GA4 Custom Reports for AI Traffic: 6 Reports Every Marketing Team Needs

GA4's default acquisition reports were designed for a world where traffic came from Google, Facebook, and email. They were not designed for a world where ChatGPT sends 50,000 visitors to your site in a week after recommending your product in a trending thread.

If you're relying on the standard Traffic Acquisition report to understand your AI traffic, you're seeing at best 60% of the picture — and the 40% you're missing includes the most strategically important data: which pages AI recommends, how mobile AI visitors behave differently, which AI platforms drive revenue, and when traffic spikes correlate with AI mention moments you can actually act on.

This guide walks through six GA4 custom reports that together give your marketing team a complete, actionable view of AI-referred traffic. Each report includes exact dimensions, metrics, filters, and step-by-step build instructions. Read the foundational guide on how to track AI referral traffic in GA4 first if you're starting from zero.

https://cdn.sanity.io/images/en9d2pb2/production/bf46a3b0a15610f5f2e001fa0a4e3c009dcae17e-1200x630.png

Want this built for you automatically? The AI traffic decoder tracks all six dimensions in one dashboard without manual GA4 setup. But if you want to own this in GA4 yourself — read on.

---

Why Default GA4 Reports Miss the Nuances of AI Traffic

GA4's built-in Traffic Acquisition report gives you session source/medium as the primary dimension and lets you see sessions, conversions, and revenue per channel. That's sufficient for paid search and email. It's inadequate for AI traffic for four specific reasons:

1. No AI domain grouping. GA4 doesn't group chatgpt.com, perplexity.ai, gemini.google.com, claude.ai, and copilot.microsoft.com into a single "AI Referral" channel by default. You'd need to scroll through hundreds of referral sources to manually sum them.

2. Dark social blindspot. The default report can only show you sessions with referrer headers. Approximately 25–35% of AI-originated sessions arrive without referrer headers — ChatGPT's mobile app, Perplexity on iOS, and any AI platform accessed in private mode all strip referrers. These sessions appear as direct traffic and never appear in your AI source filter.

3. No landing page context. Default acquisition reports don't show you which pages AI is sending traffic to. This is arguably the most important dimension for AI traffic — it tells you which content AI models are citing, which is the foundation of any answer engine optimization strategy.

4. No funnel visibility. The default report shows last-click channel performance but doesn't let you trace AI-referred sessions through your conversion funnel. You can't see where AI visitors drop off, which is essential for fixing AI-specific landing page and UX issues.

The six reports below address each of these gaps.

---

The 6 Reports: Summary Overview

Report	Primary Use	Report Type in GA4	Build Time
1. AI Source Performance Dashboard	Cross-platform AI revenue benchmark	Detail Report (Library)	10 min
2. AI Traffic by Landing Page	Content citation intelligence	Detail Report (Library)	10 min
3. AI Traffic by Device	Mobile UX and rendering diagnosis	Detail Report (Library)	8 min
4. AI Conversion Funnel	Drop-off identification by AI source	Funnel Exploration (Explore)	15 min
5. AI Traffic Trend (Weekly Cohorts)	Spike detection and AEO ROI	Free Form Exploration (Explore)	12 min
6. AI vs All Channels Comparison	Channel benchmarking and budget decisions	Free Form Exploration (Explore)	12 min

---

Report 1: AI Source Performance Dashboard

What It Does

This is the foundation report — it gives you a side-by-side view of every major AI platform's traffic performance including sessions, engagement, conversions, and revenue. It replaces the manual process of filtering the Traffic Acquisition report one source at a time.

Report Configuration

Setting	Value
Report type	Detail Report (via Reports > Library)
Primary dimension	Session source / medium
Secondary dimension	Session default channel group
Metrics	Sessions, Engaged sessions, Engagement rate, Conversions, Revenue, AOV
Filter	Session source contains: chatgpt.com OR perplexity.ai OR gemini.google.com OR claude.ai OR copilot.microsoft.com OR you.com OR bing.com/chat
Date comparison	Previous period (week-over-week)
Sort	Revenue (descending)

Step-by-Step Build Instructions

In GA4, navigate to Reports > Library (bottom of the left sidebar)

Click + Create new report and select Create detail report

Select Traffic acquisition as the template

In the Dimensions panel on the right, confirm Session source / medium is primary

Add secondary dimension: Session default channel group

In the Metrics panel, add: Sessions, Engaged sessions, Engagement rate, Conversions, Purchase revenue, Average purchase revenue per user

Click Add filter at the top of the report → Dimension: Session source → Contains → type each AI domain separated by OR logic (use separate filter conditions connected by OR)

Click Save → Name it AI Source Performance Dashboard

Click Add to report collection and choose your main collection to pin it

What You'll See

Platform	Typical Sessions %	Typical CVR	Typical AOV vs. Site Avg
chatgpt.com / referral	55–65% of AI total	1.8–2.8%	+18–22%
perplexity.ai / referral	20–28% of AI total	2.2–3.2%	+20–28%
gemini.google.com / referral	5–10% of AI total	1.5–2.2%	+10–15%
claude.ai / referral	3–6% of AI total	2.0–3.0%	+15–22%
copilot.microsoft.com / referral	2–5% of AI total	1.6–2.4%	+12–18%

Perplexity typically outperforms ChatGPT on CVR because Perplexity users are in a more active research mode — they're looking at cited sources as part of the answer, not just following a link. ChatGPT drives more volume but at slightly lower intent per session.

---

Report 2: AI Traffic by Landing Page

What It Does

This report answers the question: which pages on your site are AI models recommending? It maps landing pages to AI sources, revealing your citation footprint across platforms. This is content intelligence — it tells you where your AI visibility is concentrated and where gaps exist.

Report Configuration

Setting	Value
Report type	Detail Report (via Reports > Library)
Primary dimension	Landing page + query string
Secondary dimension	Session source / medium
Metrics	Sessions, Engaged sessions, Conversions, Conversion rate, Revenue
Filter	Session source contains: chatgpt.com OR perplexity.ai OR gemini.google.com OR claude.ai
Sort	Sessions (descending)
Date range	Last 90 days (minimum for meaningful data)

Step-by-Step Build Instructions

Go to Reports > Library > + Create new report > Create detail report

Select Pages and screens as the template

Change the primary dimension to Landing page + query string

Add secondary dimension: Session source / medium

Add metrics: Sessions, Engaged sessions, Conversions, Session conversion rate, Purchase revenue

Add filter: Session source contains [all AI domains as above]

Save as AI Traffic by Landing Page

How to Read This Report

The pages with the highest AI-referred session counts are the pages AI models are citing most frequently. Cross-reference these with your brand visibility tracker data to confirm which AI platforms are citing which pages.

Pattern	Interpretation	Action
Homepage receives majority of AI traffic	AI citing brand generally, not specific content	Build more specific, citable content pages
Product pages receiving AI traffic with high CVR	AI models recommending specific products	Optimize these pages, expand with comparison content
Blog posts receiving AI traffic with low CVR	AI citing your content but users don't convert	Add strong CTAs and product links to cited blog posts
Category pages receiving AI traffic	AI recommends your category page as a destination	Create dedicated AI-query landing pages below category level

The landing page report is also how you identify content gaps. If your competitor's review page appears in Perplexity citations but yours doesn't, you now know which content to build.

---

Report 3: AI Traffic by Device

What It Does

Mobile and desktop AI traffic behaves very differently. Mobile AI users — especially those clicking from ChatGPT's iOS app — often have shorter session times, higher bounce rates on certain page types, and lower conversion rates that are frequently misattributed to "low AI traffic quality" when the real cause is a rendering or UX issue specific to mobile.

This report isolates AI traffic by device to diagnose mobile-specific problems.

Report Configuration

Setting	Value
Report type	Detail Report (via Reports > Library)
Primary dimension	Device category
Secondary dimension	Session source / medium
Metrics	Sessions, Engaged sessions, Engagement rate, Average session duration, Conversions, Conversion rate, Revenue
Filter	Session source contains: chatgpt.com OR perplexity.ai OR gemini.google.com OR claude.ai
Sort	Sessions (descending)
Comparison	Enable date comparison to previous period

Step-by-Step Build Instructions

Go to Reports > Library > + Create new report > Create detail report

Select Tech details as the template

Set primary dimension to Device category

Add secondary dimension: Session source / medium

Add metrics: Sessions, Engaged sessions, Engagement rate, Average session duration, Conversions, Session conversion rate, Purchase revenue

Add AI source filter (same as above)

Save as AI Traffic by Device

https://cdn.sanity.io/images/en9d2pb2/production/de333cabeb9c6e85dc37a7eb7723849d4b5c7748-1200x630.png

The Mobile AI Traffic Problem

A consistent finding across AI traffic datasets: mobile AI sessions frequently show average session durations under 5 seconds on sites that perform normally for other mobile traffic. This is almost never a traffic quality issue. The causes are:

Form and widget rendering delays. Mobile pages with quiz widgets, size selector components, or email capture popups that trigger on load create friction specific to cold AI-referred visits. The visitor hasn't pre-loaded a session context.

Pop-up timing conflicts. Exit-intent popups triggered on mobile scroll-up behavior fire almost immediately for AI-referred users who are evaluating the page quickly. A popup appearing within 3 seconds of arrival on mobile kills engagement.

Mismatched content format. AI-referred mobile visitors often arrive expecting specific information from a prompt. If the mobile page leads with a hero banner and navigation before reaching the content, bounce rates spike.

Device	Typical AI CVR	Flag If CVR Falls Below	Common Fix
Desktop	2.0–3.5%	1.5%	Landing page content mismatch
Mobile	1.2–2.2%	0.8%	Page speed, popup timing, form friction
Tablet	1.5–2.5%	1.0%	Layout breakpoint issues

If mobile AI CVR is more than 40% below desktop AI CVR, run a mobile UX audit on your top AI landing pages specifically. Use GA4's Session recording via FullStory or Hotjar filtered to AI source sessions on mobile for direct observation.

---

Report 4: AI Conversion Funnel

What It Does

This funnel exploration tracks AI-referred visitors from their landing page through your full conversion path: product view, add to cart, checkout initiation, and purchase. It identifies the exact step where AI traffic drops off disproportionately compared to other channels.

Report Configuration

Setting	Value
Report type	Funnel Exploration (Explore > Funnel exploration)
Segment	AI Referral (session source contains AI domains)
Comparison segment	Organic Search
Funnel steps	Step 1: Session start with AI source filter; Step 2: view_item event; Step 3: add_to_cart event; Step 4: begin_checkout event; Step 5: purchase event
Visualization	Standard funnel (not open funnel)
Breakdown dimension	Session source / medium

Step-by-Step Build Instructions

Go to Explore (compass icon in left sidebar) → click Funnel exploration

In the Segments panel, click + and create a new segment:

• Segment name: AI Referral Traffic
• Condition: Sessions → Session source contains → chatgpt.com (add OR conditions for each AI domain)
• Click Save
• Drag the AI Referral Traffic segment into the Segment comparisons area
• Add a second comparison segment: Organic Traffic (built-in segment)
• In the Steps panel, define your funnel:
• Step 1: Event = session_start (this captures all sessions; the segment filter handles AI source)
• Step 2: Event = view_item
• Step 3: Event = add_to_cart
• Step 4: Event = begin_checkout
• Step 5: Event = purchase
• Set the Funnel type to Closed funnel (users must complete steps in order)
• Add breakdown dimension: Session source / medium
• Set date range to last 60 days
• Save the exploration as AI Conversion Funnel

Interpreting Funnel Drop-Off

Drop-Off Point	Diagnosis	Fix
Session start → view_item (high drop-off)	Landing pages don't match AI visitor intent; or slow page load on mobile	Build dedicated landing pages per AI query cluster; improve Core Web Vitals
view_item → add_to_cart (high drop-off)	Product page lacks the information AI visitors came looking for	Add comparison tables, specific specs, and trust signals to product pages
add_to_cart → begin_checkout (high drop-off)	Cart experience friction; unexpected shipping costs	Show shipping costs earlier; simplify cart UI
begin_checkout → purchase (high drop-off)	Payment friction; too many form fields	Enable accelerated checkout (Shop Pay, Google Pay); reduce form fields

Benchmark: AI-referred traffic that has a product-specific landing page typically shows 15–25% higher view_item-to-add_to_cart conversion than AI traffic landing on a homepage or category page. The funnel will show this explicitly.

---

Report 5: AI Traffic Trend (Weekly Cohorts)

What It Does

This report plots AI sessions over time by week, broken down by AI platform. It reveals spike weeks that correlate with AI mention moments — a new ChatGPT conversation citing your brand going viral, a Perplexity featured snippet update, or a new AI platform gaining traction. Identifying these spikes and their causes is how you measure the ROI of your answer engine optimization efforts.

Report Configuration

Setting	Value
Report type	Free Form Exploration (Explore)
Rows	Week (date granularity: Week)
Columns	Session source / medium
Values	Sessions
Segment	AI Referral Traffic (custom segment from Report 4)
Date range	Last 26 weeks (6 months minimum)
Visualization	Line chart
Secondary metric	Conversions

Step-by-Step Build Instructions

• Go to Explore > Blank exploration
• Name it AI Traffic Weekly Trend
• In the Variables panel, add Dimensions:
• Week (under Time dimensions)
• Session source / medium
• Add Metrics: Sessions, Conversions, Purchase revenue
• Drag Week to the Rows section
• Drag Session source / medium to the Columns section
• Drag Sessions to the Values section
• Apply your AI Referral Traffic segment
• Set date range: last 26 weeks
• Change visualization to Line chart
• Save the exploration

How to Identify What Drove a Traffic Spike

When you see a week where AI sessions jump 40%+ above the rolling average, run this diagnostic process:

Step	Action	Tool
1	Identify the spike week and which AI platform drove it	Report 5 (this report)
2	Check which landing pages received the new AI traffic	Report 2 (AI Traffic by Landing Page)
3	Search for your brand mentions in that AI platform during that week	Asva AI prompt intelligence
4	Check if any content went live, earned backlinks, or was cited in media during that week	Internal content calendar cross-reference
5	Check Google Search Console for branded keyword spikes in the same week	GSC Performance report, filter Brand queries
6	Document the correlation: what you published or earned vs. traffic lift	AEO ROI log

The weekly cohort view also reveals seasonal patterns specific to AI traffic — often different from organic search seasonality because AI citation patterns depend on what users are asking about, not just what they're searching for.

---

Report 6: AI vs All Channels Comparison

What It Does

This is the benchmark report every marketing team needs for budget and strategy decisions. It places AI referral traffic side-by-side with Paid Search, Organic Search, Direct, and Email across the metrics that matter most: sessions, conversion rate, revenue, and AOV. It answers the question every CMO eventually asks: "Is AI traffic actually valuable, or is it just interesting?"

Report Configuration

Setting	Value
Report type	Free Form Exploration (Explore)
Rows	Session default channel group
Columns	Metrics
Values	Sessions, Engaged sessions, Engagement rate, Conversions, Session conversion rate, Purchase revenue, Average purchase revenue per user
Segments	AI Referral (custom), Paid Search (built-in), Organic Search (built-in), Direct (built-in)
Date range	Last 90 days
Sort	Purchase revenue (descending)

Step-by-Step Build Instructions

• Go to Explore > Blank exploration
• Name it AI vs All Channels Comparison
• Add the AI Referral Traffic custom segment
• Also apply GA4's built-in segments: Organic Traffic, Paid Traffic, Direct Traffic
• Add Dimension: Session default channel group
• Add Metrics: Sessions, Engaged sessions, Engagement rate, Conversions, Session conversion rate, Purchase revenue, Average purchase revenue per user (AOV)
• Drag Session default channel group to Rows
• Drag all metrics to Values
• Switch visualization to Table for clean comparison
• Sort by Purchase revenue descending
• Save the exploration

Benchmark Interpretation Table

Channel	Typical CVR Range	Typical AOV vs. Site Avg	Strategic Implication
AI Referral	1.5–3.2%	+15–25%	High intent, scalable via AEO
Paid Search (brand)	3.5–6.0%	+5–10%	High CVR but high CPC; AI competes on cost
Paid Search (non-brand)	1.0–2.5%	-5–10%	Volume but lower quality; AI often wins on AOV
Organic Search	1.5–2.5%	Baseline	Established benchmark
Direct	3.0–5.5%	+10–20%	High intent but often AI-influenced (dark social)
Email	2.5–4.5%	+5–15%	High CVR from warm audiences

If AI Referral CVR exceeds non-brand Paid Search CVR, you have a strong case for reallocating a portion of paid search budget into AEO content investment — content designed to earn AI citations rather than ad impressions. Use your AI visibility score as the leading indicator of that investment's performance.

---

How to Set Up Custom Alerts for Each Report

Custom Insights in GA4 let you set automated alerts that trigger when thresholds are crossed. Here are the recommended alerts for each of the six reports:

Alert Name	Linked Report	Condition	Threshold	Notification
AI Traffic Spike	Report 1	AI source sessions increased	+40% week-over-week	Email immediately
AI Traffic Drop	Report 1	AI source sessions decreased	-25% week-over-week	Email immediately
New AI Landing Page	Report 2	New landing page receives AI sessions	First session from AI source	Weekly digest
Mobile AI CVR Alert	Report 3	Mobile AI CVR drops	Below 0.8%	Email immediately
AI Funnel Drop Alert	Report 4	view_item-to-add_to_cart drops	Below 15% for AI segment	Email immediately
AI Revenue Milestone	Report 6	Weekly AI revenue exceeds	Custom target (set your own)	Email immediately

Alert Setup in GA4

• Go to Admin > Property > Custom Insights
• Click + Create
• Set evaluation frequency: Weekly (for trend-based alerts) or Daily (for anomaly alerts)
• Add conditions using the dimension/metric combinations from the relevant report
• Set the notification email to your full marketing team distribution list
• Click Create

---

How to Share Reports and Schedule Email Exports

Sharing Custom Reports

GA4 custom reports built in the Library are property-level — any user with at least Viewer access to the property can see them once published. To share:

• Reports: Go to Reports > Library, find your saved report, click the three-dot menu → Share with all users or copy the direct link for the report
• Explorations (Explore): Go to Explore, find your exploration, click the three-dot menu → Share → choose Anyone in this Google Analytics property — this makes the exploration visible to all property users

Scheduling Email Exports

GA4's native email scheduling is available for standard reports only (not Explorations). For Explorations, use Looker Studio:

• Connect Looker Studio to your GA4 property
• Recreate the relevant report using GA4 as the data source
• In Looker Studio, click Share > Schedule email delivery
• Set frequency (daily, weekly, monthly), recipients, and format (PDF)
• Reports arrive in team inboxes automatically on schedule

For teams using the AI traffic decoder, all six report views are available natively in the dashboard with scheduled Slack and email delivery built in — no Looker Studio required.

---

What to Do When AI CVR Is Below Organic

If your Report 6 shows AI Referral CVR consistently below Organic Search CVR, the cause is almost always one of three things:

1. Landing page mismatch. The most common cause. An AI model recommends your brand in response to a specific query — "best lightweight moisturizer for oily skin" — but your AI referral landing page is the homepage. Organic search visitors land on a more relevant page because Google matches query intent to URL. Fix: use Report 2 to identify your top AI landing pages and create dedicated versions aligned to the most common AI queries. Use prompt intelligence to discover the actual prompts driving your AI traffic.

2. Page speed on mobile. AI platforms generate a high proportion of mobile traffic. If your top AI landing pages load slowly on mobile (Core Web Vitals LCP above 4 seconds), AI visitors bounce before they can convert. Organic search performance already filters this out to some extent because Google's ranking factors reward fast pages. Fix: run PageSpeed Insights specifically on your top AI landing pages. Prioritize mobile LCP improvements.

3. Trust deficit on first exposure. AI-referred visitors have never heard of you before the AI recommendation in many cases. They arrived via a trusted intermediary (ChatGPT, Perplexity) but haven't yet formed trust in your brand directly. Organic search visitors often arrive with some prior brand awareness from earlier research. Fix: add trust signals specific to first-time visitors on AI landing pages — customer review counts, press mentions, clear return policy, and social proof above the fold.

What to Do When AI CVR Is Above Paid Search

If AI Referral CVR in Report 6 consistently exceeds non-brand Paid Search CVR, you have an exceptionally strong signal to scale your AEO investment:

Calculate the cost-per-acquisition difference. AI traffic has a near-zero marginal CPA once you have citations — your cost is content creation and brand visibility tracking, not per-click spend. If AI CVR matches paid search at 1/10th the CPA, the ROI math is compelling.

Identify which AI platforms drive the highest CVR. Use Report 1 to rank platforms by CVR. Allocate content and AEO efforts toward the platforms where you already convert well — doubling down on strength compounds faster than building from weakness.

Use AI traffic trend data to forecast AEO ROI. Report 5 shows you the week-over-week growth trajectory. Extrapolate 6-month AI session growth at current trajectory and multiply by observed CVR and AOV to build a business case for AEO budget in your next quarterly review. Book a demo to see how Asva AI builds this forecast automatically from your GA4 data.

Scale the content that's already earning citations. Report 2 tells you which pages AI models are recommending. Create more depth on those topics, build related content clusters, and earn additional citations on adjacent queries. Your AI visibility score will rise as citation density increases.

---

Putting It All Together: Your AI Traffic Analytics Stack

Here's how the six reports work together as a weekly and monthly analytics workflow:

Weekly (15 minutes):

• Check Report 1 for any significant week-over-week session or revenue changes per AI platform
• Check Report 5 for trend direction — is AI traffic growing, flat, or declining?
• Review any Custom Insight alerts that fired during the week

Monthly (45 minutes):

• Run Report 2 to identify new landing pages receiving AI traffic and pages losing AI traffic
• Review Report 3 for mobile vs. desktop CVR gaps; escalate UX issues if mobile AI CVR dropped
• Run Report 4 funnel analysis to check for funnel step drop-offs that worsened
• Pull Report 6 for channel comparison; update your channel benchmark table

Quarterly (2 hours):

• Full analysis of Report 5 trends over the quarter with AEO initiative correlation
• Report 6 quarterly comparison: is AI's share of total revenue growing?
• Funnel analysis to measure impact of landing page improvements made during the quarter
• Update AEO investment business case with observed ROI

The brands that build this infrastructure now will have 12–18 months of AI traffic data when this channel matures into one of the primary acquisition sources for most categories. That historical data — the cohort trends, funnel benchmarks, and platform CVR comparisons — becomes a competitive moat that late-movers can't replicate quickly.

For teams that want this entire system built and maintained automatically, the AI traffic decoder provides all six report views, automated AI citation monitoring, dark social estimation, and real-time alerts in one platform. Book a demo to see it live.