The “All products” page in your Merchant Center is your central hub for managing your product catalog. This page features improved AI-powered filtering to help you find, filter and manage your products more efficiently using everyday, natural language.
On this page
- How to use improved product filters
- Example questions to ask the AI
- Refining your search results with AI
- Which product attributes are supported?
How to use improved product filters
On the “All products” page, a single intelligent search bar handles both natural language queries and traditional filtering.
- To search with AI, simply type your request in natural language into the search bar. The system’s AI will translate your words into the appropriate filters to show you the exact products you’re looking for.
- The filters appear as chips which you can then manually add, remove, or modify to further adjust your results.
Example questions to ask the AI
You can ask to filter the products by a wide range of product attributes, performance metrics, and statuses. This saves you the time and effort of manually applying filter conditions. Here are some examples of what you can ask the AI:
By performance:
-
Find visible products that have zero clicks -
Show top-clicked products in “Free Listings” -
Which products cost me clicks but are out of stock?
By status and attributes:
-
Show me products from “Nike” that aren’t approved -
Manually edited approved products -
Show products with manual edits that have item issues
Example questions that don't work
The system doesn’t understand personalized queries yet. Things like
-
Products with a lot of clicks -
Show me 10% of my products -
Products that amount to 80% of my traffic
The system operates as a natural language query interpreter allowing for informal language in your requests. However, it’s important to understand that this flexibility can sometimes lead to loss of precision. For example, when users frequently employ “and” in their spoken language, they may implicitly mean “or”. This ambiguity makes it challenging to create a precise, formal filter set. Queries like:
-
Marketing method free local listings and shopping ads
Might be misinterpreted by the LLM and need to be refined.
Refining your search results with AI
Our AI-powered improved product filters are aware of the filters you already have active. If you issue a natural language query while other filters are applied, the system will attempt to refine your current results based on the new input.
For example, after searching for Show top-clicked products in “Free Listings”
, you can then type also limit products with more clicks than 100
to further narrow down the results without starting over.
Which product attributes are supported?
The AI can generate filters based on many of your product’s attributes. Here are the fields it currently understands:
| Attribute name | Technical name |
| Title | title
|
| Product ID | product_id
|
| All Clicks | all_clicks
|
| Has UI Edit | has_ui_edit
|
| Target Countries | intended_countries
|
| Brand | brand
|
| Product Type | product_type_full
|
| Feed Label | feed_label
|
| availability | availability
|
| Product Status | calculated_status
|
| Visibility (Is Showing) | is_showing
|
| Price | price
|
| Custom Labels (Tags) | tags
|
| Data Source |
primary_source_info
|
| MPN |
mpn
|
| Has Item Issues | has_item_issues
|
| Reporting Destinations | intended_reporting_contexts
|
| Item Group ID | item_group_id
|
