Google Merchant Center’s 6 conversational attributes explained

Product data feeds are becoming a key source of information for AI shopping tools. For ecommerce advertisers, a well-established feed with detailed product attributes can help products get discovered across AI-powered experiences like AI Mode and Gemini.

Google Merchant Center’s six conversational attributes build on existing product feeds by adding richer, machine-readable information, including customer questions and answers, product documentation, related products, and variant details. According to Google, these attributes are designed to make products easier to discover in nuanced searches and showcase complementary items.

What each conversational attribute does

1. Question and answer (question_and_answer)

The question_and_answer attribute lets you address specific customer queries and use cases for your product. This attribute is submitted in a natural-language conversational style, so consider using your existing FAQ section or product documentation to populate this category. Consider user, merchant, and manufacturer questions and answers when compiling this data.

To format this attribute, you must use the two subattributes, which are [question] and [answer]. Each of these subattributes has a 1,000-character limit and you can have up to 30 questions and answers per product.

2. Document link (document_link)

In this attribute, you can attach relevant product documentation as PDF files. Specific examples that work best for the document_link attribute include training manuals, user guides, assembly instructions, or packing inserts. Conversational AI uses this information to answer detailed user questions about the product.

To format this attribute, you need a PDF document URL that starts with http:// or https://. There is a 2,000-character limit for the URL and you can attach up to five documents.

3. Related product (related_product)

The related product category features products that are frequently bought together, accessories to the line-item product, substitutes, or alternatives. Keep in mind that a product can have multiple relationship types with other products.

There are three sub-attributes for related_product that must be used. These are [relationship_type], [identifier_type], and [identifier]. The relationship type can indicate whether it is part of a set, a required part, often bought with, a substitute, a different brand, or an accessory.

The identifier type indicates whether you are using GTIN or ID and the identifier subattribute is the GTIN or ID of the related product. You can have up to 30 related products per line item.

4. Item group title (item_group_title)

The item_group_title attribute allows you to showcase all the variants in your item group with a common title — essentially a name for the entire product family. You use this attribute with the item_group_id attribute.

The format of this attribute is a 1- to 150-character string, but make sure it is free of variant information. Use a relevant title that clearly describes your product group and submit the same variant-identifying attributes for each variant in the product group.

Google recommends several best practices for this attribute, including using a succinct title, putting important details first, using keywords that you want your product to be found for, and only using details that apply to the product group.

5. Variant option (variant_option)

The variant_option attribute provides structured details for specific product options (such as color, size, and material). This attribute will be used to showcase your product with its other available variants. Only certain products are eligible for this particular attribute and it is used in conjunction with [item_group_id] and [item_group_title].

There are two sub-attributes with the variant_option: name and value. Make sure to keep the name of each variant consistent across your product catalog.

6. Popularity rank (popularity_rank)

Lastly, the popularity_rank attribute submits a relative internal ranking number (0 to 100) indicating how popular a specific item is compared to the rest of your catalog. You can use this attribute to identify your top-selling products, with the higher number being the more popular product.

Formatting is a numerical text value between 0 and 100. Try to specify an accurate rating and do not add a percentage sign. Google recommends updating this number if and when your top-selling products change, based on recent sales.

If you’d like to read more about conversational attributes and their formatting, Google Merchant Center Help has an excellent article titled “How to use conversational attributes.”

Dig deeper: AI shopping starts with your product feed, not your product page

How to implement and track conversational attributes

Google recommends using a supplemental data feed to add conversational attributes without rewriting your existing product data. This also makes it easier to update these fields as needed.

To assess visibility across AI shopping experiences, eligible merchants can use the AI performance insights report in Google Merchant Center. It shows how products and brands perform in conversational shopping queries across AI Mode and AI Overviews, including share of voice and visibility across the discovery, evaluation, and ready-to-buy stages.

The report also highlights popular product attributes and search terms shoppers use in AI queries, helping you identify gaps in your product data and opportunities to improve visibility.

Preparing your product data for AI shopping agents

As AI agents move from guiding purchasing decisions toward shopping on consumers’ behalf, the exchange of product information between websites and AI systems will become increasingly important.

Technologies like Universal Commerce Protocol (UCP) are designed to help websites and AI agents communicate and exchange information. Think of UCP as a universal translator for AI shopping.

Meanwhile, adding conversational attributes through supplemental feeds can provide richer product information to support visibility across AI-powered shopping experiences. Keeping these attributes accurate and up to date can help your catalog remain relevant as conversational commerce evolves.

Dig deeper: How ecommerce brands actually get discovered in AI search