How to Optimise Your Product Feeds for
Better Results in AI Search
AI-powered search is rapidly changing how customers discover products online. From Google AI Overviews and Gemini to emerging AI shopping assistants, product feeds are becoming a critical source of information that helps AI systems understand, recommend and compare products.
For ecommerce businesses, simply providing basic product feed data is no longer enough. To maximise visibility in AI search results, merchants need to enrich their product feeds with detailed, structured information that helps AI answer customer questions and make better recommendations.
At Dream Agility, we’re already helping brands prepare for the future of AI commerce by optimising product feeds for Google Merchant Center, Shopping campaigns, Performance Max and AI-driven search experiences. We were first in 2013 when Google Shopping launched to build our award winning “paid search first” feed tech instead of trying to reuse Market Place feeds which were completely unsuitable and expensive to use in a paid environment. We’re first again to get all our clients, AI Search ready.
Why Product Feed Enrichment Matters for AI Search
Traditional search engines relied heavily on keywords. AI search works differently.
AI systems use product feed data to answer questions such as:
- Which product is best for my needs?
- What products are compatible with this item?
- What are the key features and benefits?
- How does this compare to alternatives?
The richer your product feeds, the more likely AI systems are to understand and recommend your products.
8 Product Feed Attributes to Support AI Search
Customers increasingly search by asking detailed questions: “Which running shoes come in a wide fit?” or “Does this appliance need professional installation?”
Your product data needs to provide the answers. Google’s six conversational attributes, alongside product highlights and technical details, give you more ways to explain what your products offer and who they’re suitable for.
1. Product Highlight (product_highlight)
Use short, clear bullet points to explain the product’s main benefits and features. Google recommends four to six highlights per product.
Focus on useful information, such as “Removable, dishwasher-safe parts” or “Up to 10 hours of battery life”, rather than vague claims like “Outstanding quality”.
2. Product Detail (product_detail)
Use this field for structured technical specifications, with an attribute name and value, plus an optional section name.
For example: Battery life:10 hours.
This helps provide precise information for shoppers comparing products against specific requirements. Keep the details factual and avoid repeating information already supplied in other attributes.
3. Question and Answer (question_and_answer)
Add product-specific questions and clear, accurate answers that help customers make a buying decision.
Think about the questions your sales or customer service team regularly receives, such as “Can this pump supply two showers at once?” Include useful information that isn’t already covered elsewhere in your submitted product data or supporting documents.
4. Variant Option (variant_option)
Describe the properties that distinguish one version of a product from another, such as shoe width, fit or storage capacity.
Used alongside item_group_id and item_group_title, this helps Google understand the available options when a customer asks for a particular combination of features.
5. Item Group Title (item_group_title)
Provide a shared title for a product family, with individual sizes, colours or configurations grouped underneath it.
For example, “Men’s Lightweight Trail Running Shoes” could be the shared title across several colour and size variants. Use it alongside item_group_id to make those relationships clear.
6. Related Product (related_product)
Identify up to 30 related products from your inventory, specifying the relationship, such as an accessory, required part, substitute or an item often bought alongside the product.
For a camera, this might include a compatible lens, spare battery or carrying case. These relationships can support more useful recommendations during conversational shopping.
7. Document Link (document_link)
Link to publicly accessible PDFs, such as user manuals, specification sheets and installation guides.
These documents give Google access to more detailed product information, which can help answer questions that a short description cannot cover. This is particularly useful for technical products with complex compatibility or installation requirements.
8. Popularity Rank (popularity_rank)
Provide a score between 0 and 100 to indicate how popular a product is within your own catalogue, with higher values representing greater popularity.
This reflects your assessment of the product’s popularity, rather than its customer review rating. At Dream Agility, we recommend using consistent performance data to inform these scores and reviewing them regularly.
Pro Tip: Add the Extra Detail Through a Supplemental Feed
You don’t need to rebuild your primary feed to get started. Google recommends using a supplemental data source to add conversational attributes to your existing product data.
Match each row to the product’s existing id, then add the relevant attribute columns using Google’s required formatting.
Start with your priority products, check that the information is being processed correctly in Merchant Center, and expand from there. Focus on filling information gaps rather than repeating details you already provide.
Build an Agentic Information Structure
AI commerce is moving towards agentic shopping, where AI assistants help customers research, compare and recommend products.
To support this, businesses should create an agentic information structure around their product feeds.
This includes:
- Detailed product information
- Product highlights
- FAQs
- Technical specifications
- Related products
- Supporting documentation
- Popularity signals
Together, these elements create a richer product knowledge base that AI systems can understand and trust.
Prepare Your Product Feeds for Universal Commerce Protocol (UCP)
Universal Commerce Protocol (UCP) is emerging as a framework for machine-readable commerce data that can be used across AI shopping platforms.
By enriching your product feeds today, you’re already taking important steps towards UCP readiness by providing:
- Structured product data
- Detailed attributes
- Product relationships
- Customer-focused content
- Technical documentation
The businesses that invest in product feed optimisation now will be better positioned as AI commerce continues to evolve.
Don’t Forget Your Brand Profile
Google also recommends strengthening your brand presence by completing your Merchant Center business profile and ensuring your brand information is accurate and consistent.
US-based businesses should also explore Google’s Business Agent feature, which provides additional information AI systems can use when representing brands within conversational experiences.
How Dream Agility Can Help
At Dream Agility, we help ecommerce businesses transform standard product feeds into AI-ready commerce assets.
As AI search becomes a major driver of product discovery, enriched product feeds will play a crucial role in helping brands remain visible and competitive.
Ready to Optimise Your Product Feeds for AI Search?
If you want to improve your product visibility across Google AI Overviews, Gemini, Shopping and future AI-powered commerce experiences, contact us here at Dream Agility today.
Our team can help you build AI-ready product feeds that increase discoverability, improve product understanding, and support long-term ecommerce growth.
Book A Call Today!
