How Do You Set Up an AI Assistants Channel Group in GA4?

Why are AI assistant visits hiding inside your Referral traffic in GA4?

Because unless you tell Google Analytics 4 otherwise, AI assistant visits get sorted into broader channels, most often Referral. A custom channel group fixes that. You define an AI Assistants channel with a source rule, place it above Referral, and GA4 reports those visits separately.

This tutorial is for one specific reader. You are a marketer or founder at a B2B company, you have a standard GA4 property, and you want to see how much traffic AI assistants like ChatGPT, Perplexity, Gemini, Copilot, and Claude send to your site. You want it as a clean line in your reports, not a manual filter you rebuild every month.

Everything below follows Google's own help documentation for custom channel groups, which even uses an AI assistants channel as its example. I will add the decisions I make around it, because the setup is quick but the details decide whether the numbers are trustworthy.

What is a custom channel group in GA4?

A custom channel group is your own set of rules for sorting traffic into channels, built alongside GA4's default channel group. You define each channel with conditions on dimensions such as Source or Medium. Google's documentation says custom channel groups can be applied to your reports retroactively, so you see historical data too.

The default channel group is useful but generic by design. A custom group lets you keep the default logic for everything else while carving out the channel you care about.

There are limits. According to Google's help page, a standard property can have two custom channel groups in addition to the predefined one, and each group can hold up to fifty channels. Analytics 360 properties get five custom groups. For most B2B teams, one custom group is plenty.

You also need the right access. Google says you must be an Editor or above at the property level to create or edit channel groups. If you cannot see the option, check your role before troubleshooting anything else.

Step 1: How do you find which AI sources already send traffic?

Open the Traffic acquisition report, switch the primary dimension to session source, and look for the domains of AI assistants in the list. Copy each source value exactly as GA4 shows it. Your rule will only be as good as this list, so take the values from your own data, not from someone else's template.

Set a long date range, such as the last six months, so smaller sources appear. Search the table for the names of the assistants your buyers are likely to use. Write down every matching source value in a simple note.

Expect the list to change over time. AI products launch new apps, change domains, or add new surfaces. Plan to repeat this check every quarter and update your rule, the same way you would maintain any tracking convention.

Be aware of a blind spot. Some visits from AI apps may arrive without referrer information, and those can land in Direct instead. A channel group can only sort what GA4 can see, so treat your AI Assistants number as a floor, not a full count.

Step 2: How do you create the custom channel group?

In GA4, go to Admin, then Data display, then Channel groups. Click Create new channel group, which starts from a copy of the default group. Give it a clear name, such as "Default plus AI assistants," and an optional description that explains what you changed and when.

Starting from a copy of the default group is the safest approach. Every other channel keeps its familiar definition, so your Organic Search, Paid Search, and Email numbers stay comparable with what you reported before.

Use the description field properly. Note the date, your name, and the reason for the group. Six months from now, someone will wonder why the property has two channel groups, and that line will save a confused conversation.

Do not save yet. The group needs its new channel and the right order before it is useful.

Step 3: How do you add the AI Assistants channel?

Click Add new channel, name it AI Assistants, then click Add condition group. Choose Source as the dimension, choose the matches regex match type, and enter a pattern that covers the source values you collected in step one. Save the channel. Google notes that channel definitions are not case sensitive.

Keep the regex simple and readable. A pattern that lists each source separated by a pipe character is easy for the next person to check and update. Clever patterns that try to catch every possible variation are hard to maintain and easy to break.

Test your pattern before you rely on it. Paste your list of source values into any regex tester along with your pattern and confirm each one matches. Also test a few normal referral sources to make sure the pattern does not catch them by accident.

If you track AI traffic from campaigns with UTM parameters, for example links you placed in an AI tool's directory, decide whether those belong in this channel or in your campaign channels. I wrote about a UTM naming convention spreadsheet for GA4, and keeping the rules consistent there avoids double counting here.

Step 4: Why does channel order matter, and how do you set it?

GA4 assigns each session to the first channel in the group whose rules it matches, so order decides the result. Google's guide recommends placing AI Assistants above Referrals. Click Reorder, drag AI Assistants above Referral, click Apply, then click Save group. If you skip this, AI visits will still land in Referral.

This is the step people miss most often. The channel looks right, the regex is right, and the numbers still show zero, because Referral sits higher and claims the traffic first. Always check order after adding any channel.

Think about other channels too. If you have custom channels for partner sites or review platforms, decide whether they sit above or below AI Assistants. The rule is simple: the more specific channel goes higher.

Once saved, the group is available in your reports. Because Google says custom groups apply retroactively, you can immediately compare past months without waiting for new data.

Step 5: How do you report on the new channel?

Open the Traffic acquisition report and switch the primary dimension to your new custom channel group. AI Assistants appears as its own row. Add key events, such as form submissions or demo requests, to see whether these visits convert, not just whether they arrive. That turns a curiosity metric into a pipeline metric.

Look beyond sessions. Check engagement, landing pages, and conversions for the AI Assistants row. Landing pages are especially useful, because they show which of your pages assistants send people to. Those pages deserve priority when you update content.

Save a comparison or build a simple exploration so you can return to the view each month without rebuilding it. Share it with sales leadership if AI traffic is converting, because that is the kind of evidence that changes content budgets.

For the bigger picture, combine this with how often assistants mention your brand. I covered how to measure AI visibility for your brand, and the channel group supplies the traffic half of that story.

Step 6: How do you connect AI traffic to your CRM?

Capture the session source or a channel label when someone submits a form, and store it on the contact record in your CRM. Then you can see whether leads who first arrived from AI assistants turn into meetings and deals. GA4 shows the visit. Your CRM shows the revenue.

The simplest method is a hidden form field filled with the referrer or UTM data at the moment of submission. Many form tools and CRMs can do this. Store it in a dedicated property, such as "First touch source," so it does not get overwritten later.

Once the data is in your CRM, build a small report: leads, meetings, and opportunities by first touch source, with AI assistants as one group. Review it monthly alongside the GA4 view.

I wrote about AI search referral traffic attribution in more depth. The channel group is the starting point, and the CRM connection is what makes the numbers matter to sales.

What should you do next?

Pull your AI source values from the Traffic acquisition report, create a custom channel group from a copy of the default, add an AI Assistants channel with a tested regex on Source, move it above Referral, and save. Then add key events to the report and pass first-touch source into your CRM.

Set a quarterly reminder to recheck sources and update the regex. AI products change faster than most tracking setups, and a stale rule quietly undercounts.

I build websites and the automations behind them for lead generation, and measurement like this is what turns AI search from a buzzword into a line on a pipeline report. It is a small setup with a big effect on the conversations you can have with leadership.

If you want help setting up AI traffic tracking in GA4 or connecting it to your CRM so you can see what it is worth, reach out. Let's chat about what your AI traffic is really doing.