Skip to main content

AgentIQ Overview

The AgentIQ Overview is your command center for agentic performance. It tells you where your catalog stands, which products AI agents are surfacing and which they're skipping, and what to do about it.

Agent Traffic Analytics

This section shows how agentic traffic is performing across your entire catalog. Use it to understand the revenue contribution of AI shopping agents and track changes over time.

Metrics displayed:

  • Sessions — total visits attributed to AI agent sources

  • Attributed Orders — purchases originating from agentic sessions

  • Attributed Revenue — revenue tied to those orders

  • Conversion Rate — sessions that resulted in a completed order

  • Average Order Value — mean order size from agentic traffic

  • Best attributed products — a list of your best products performing

  • AI agents distribution — a breakdown of different ai agent channels

  • Breakdown by collections — whats your best collections in agentic channels

Catalog Health Tracker

The Catalog Health Tracker scores every product in your catalog on AI readiness. Products are graded on a scale from Bad to Excellent based on how completely and accurately their listing data aligns with what AI agents need to match, rank and recommend them.
​

Use this view to spot clusters of underperforming listings before they cost you recommendations.
​

Tracked Prompts Rate

Tracked Prompts Rate section on the AgentIQ Overview, showing the rate, the search bar and the tracked prompts table

Every prompt you use in an Agentic Product Listing (APL) is tracked here automatically. AgentIQ regularly searches the Shopify Catalog for each tracked prompt and shows how often your products come back. A prompt stays tracked even after a listing stops using it, so you keep its history.

Your Tracked Prompts Rate

This is the number at the top of the section. It shows the share of Shopify Catalog searches that return one of your products, across all your tracked prompts. Prompts used by more Product Listings count for more. Prompts that haven't been searched yet are left out, so they never pull your rate down.

Next to the rate you'll see:

  • Change over the last 7 days: shown as a percentage. A green up arrow means AI agents are finding your products more often than a week ago. An amber down arrow means less often.

  • Prompts and Product Listings: how many prompts you track and how many Product Listings use them.

  • Pending: how many prompts haven't had their first catalog search yet.

The tracked prompts table

The table shows 10 prompts per page, with page controls below it. Use the search bar to find any prompt across your whole list. Hover the ⓘ next to a column name for a quick explanation.

  • Prompt: a shopper query your APLs target. Click it to expand the row.

  • Rank rate: how often a Shopify Catalog search on this prompt returns one of your products, across every search we've run. A dash means the first search is still pending.

  • Used in products: how many of your products use this prompt in their APL. Hover to see which ones. A 0 means no listing uses it anymore, but it stays tracked.

  • Alignment: how well the products returned by your own store's search match this prompt, from 0 to 100. "n/a" means your store's search returned no products for it.

  • Competitors: other stores that came back in the latest catalog search. Hover to see their names.

See what actually came back

Expand a prompt to see its latest Shopify Catalog search: the date, the total number of results, your products with their position and price, and the competing stores that appeared, each with its best position and top products.

Actions

  • Run in Catalog Browser: opens the prompt in the Catalog Browser.

  • Re-run search: runs a new catalog search now. Use it after you change a listing.

  • Remove prompt: shows on hover, and only for prompts no listing uses. It stops tracking the prompt and deletes its rank-rate history.

  • Track a prompt via APL: you start tracking a prompt by using it in an APL. This button opens the product picker to create one.

Did this answer your question?