What's new on Shopify's Agentic Dashboard (and what it shows)
products listed might not show up in AI channels
I check Shopify's Agentic Dashboard on a handful of stores every week, mostly just watching how the ranking numbers move.
So when two new filters showed up last week, it felt worth sharing.
1/ Country selector on catalog search
2/ Category dropdown on relevant searches
What actually changed
In the dashboard, test searches against your catalog run the same way an AI agent would and show where your products land.
Earlier, catalog search returned a blended ranking number - no matter which market a shopper was searching from.
Now there is a country selector that splits it.
Pick a market and you can check how your catalog ranks specifically in that instead of the average across all countries you ship to.
Next, the category dropdown does something similar for the two search lists that were already there: recent searches where your products appeared and relevant searches in your category where you should be coming up.
Set it to a specific category like Apparel and Accessories, and both lists narrow to that category alone - instead of mixing in queries from a store's entire catalog.
Shopify Agentic Dashboard, category and country filtering
Both turn one averaged-out number into something a merchant can actually use.
The part underneath both the new filters
Neither filter fixes anything by itself - they just narrow the view.
What decides whether there's anything worth seeing?
At DotDev, Ravi Theja Yada (Shopify PM on Catalog API) called it a product graph.
A product graph is Shopify's own map of your catalog data. It's what an AI agent actually reads before it shows a shopper anything.
Yada splits that graph work into two separate requirements.
1/ Supply means your products are in the graph at all.
2/ Identity means the graph can tell your product apart from everyone else's.
If a product's identity isn't resolved, it can look like a near-duplicate of something else. The system may just skip it, no matter how you filter the dashboard.
Where to start
1/ Filter by your biggest market first. A blended rank that looked fine before can hide a market where you're barely showing up at all.
2/ Pull up relevant searches in your top category, then compare it against recent searches where you actually appeared. That gap is usually where an identity problem shows up, before it ever becomes a ranking problem.
3/ Check whether your best sellers show up in a specific category, not a parent node one or two levels up. That's the identity question the dashboard can't answer for you.
Most of what shows up in either list comes down to the same thing: whether a product's category and attributes are specific enough for Shopify's graph.
This is what Atomz's catalog enrichment does for a Shopify store.
It assigns the leaf-level category and fills in what's underneath it, the same structured data that agentic channels like ChatGPT and Copilot end up reading from.
It checks the same category and attribute fields that decide whether your products resolve as their own identity in Shopify's graph, or blend into everyone else's.
- Ankit
If this was useful, the next one will be too.
Weekly. Free.
Unsubscribe with one click.
