Creating GA4 Segments for AI Traffic Analysis (Step-by-Step)

Your AI Chatbots channel in GA4 shows 3,200 sessions last month with an average engagement time of 4 minutes and 22 seconds.

But that number hides critical differences:

Claude visitors who spent 14 minutes and converted to email at 8 percent are averaged with Perplexity visitors who spent 90 seconds and bounced. First-time ChatGPT visitors discovering your brand are mixed with returning Claude users who've visited six times.

Aggregate AI traffic metrics obscure the patterns that matter.

You need segments—slices of your AI traffic isolated by specific characteristics—to understand behavior, optimize conversion paths, and measure true value.

Here's how to build the six most valuable AI traffic segments in GA4.

Segment 1: AI First-Time Visitors

This segment isolates users discovering your brand through AI for the first time.

Why it matters: First-time visitors have different behavior and conversion potential than returning visitors. Tracking them separately reveals whether AI is an effective discovery channel.

How to create:

Navigate to Explore → Blank exploration

Click the "+" next to Segments

Choose "Create a custom segment"

Segment conditions:

Segment type: User segment

Add condition group:

• First user source / medium → matches regex → your AI tracking pattern
• Session number → equals → 1

Pattern to use:

(chatgpt|openai|anthropic|deepseek|grok)\\.com|(gemini|bard)\\.google\\.com|(perplexity|claude)\\.ai|copilot\\.microsoft\\.com

Name: "AI First-Time Visitors"

Save and apply to your exploration.

What to analyze with this segment:

• Landing page distribution (which content drives discovery)
• Engagement rate (do first-timers engage or bounce)
• Email capture rate (top-funnel conversion)
• Return visitor rate within 30 days (retention signal)

Compare this segment to "All AI Traffic" to see if newcomers behave differently than your established AI audience.

Segment 2: AI Returning Visitors

Users who discovered via AI and came back—these are your highest-value AI-influenced visitors.

Why it matters: Returning visitors signal content satisfaction and brand recall. They're more likely to convert because they've moved past initial awareness into consideration.

How to create:

Explore → Blank → Create custom segment

Segment conditions:

Segment type: User segment

Add condition group:

• First user source / medium → matches regex → AI pattern
• Session number → greater than → 1

Name: "AI Returning Visitors"

What to analyze:

• Conversion rate comparison (returning vs first-time)
• Content consumption patterns (what brings them back)
• Days between first and second visit (engagement timeline)
• Device switching (mobile discovery, desktop conversion)

If returning AI visitors convert at 5x the rate of first-time visitors, that validates AI as a top-funnel channel that pays off over time.

Segment 3: Platform-Specific Segments (ChatGPT, Claude, Perplexity)

Isolate individual AI platforms to compare performance.

Why it matters: ChatGPT visitors behave differently than Claude visitors. Platform-specific segments reveal which AI sources deserve optimization priority.

How to create (example for ChatGPT):

Explore → Create custom segment

Segment conditions:

Segment type: Session segment

Add condition:

• Session source / medium → matches regex → (chatgpt|openai)\\.com

Name: "ChatGPT Traffic"

Repeat for each platform:

Claude pattern: (claude\.ai|anthropic\.com) Perplexity pattern: perplexity\\.ai Gemini pattern: (gemini|bard)\\.google\\.com Copilot pattern: copilot\\.microsoft\\.com

Comparison analysis:

Apply all platform segments simultaneously to one exploration.

Compare metrics:

• Engagement time (which platform sends most engaged visitors)
• Pages per session (depth of interaction)
• Conversion rates (which platform drives action)
• Device breakdown (mobile vs desktop patterns)

This reveals whether you should optimize content differently for ChatGPT's broad audience vs Claude's technical users.

Segment 4: High-Engagement AI Visitors

AI visitors who showed strong engagement signals but didn't convert yet—prime retargeting candidates.

Why it matters: These users demonstrated interest (long session, multiple pages) but need nurturing. Segment them for retargeting campaigns.

How to create:

Explore → Create custom segment

Segment conditions:

Segment type: User segment

Add condition group 1:

• Session source / medium → matches regex → AI pattern

Add condition group 2 (any of these):

• Session engaged → equals → true
• Engagement time → greater than → 180 (3 minutes in seconds)
• Page views → greater than → 3

Add exclusion condition:

• Conversion event → occurred → (your key conversion event)

Name: "High-Engagement AI Non-Converters"

How to use:

Export this segment as a GA4 Audience (Configure → Audiences → New audience → Import from segment)

Once it's an audience, you can:

• Use it for Google Ads retargeting
• Build lookalike audiences
• Track conversion rate over 90 days
• Create nurture email campaigns (if integrated with your email platform)

These visitors are warm leads—they engaged deeply but didn't convert in first session. Perfect for retargeting.

Segment 5: AI Mobile vs Desktop

Separate mobile and desktop AI traffic to understand device-specific behavior.

Why it matters: Mobile app AI citations often don't pass referrer data, appearing as Direct traffic. Comparing mobile AI traffic (when trackable) to desktop reveals device preferences and potential attribution issues.

How to create:

Explore → Create two segments

Segment A - AI Mobile:

Segment type: Session segment

Conditions:

• Session source / medium → matches regex → AI pattern
• Device category → exactly matches → mobile

Segment B - AI Desktop:

Segment type: Session segment

Conditions:

• Session source / medium → matches regex → AI pattern
• Device category → exactly matches → desktop

Analysis to run:

Compare conversion rates, engagement time, and pages per session between mobile and desktop.

If mobile AI traffic shows:

• Much lower volume than expected → mobile attribution issue (referrers stripped)
• Higher bounce rate → mobile landing page optimization needed
• Lower conversion rate → mobile checkout friction

This helps diagnose technical tracking issues vs genuine behavioral differences.

Segment 6: AI Traffic by Content Category

Segment AI visitors by which content category they landed on.

Why it matters: AI platforms may cite your technical documentation heavily but ignore your marketing blog. Segmenting by landing page category reveals content strengths and gaps.

How to create:

Explore → Create custom segment

Segment conditions:

Segment type: Session segment

Condition group:

• Session source / medium → matches regex → AI pattern
• Landing page → contains → /blog/ (or your content category path)

Name: "AI Blog Visitors"

Repeat for each content category:

• AI Docs Visitors: Landing page → contains → /docs/
• AI Guides Visitors: Landing page → contains → /guides/
• AI Product Visitors: Landing page → contains → /product/

Comparison insights:

If AI blog traffic converts to email at 6 percent while AI docs traffic converts at 12 percent, your documentation is more valuable for top-funnel capture than blog content.

Shift content strategy accordingly—expand docs, improve blog quality, or adjust CTAs by section.

Building a Complete AI Analysis Exploration

Combine multiple segments into one comprehensive analysis dashboard.

Step 1: Navigate to Explore → Blank exploration

Step 2: Add all your segments:

• AI First-Time Visitors
• AI Returning Visitors
• ChatGPT Traffic
• Claude Traffic
• Perplexity Traffic
• High-Engagement AI Non-Converters

Step 3: Add dimensions:

• Landing page
• Device category
• Country

Step 4: Add metrics:

• Sessions
• Engaged sessions
• Average engagement time
• Conversions
• Conversion rate
• Pages per session

Step 5: Build comparison tables:

Tab 1: Platform comparison (ChatGPT vs Claude vs Perplexity) Tab 2: New vs returning AI visitors Tab 3: Landing page performance by AI segment Tab 4: Device analysis

Save this exploration as "AI Traffic Deep Dive" and review monthly.

This single exploration replaces hours of manual filtering and provides consistent month-over-month comparisons.

Using Segments in Standard Reports

Segments aren't just for Explorations—apply them to standard GA4 reports too.

In any standard report:

Click "Add comparison" at the top of the report

Select your saved segments from the dropdown

Useful applications:

Acquisition reports: Compare AI traffic to Organic Search and Paid Search using segments

Engagement reports: Filter Pages and Screens report to show only AI traffic landing pages

Monetization reports: Compare revenue attribution across traffic source segments

Comparisons persist as you navigate between reports, making it easy to analyze AI traffic across the entire GA4 interface.

FAQ

What's the difference between segments and audiences?

Segments are analysis tools used in Explorations and reports. Audiences are user groups you can activate for marketing—retargeting, email campaigns, or exclusions. Create segments for analysis, export them as audiences for activation.

How many segments can I create?

No practical limit in Explorations. However, only 10 audiences can be active per property (standard GA4), so prioritize which segments to export as audiences carefully.

Do segments apply to historical data?

Yes. Segments analyze all historical data that meets the conditions. This is different from audiences, which only include users going forward from creation date.

Can I share segments with team members?

Yes. When you save a segment in an Exploration, it's available to anyone with access to that GA4 property. Save meaningful segments with clear names so colleagues can reuse them.

Should I create segments for every AI platform?

Only if you have sufficient traffic volume (100+ sessions monthly per platform). Below that threshold, noise overwhelms signal and segment-specific insights aren't reliable.

How often should I review segment performance?

Monthly for tactical insights (which content performs well, engagement patterns). Quarterly for strategic decisions (budget allocation, optimization priorities, platform focus).

Related guides:

• Measuring the Real Value of AI Chatbot Traffic in GA4
• Track All AI Chatbot Traffic in GA4 with One Filter
• Using AI Traffic Data to Guide Your Content Strategy