How can I identify bot traffic in Google Analytics?

Bot traffic and LLM referral traffic are different: GA4 already filters known bots/crawlers by default under data settings. LLM traffic is human visitors clicking a link inside an AI assistant's answer, so it shows as a referral session, not a bot hit

How to track LLM usage?

If the question means visits arriving from LLM tools, use the traffic acquisition report and filter or segment by the assistant's referral domain. If it means usage of an LLM product itself, that's a separate analytics setup outside GA4's website traffic reports

How to see where traffic is coming from Google Analytics?

Open Reports > Acquisition > Traffic acquisition to see sessions broken down by default channel, source and medium. Sort or filter by session source to isolate a specific referral domain, including AI assistant domains

How to identify bot traffic in GA4?

GA4 excludes known bots and spiders automatically based on the IAB/ABC international spiders and bots list. Unusual patterns like unrealistic session counts or a single source spiking without matching engagement can indicate unfiltered bot activity, but this is distinct from AI assistant referral traffic

How to Identify LLM Traffic in Google Analytics (GA4)

Setting up custom filters is the only reliable way to identify LLM traffic in Google Analytics (GA4) before your AI referral data gets permanently lost in generic direct and referral buckets. Relying on default GA4 channel groups means missing visits from ChatGPT, Gemini, Claude, and Perplexity entirely. You can isolate these visits by changing your reporting dimensions to Session source/medium and building custom channel rules to group known AI domains. Since organic AI referral volume is often too low to test new filters immediately, using a controlled traffic source allows you to confirm your setup works in real time.

To isolate AI referral traffic, change your GA4 report dimension to Session source/medium and apply a filter containing known AI domains like chat.openai.com. For permanent tracking, create a custom channel group that automatically pools these specific domains into a dedicated "AI Traffic" channel.

What counts as LLM traffic in Google Analytics

Google Analytics 4 (GA4) does not have a dedicated default channel for AI assistants. Instead, GA4 usually buckets this traffic under the generic Referral category. If an assistant strips out the referrer data during the transit, the visit will instead end up categorized as Direct or Unassigned. This fragmentation means your raw channel reports will heavily undercount your actual AI audience.

Finding LLM traffic in the GA4 traffic acquisition report

The default reports bundle all inbound links into broad buckets. To find the raw data, navigate directly to your detailed acquisition metrics.

Open the Reports section located on the left panel of your Google Analytics 4 property.
Expand the Acquisition dropdown menu.
Click on the Traffic acquisition report.
Change the primary dimension of the table from "Session default channel group" to Session source/medium.

Selecting the Session source/medium dimension forces GA4 to surface the exact referral domains. This is the necessary first step for isolating AI assistant data before building permanent reporting filters.

How to identify LLM traffic in Google Analytics with filters and a custom channel group

You can use a temporary filter for a quick check, or build a permanent custom channel group for ongoing reporting.

Filtering the standard traffic acquisition report

A temporary filter on your standard traffic acquisition report isolates sessions where the source matches known AI domains.

Navigate to Reports, expand Acquisition, and click Traffic acquisition.
Click Add filter below the report title.
Select Session source from the dimension menu.
Choose contains as the match type.
Enter the specific AI domain you want to analyze, such as chat.openai.com, and click Apply.

Building a custom channel group for repeat reporting

Temporary filters fail when you need to analyze multiple AI platforms simultaneously. An exploration report allows you to combine various AI domains into a single segment, which you can then break down by landing page or geographic location.

For daily monitoring, a custom channel group is the most efficient option. This configuration creates a permanent classification rule within GA4. Any inbound traffic matching your specified list of AI domains is automatically grouped into a dedicated "AI Traffic" channel.

Recognizing referral sources for the major AI assistants

A single filter rarely catches all AI referral traffic at once because each AI assistant sends its own unique referral domain. To build accurate custom channels, you must know which domains to target. The bulk of your natural AI traffic will typically originate from four major engines: ChatGPT, Perplexity, Claude, and Gemini.

Identifying these main drivers requires looking for specific hostnames in your GA4 reports. You must know how the ChatGPT referral source appears to isolate traffic from OpenAI. It is equally important to recognize what a Perplexity referral looks like, as conversational search engines format their referral headers differently.

Depending on your audience, you may also receive traffic from a broader ecosystem of platforms. Monitoring this wider group ensures you capture the full spectrum of AI-driven discovery as user habits shift away from traditional search.

Why natural LLM traffic is often too thin to confirm your setup works

A newly built filter or custom channel group that displays zero sessions often means the configuration is correct, but your website simply does not receive natural AI referral traffic yet. Waiting weeks or months for organic mentions to slowly accumulate is an impractical way to verify your GA4 tracking rules are working.

You can proactively verify your tracking configuration by sending real LLM referral traffic to test your GA4 setup. PerkFuel sends real, targeted website visitors from AI assistants such as ChatGPT, Gemini, Claude, and Perplexity directly to your chosen landing pages. This allows you to dispatch a known batch of AI referral visits to confirm your GA4 filters catch them as expected.

These visits are fully tracked in the PerkFuel dashboard, and with analytics tracking enabled, they will appear within your Google Analytics reports. Watching these controlled visits populate your reports provides immediate confirmation that your custom channel setup is correct.

Transition your tracking from temporary filters to a permanent custom channel group to build a clean, ongoing record of your AI referral traffic. Deploy a targeted test campaign of live AI assistant visits to your site to instantly verify that your GA4 configuration correctly catches and groups your traffic.