How to See ChatGPT and Perplexity Visitors in Google Analytics

How to See ChatGPT and Perplexity Visitors in Google Analytics

AI assistants are already sending people to business websites — but GA4 buries them under "Referral". Here is how to make them visible, step by step.

8 July 2026
4 min read
By Codifyany Team
AnalyticsAI SearchHow-To

Someone asks ChatGPT for a recommendation, gets your business in the answer, clicks through, and becomes a customer. Did that ever happen to you? For most business owners the honest answer is: no idea. Google Analytics does not have an "AI" bucket, so these visits get filed under generic Referral traffic — or disappear into Direct.

This guide shows you how to see them. It takes about fifteen minutes in GA4 and does not require any code on your website.

Step one — check what is already there

Before building anything, see if AI referrals are already in your data.

1. Open Google Analytics and go to Reports, then Acquisition, then Traffic acquisition
2. Change the primary dimension to "Session source / medium"
3. In the search box above the table, type chatgpt

Repeat the search for perplexity, copilot, gemini, and claude. If you see sources like chatgpt.com / referral, those are real people who clicked from an AI conversation to your website. Many NZ businesses already have a trickle of these and have simply never looked.

Step two — build a channel for AI traffic

Searching manually gets old. GA4 lets you define a custom channel group so AI referrals get their own row in every report.

1. Go to Admin, then under Data display choose Channel groups
2. Click "Create new channel group" (this copies the default one — the default itself cannot be edited)
3. Add a new channel called AI Referrals and move it to the top of the list, because GA4 applies the first rule that matches
4. Set the condition to: Source matches regex

For the regex value, use a pattern like this:

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

Save it. From now on, any report that supports channel groups can show AI Referrals as its own line — sessions, conversions, revenue if you have e-commerce tracking.

Step three — understand what the number really means

Two honest caveats before you read too much into your new channel.

First, it undercounts. Visits from mobile apps often arrive with no referrer at all, so a chunk of genuine AI-driven traffic lands in Direct where nothing can identify it. Treat your AI Referrals number as a floor, not the full picture.

Second, it only counts clicks. When an AI assistant answers a question using information from your website but the person never clicks through, no analytics tool on your own site can see that. The click-through is the visible tip of a larger effect.

What counts as a good result

There is no universal benchmark, and anyone quoting one is guessing. What matters is the trend: if AI referrals are growing month over month while you improve how clearly your site describes what you do, the work is paying off. In our experience these visitors tend to arrive well-qualified — they asked a specific question and were pointed specifically at you — so watch their conversion rate, not just the count.

If you would rather have this set up and monitored for you, measurement like this is part of our monthly marketing service — but the steps above are genuinely all you need to do it yourself.

More Articles

8 July 2026
4 min read

What Our Free Site Audit Checks — and Why Each Check Matters

A plain-language tour of every category in our free website audit: what we test, what the scores mean, and what the tool honestly can and cannot tell you.

Read Article
8 July 2026
3 min read

One Question Becomes Twenty: How AI Search Splits Up What Your Customers Ask

AI search does not process your customer's question — it explodes it into many smaller ones. Understanding this changes how you should write content.

Read Article
8 July 2026
3 min read

Your Next Website Visitor Might Be an AI Agent

A growing share of website visits are made by AI tools acting on someone's behalf. They are immune to good looks — and picky about substance.

Read Article