- Step-by-step guide to building a GA4 dashboard that tracks AI referral traffic separately from organic and paid
- Monitors traffic from ChatGPT, Perplexity, Gemini, Claude, and Copilot in one view
- Covers custom channel grouping, calculated fields, and conversion tracking for AI visitors
- Takes about 20 minutes to set up in Looker Studio with your own GA4 property
GA4 Wasn’t Built for AI Traffic. Here’s How to Fix That.
Open your GA4 right now and try to find how much traffic you’re getting from AI search engines. Go ahead, I’ll wait.
You probably can’t find it easily. That’s because GA4 lumps AI referral traffic into the generic “Referral” or “Organic Search” bucket depending on how the traffic arrives. ChatGPT clicks might show as referrals. Google AI Overview traffic gets mixed into organic. It’s a mess.
I built this dashboard setup because I was tired of manually filtering referral data every time a client asked, “How much traffic are we getting from AI?” Now you can see it in one place, updated in real time.
The 527% growth stat comes from BrightEdge’s year-over-year AI search traffic study. If your dashboard doesn’t separate AI traffic, you’re missing the fastest-growing channel in search.
What This Dashboard Tracks
The dashboard has five main sections. Here’s what each one shows:
AI Traffic Overview
Total sessions from all AI platforms combined. Trend line showing growth over time. Comparison to previous period.
Platform Breakdown
Sessions split by platform: ChatGPT, Perplexity, Gemini, Claude, Copilot, and “Other AI.” See which platforms drive the most traffic.
Top Landing Pages
Which pages are AI engines sending traffic to? This shows you what content gets cited most often.
Engagement Metrics
How AI traffic compares to organic and paid on engagement rate, time on page, bounce rate, and pages per session.
Conversion Tracking
AI traffic conversion rate vs. other channels. According to First Page Sage, AI search converts at 4.4x the rate of organic. This dashboard lets you verify that for your own site.
Step 1: Set Up Custom Channel Grouping in GA4
Before building the dashboard, you need GA4 to recognize AI traffic as its own channel. Here’s how:
Go to GA4 Admin
Open your GA4 property. Click Admin (gear icon) > Data display > Channel groups.
Create New Channel Group
Click “Create new channel group.” Name it “Custom with AI Search.”
Add AI Search Channel
Click “Add new channel.” Name it “AI Search.” Set the conditions to: Source matches regex: chatgpt\.com|chat\.openai\.com|perplexity\.ai|gemini\.google\.com|copilot\.microsoft\.com|claude\.ai|you\.com|phind\.com
Copy Default Channels
Copy all existing default channel definitions so you keep Organic Search, Paid, Direct, etc. alongside your new AI Search channel.
Important: The channel group only applies to new data going forward. You won’t see historical AI traffic in this channel. But you can still filter historical referral data in the dashboard using the source dimension.
Step 2: Build the Looker Studio Dashboard
Open Looker Studio (it’s free with any Google account) and connect your GA4 property as a data source.
Page 1: AI Traffic Overview
Create these components:
Overview Page Components
Metric: Sessions. Filter: Source matches the AI regex pattern. Add comparison to previous period.
Calculated field: AI Sessions / Total Sessions * 100. This number will likely be 1-5% for most sites in early 2026, but growing fast.
Dimension: Date. Metric: Sessions. Filter: AI sources only. Set to daily granularity with a 90-day default range.
Dimension: Session source. Metric: Sessions. Filter: AI sources only. Shows which AI platform drives the most traffic.
Page 2: Platform Deep Dive
Platform Breakdown Components
Horizontal bar chart comparing ChatGPT, Perplexity, Gemini, Claude, and Copilot sessions side by side.
Rows: Each AI platform. Columns: Sessions, Engagement Rate, Avg Session Duration, Conversions. This shows you which platform sends the best traffic, not just the most.
Separate time series for each AI platform. Useful for spotting when a specific platform starts (or stops) sending traffic.
Page 3: Content Performance
Content Page Components
Dimension: Landing page. Metrics: Sessions, Engagement Rate, Conversions. Filter: AI sources. Sorted by sessions descending.
Shows your top 20 pages with columns for AI Sessions, Organic Sessions, and Paid Sessions side by side. Helps you see which content works for AI vs. traditional search.
Page 4: Conversions
Conversion Components
Conversion rate for AI traffic specifically. Compare to site-wide conversion rate.
Compare AI Search vs Organic vs Paid vs Direct conversion rates. This is the chart that usually gets executives excited.
Which AI platform drives the most conversions? Not just traffic. This helps you prioritize your AI search improvement efforts.
Step 3: Create Calculated Fields
You’ll need a few custom fields to make the dashboard work properly:
| Field Name | Formula | Purpose |
|---|---|---|
| Is AI Traffic | REGEXP_MATCH(Session source, “chatgpt|openai|perplexity|gemini\.google|copilot\.microsoft|claude\.ai”) | Boolean filter for AI traffic |
| AI Platform Name | CASE WHEN REGEXP_MATCH(Source, “chatgpt|openai”) THEN “ChatGPT” WHEN … END | Clean platform labels |
| AI Traffic Percentage | SUM(AI Sessions) / SUM(Total Sessions) * 100 | Share of total traffic |
What the Data Usually Shows
I’ve set up this dashboard for about 40 sites now. Here are the typical patterns:
Perplexity usually leads on referral volume. It links to sources every single time, so click-through rates are higher. ChatGPT traffic is growing faster, but many ChatGPT interactions don’t generate clicks because users get their answer in the chat.
AI traffic has higher engagement. On average, I see 15-25% higher engagement rates from AI referrals compared to organic. My theory: people who click through from an AI citation are more intentional. The AI already told them the basics; they’re clicking because they want more depth.
Conversion rates are genuinely higher. The 4.4x number from First Page Sage holds up in my experience, especially for B2B sites. AI traffic tends to convert better because these visitors already understand the topic. They’re further down the funnel.
The absolute numbers are still small. For most sites, AI traffic is 1-5% of total traffic in early 2026. But it’s growing month over month. The sites that are tracking it now will be ready when it hits 10-20%.
For the latest market data on AI search growth, check our AI search market share report.
Connecting to Your Broader AI Strategy
This dashboard tells you what’s happening with AI traffic. But to improve the numbers, you need to work on the why.
- Low AI traffic overall? Start with the AI search improvement fundamentals
- Traffic from Perplexity but not ChatGPT? Review how ChatGPT selects sources
- High traffic but low conversions? Your landing pages might need better CTAs for AI-referred visitors
- Want to check if you’re being cited at all? Use our free AI citation checker
Keeping the Dashboard Updated
AI platforms change their referral patterns. New AI search engines launch. Here’s how to keep your dashboard current:
Quarterly Maintenance
Look at your full referral list in GA4. Are there new AI-related domains showing up that aren’t in your filter?
Add any new AI platform domains to your channel grouping and dashboard filters.
Make sure your GA4 conversion events are still firing correctly. Dashboard data is only as good as the tracking.
Send the dashboard link to your team or clients. Looker Studio sharing is free and makes you look good.
Troubleshooting Common Issues
No AI traffic showing up at all?
Three possible reasons: (1) You genuinely aren’t getting AI referral traffic yet, which means you need to work on your AI search improvement checklist. (2) Your regex pattern has a typo. Double-check the source names. (3) Traffic is coming in but categorized as “direct” because the referrer header is stripped. This happens sometimes with ChatGPT’s mobile app.
Numbers seem too low?
Remember that not all AI traffic comes with a clean referrer. Some ChatGPT and Gemini traffic gets categorized as direct or organic. The dashboard captures identifiable AI traffic, which is a floor, not a ceiling.
Gemini traffic hard to separate?
Yes. Google AI Overviews send traffic that looks identical to regular Google organic traffic in GA4. There’s no reliable way to separate them as of March 2026. This is a known limitation. I track it as a combined “Google AI” signal and note the caveat.
Want a Custom AI Traffic Dashboard for Your Site?
We set up tracking, build custom dashboards, and help you grow AI referral traffic. Start with a free AI visibility audit to see where you stand.
Frequently Asked Questions
Yes. Looker Studio (formerly Google Data Studio) is completely free. You need a Google account and access to a GA4 property. The template itself just requires you to set up the components described in this guide. No paid tools needed.
Not reliably as of March 2026. Google AI Overview clicks show up as regular organic traffic in GA4 because they come through google.com. There’s no separate referrer for AI Overviews. This is an industry-wide limitation that Google may address in the future.
For most sites in early 2026, AI referral traffic is 1-5% of total traffic. Sites with strong citation presence can see higher percentages. The key metric is growth rate, not absolute numbers. If your AI traffic is growing month over month, you’re on the right track.
Yes. Each GA4 property needs its own Looker Studio dashboard connected to its specific data source. But once you build the template for the first client, you can copy the dashboard and just swap the data source for subsequent clients. Takes about 5 minutes per client after the first one.
