Agentic offer discovery is the pre-transaction process through which an AI agent or AI product uses a user’s expressed goal, explicit constraints, and appropriate context to retrieve and evaluate seller-specific commercial offers.
The process can include candidate retrieval, deterministic eligibility filtering, relevance scoring, ranking, and presentation. In AON, discovery returns attributable structured offers rather than product records alone. The merchant remains authoritative for current price, availability, transaction eligibility, and binding terms.
Offer discovery in an AON query
An AI product sends the user's text or image intent plus only the context and explicit constraints its active integration supports. AON applies eligibility rules and relevance signals to return structured offers or a valid empty result. The publisher controls whether and how to present a returned offer.
See the Offer Query APIHow agentic offer discovery works
Capture the request
The AI product supplies the user’s expressed goal, accurate context, and any explicitly supported constraints without inventing attributes.
Retrieve candidate supply
The offer system finds commercial offers that could plausibly address the request.
Apply eligibility
Deterministic serving rules remove offers that should not be returned for the available context.
Estimate relevance and rank
Eligible candidates are ordered by their estimated usefulness to the intent and placement.
Return offers or an empty result
The product receives ranked machine-readable offers, or an empty result when no active eligible offer matches.
- Relevance cannot override a failed deterministic eligibility rule.
- Stated preferences can influence relevance, but only explicitly supported constraints should be treated as deterministic eligibility filters.
How offer discovery and intent matching work together
Intent matching interprets how well an offer fits the user's expressed goal. Offer discovery is the broader workflow that retrieves candidate supply, applies eligibility rules, uses relevance signals to rank the remaining offers, and returns structured results or an empty result. See the Intent matching definition.
Example in practice
A user asks for a compact travel camera under $800. The publisher sends the request in the intent field and can add an AON camera category ID when that category is an explicit deterministic constraint. AON filters candidate offers for eligibility, ranks the remaining offers against the request, and returns structured results or an empty offers array.
Common questions about agentic offer discovery
Is intent matching the same as agentic offer discovery?
No. Intent matching measures how well the expressed need fits candidate offers. Agentic offer discovery is the broader process that retrieves supply, applies eligibility rules, ranks candidates, returns structured offers, and handles empty results.
What happens when no offer matches an AON query?
AON can return a successful response with an empty offers array when no offer should be returned for the request. This can reflect no suitable eligible supply or a runtime rule that suppresses an offer. An empty result is valid, not an API failure or a reason to substitute a weak or ineligible offer.
What should an AI product do after AON returns no offers?
Treat the response as a valid empty state rather than an integration failure. Keep data.request_id for diagnostics and, when appropriate, use documented refinements or relax a supported constraint before sending a new query. Do not replace an empty result with an unrelated offer.
Does an offer-discovery response confirm final price and availability?
No. It reflects the information available during discovery. The merchant or checkout system must confirm current price, availability, eligibility, and binding terms.
Sources
- AON Offer Query APIAON documentation
- UCP Catalog capabilityRelated standard
- OpenAI product discovery in ChatGPTIndustry source
- Schema.org OfferRelated standard