added to create your exploration report:

Save the channel.
Reorder the channels, giving “LLM traffic” priority over Referral.
Save the group.

GA4 traffic acquisition report

Give your channel time to collect data before checking your AI report by selecting Reports > Acquisition > Traffic acquisition. Select your new channel group from the drop-down menu at the top and adjust the date range to your preference.

The generated report will show how your LLM traffic compares to other traffic sources like “Organic Search” and “Direct.”

Reporting in Looker Studio

You can also visualise your AI traffic source data in Looker Studio by following these steps:

Create a Blank Report.
Click Data source and select Google Analytics.
If prompted to give Looker Studio access to your GA4 account, click AUTHORISE.
Choose the new channel group and click CONNECT.
Click CREATE REPORT, and a table with fields from the channel group will appear.
Click ADD TO REPORT.
You can then start customising your AI traffic report by adding charts and controls, colouring your data, adding comparison metrics, editing fields, and more.

A new ecosystem

The search ecosystem has new players: LLMs such as ChatGPT and Gemini deliver the most relevant results faster and more efficiently. AI-powered search is quickly becoming a mainstream alternative to traditional search engines, making tracking and segmenting LLM referral traffic crucial for maintaining content relevance and discoverability.

GA44 and Looker Studio help businesses track, analyse, and leverage traffic from LLMs efficiently. Insights gained from segmenting this data help refine SEO strategies and improve user experience, contributing to overall website performance. If you want to know more about how search is changing and how it could affect your business, get in touch.