Deep Dive

Setting Up AI Source Attribution on Shopify: 4 Approaches Ranked

Four ways to wire up AI search attribution on Shopify — GA4 custom channels, Looker Studio, Triple Whale or Polar Analytics, and Inxy. Ranked by setup effort, accuracy, and cost with honest assessments of what each actually delivers.

Inxy Team · Updated May 20, 2026 · 12 min read

← Back to AI Search Attribution

There are four practical ways to track AI search revenue on Shopify. They vary dramatically in setup time, accuracy, and cost. Here’s an honest comparison.

Option 1: GA4 Custom Channel Groups (Free)

Setup time: 20–30 minutes | Accuracy: Low–medium | Cost: Free

Setup steps

  1. GA4 → Admin → Data Settings → Channel Groups
  2. Create new channel group
  3. Add rules for each AI engine:
Rule conditionChannel name
Session source contains chatgpt.comAI Search
Session source contains chat.openai.comAI Search
Session source contains perplexity.aiAI Search
Session source contains claude.aiAI Search
Session source contains gemini.google.comAI Search
Session source contains l.meta.aiAI Search
Session source contains you.comAI Search
  1. Save. GA4 will now classify matching sessions as “AI Search” in Traffic Acquisition reports.

What you get

Sessions from AI engines that pass referrer headers appear under “AI Search.” Revenue from those sessions is attributed correctly.

What you don’t get

The ~60% of ChatGPT sessions that strip referrers stay in Unassigned. Gemini sessions still show as Organic Search. No order-level attribution — you see estimated session revenue, not specific orders.

Verdict

Good enough for directional insight. Understates actual AI revenue by 2–3×. Best starting point while you evaluate more accurate options.

Option 2: Looker Studio Dashboard (Free, More Work)

Setup time: 2–4 hours | Accuracy: Medium | Cost: Free

What you build

Connect Shopify and GA4 to Looker Studio. Build a blended data source joining GA4 session data with Shopify order data. Layer in the AI channel custom grouping from Option 1.

The join problem

GA4 sessions have client IDs; Shopify orders have order IDs. They don’t share a reliable join key. Joining requires either:

  • Shopify’s GA4 integration + ga_session_id custom parameter (works when the pixel fires correctly, fails with ad blockers)
  • Accepting session-level attribution with approximate revenue (±30–50% margin of error)

Looker Studio dashboards look authoritative but the underlying accuracy problem from GA4 isn’t solved — only visualized differently.

Verdict

Better reporting interface than native GA4. Good if you already live in Looker Studio. Doesn’t improve data quality.

Option 3: Triple Whale, Polar Analytics, or Northbeam (Paid)

Setup time: 1–3 hours | Accuracy: Medium (ads) / Low (AI) | Cost: $100–$500/month

What these tools do well

Multi-touch attribution for paid channels — Meta, Google Ads, TikTok. Order-level revenue with actual Shopify order IDs. This is their core value.

What they don’t do

None of the major third-party attribution platforms have dedicated AI search channels as of May 2026. ChatGPT sessions arrive as Unassigned or Direct in Triple Whale just as they do in GA4. You still need to add custom channel rules — the same workaround, just in a different interface.

Verdict

Worth investing in if you have significant ad spend and need proper multi-touch attribution for paid channels. AI search attribution remains an unsolved problem on these platforms — it’s not their focus.

Option 4: Inxy (Dedicated AI Attribution, 5 Minutes)

Setup time: 5 minutes | Accuracy: High (85–90%) | Cost: Included in Inxy subscription

Setup

Connect Shopify to Inxy via OAuth. Done. No pixel installation, no Looker Studio setup, no custom channel rules to maintain.

How the classifier works

Inxy reads four signals at order creation:

  1. Shopify referring_site — recorded by Shopify at order creation, independent of GA4. Survives ad blockers and browser privacy settings that break GA4’s pixel.
  2. Shopify landing_site — the first page visited, cross-referenced against your AI-optimized content inventory.
  3. UTM parameters — when present in the order URL, used directly and take priority.
  4. Order timing pattern — tiebreaker for sessions where all other signals are absent.

What you see

This month's AI search revenue:
  ChatGPT     $423   7 orders
  Perplexity  $211   4 orders
  Gemini      $156   3 orders
  Claude       $89   2 orders
  ─────────────────────────
  Total        $879  16 orders

Order-level breakdown with Shopify order IDs. Export to CSV. Week-over-week trends. Attribution by product and collection.

Why Shopify order fields beat GA4 sessions

GA4 captures sessions via JavaScript pixel. If a user has an ad blocker, privacy extension, or Safari’s ITP, the pixel may not fire — and the session is lost from GA4’s data. Shopify’s referring_site and landing_site fields are recorded server-side at order creation. They’re not affected by client-side blocking.

In practice, Inxy captures 15–25% more revenue than GA4-based approaches simply because the data source is more reliable.

Verdict

The only option purpose-built for AI search attribution on Shopify. Recommended for any store where AI search attribution is strategic (typically stores with 3+ months of published AEO content and $15K+ monthly revenue).

Decision Framework

SituationRecommended path
Just starting, want quick visibilityOption 1 (GA4 custom channels)
Already in Looker Studio for everythingOption 1 + Option 2
Significant ad spend, need multi-touchOption 3 + Option 1 for AI
AI traffic is strategic, want real revenue numbersOption 4 (Inxy)

Start with Option 1 today — it’s free and takes 30 minutes. Whether the approximate numbers are sufficient depends on how much you’re investing in AEO and how much AI-source revenue you suspect you’re leaving unattributed.

Next: AI Search UTM Parameters Reference — every UTM pattern AI engines use, updated monthly.

← Back to AI Search Attribution