AI Shopping Assistant Page Type
AI shopping assistants are already generating high-intent traffic on your site. Shoppers who engage with AI shopping assistants arrive having already explored, compared, and shortlisted — they are closer to a purchase decision than almost any other visitor on your site. That traffic has real value. Using the Sponsored Products infrastructure you already have in place and with no disruption to the shopper experience, the AI assistant page type turns your onsite AI shopping assistant into a retail media surface.
How It Works
When a shopper interacts with your AI assistant, their natural language query is converted into structured shopping signals. Your system — or your LLM partner — extracts the relevant keywords and filters from the conversation and passes them to the Criteo Sponsored Products engine as a standard ad request. The engine runs an auction and returns relevant, high-quality Sponsored Products, which are embedded naturally into the assistant's response alongside organic recommendations. Here’s what the journey looks like:
A shopper asks your AI assistant a question (e.g., "show me affordable running shoes").
Your system or LLM partner extracts keywords and filters from the conversation (e.g., keyword = "running shoes," filter = price under $150).
Those signals are sent to the Criteo Sponsored Products engine as an ad request.
The engine runs an auction and returns the most relevant Sponsored Products.
Winning ads are embedded natively into the assistant's response.
Performance is measured through your existing reporting workflows.
What This Means for Your Existing Setup
AI Assistant is introduced as a new page type. This is the same mechanism used for search, category, homepage, and product detail page (PDP) inventory today. You simply pass the AI Assistant page type alongside the keywords and filters your system has already extracted. Your existing Sponsored Products integration handles the rest. No new infrastructure is required.
Brand partners benefit equally from this simplicity: their active Sponsored Products campaigns automatically become eligible to serve on the AI Assistant surface based on existing keyword targeting. They can adjust investment using Page Type Bid Multipliers and measure performance through the AI Assistant page type dimension in their existing dashboards.
What You Need to Get Started
Before activating AI Assistant placements, confirm the following with your technical team or LLM partner:
Conversation-to-keyword translation. Your system must be able to extract structured keywords and filters from shopper conversations and pass them to Criteo in the standard ad call format. This is the critical integration step that enables monetization.
LLM partner configuration. If you work with an external LLM provider to power your AI assistant, confirm they support keyword and filter extraction and can pass the structured output to Criteo as part of the ad request.
Existing search integration. You will reuse your current Sponsored Products search integration, updated to include the new AI Assistant page type. No separate integration is required.
Measuring Performance
Because AI Assistant is a standalone page type, you can isolate its performance using your existing reporting workflow:
Navigate to the Analytics section.
Select Onsite Sponsored Products Revenue.
Apply the Page Type dimension.
Filter to AI Assistant to view impressions, clicks, attribution metrics, and more, specifically for this surface.
Protecting the Shopper Experience
Shopper trust is foundational to a sustainable monetization strategy. We apply the same strict relevance standards that already power your search advertising to protect the quality of the experience:
Intent-driven placement. Sponsored Products are shown only when they are strongly aligned with what the shopper asked.
Relevance thresholds. Keyword and semantic models ensure contextual matching. If a product does not meet the relevance bar, it is not shown and no ad appears.
Privacy by design. Criteo receives only the structured output from your system: the keywords and filters extracted from the conversation. The raw conversational text stays within your own environment and is never transmitted to us.
Planning Your Monetization Strategy
Beyond the technical setup, consider how this new surface fits into your broader retail media strategy:
Define the shopper experience. Determine how sponsored recommendations will visually integrate with your assistant's organic responses, ensuring ads feel native, helpful, and consistent with your standards.
Establish relevance guardrails. Decide how strictly to apply contextual matching. Maintaining high relevance preserves shopper trust and the long-term value of your AI assistant as a discovery tool.
Match ad presence to the stage of the journey. Shoppers move through distinct stages when using an AI assistant: exploring, narrowing on needs, weighing trade-offs, shortlisting, and finally stress testing before deciding. As you define your relevance guardrails, consider how the nature of a shopper's query signals where they are in the journey, and calibrate your ad experience accordingly.
Next Steps
Contact your Criteo representative to enable the AI Assistant page type for your network.
Learn more about managing page types in your Supply-Side Dashboard.