AI Shopping Analytics

Zevin’s analytics dashboard gives you a unified view of how AI shopping agents interact with your catalogue — from the first query to a completed checkout.

Why AI analytics is different

Traditional web analytics tools (Google Analytics, Mixpanel) track humans navigating pages. AI agents behave differently:

• They query your catalogue via API, not by loading pages
• A single agent session may generate dozens of product queries before the human sees a single result
• Attribution happens at the recommendation layer, not the click layer

Zevin captures all of this — giving you data that Google Analytics can never see.

Dashboard overview

The main dashboard shows four core metrics at a glance:

Metric	What it measures
AI queries	Number of times an AI platform queried your catalogue
Product impressions	Number of times a product appeared in an AI-generated response
AI-referred sessions	Visits to your store that originated from an AI recommendation
AI-referred revenue	Revenue from orders placed by AI-referred shoppers

All metrics can be filtered by date range, AI platform (ChatGPT, Gemini, Claude, Perplexity, etc.), and product category.

Query analytics

The Queries tab shows the natural language queries that AI agents ran against your catalogue:

• Top queries — the product questions asked most often
• Zero-result queries — queries where your catalogue returned no matches (gaps to fill)
• Partial-match queries — queries where results were returned but had low relevance scores

Zero-result and partial-match queries are the most actionable data in the dashboard. They reveal exactly what shoppers are looking for that you don’t currently stock or don’t describe clearly enough.

Product performance

The Products tab ranks your catalogue by AI performance:

• AI impressions — how often each product appears in AI responses
• AI CTR — click-through rate from AI recommendation to your product page
• AI conversion rate — percentage of AI-referred product page views that result in a purchase
• AI revenue — total revenue attributed to AI-referred sessions for each product

Products with high impressions but low CTR usually need better images or titles. Products with high CTR but low conversion rate point to a landing page or checkout issue.

Platform breakdown

The Platforms tab shows the same metrics segmented by AI platform. This lets you see:

• Which platforms drive the most traffic for your product category
• Where to focus optimisation effort
• How quickly each platform picks up catalogue changes after a sync

Attribution model

Zevin uses a first-touch AI attribution model. If a shopper is first referred to your store by an AI platform and completes a purchase within 30 days, that order is counted as AI-attributed revenue — even if the shopper visited multiple times or used other channels before converting.

The 30-day window and the attribution model can be adjusted in Settings → Analytics.

Connecting to your existing analytics

Zevin can push events to:

• Google Analytics 4 — sends a zevin_referral event with platform and product data
• Segment — full event stream for all AI queries and referrals
• Webhooks — POST events to any URL for custom analytics pipelines

Configure these integrations in Settings → Analytics → Destinations.

Data retention

Analytics data is retained for 13 months on all paid plans. The Free plan retains 30 days of data.