How to run AI ads without keywords | Particular Audience
To run AI ads without keywords, let the advertiser pick products and score each product against the meaning of every incoming request. High-scoring.
How to run AI ads without keywords
To run AI ads without keywords, let the advertiser pick products and score each product against the meaning of every incoming request. High-scoring products enter the auction; the rest are skipped.
Compared with keyword targeting
Modern keyword campaigns already use broad, phrase and intent matching, and PA's own search is hybrid. The difference here is the setup: the advertiser selects products or a collection and semantic matching finds candidate requests, instead of building and maintaining keyword lists. It is not a guarantee against poor matches: set a testable objective, review the requests a campaign matched, add exclusions, and keep hard constraints such as category and availability checks.
Why keywords break in chat
Shoppers ask in full sentences, with constraints and context. No keyword list can cover every phrasing.
How meaning scoring works
Adaptive Transformer Search represents requests and products as meanings and scores how well they fit. A threshold decides eligibility.
What advertisers gain
Reach on phrasings they would never have bought, with relevance review and exclusions to limit poor matches.
Optional include and exclude terms
No keywords are required, but advertisers can optionally add include terms (prompts, keywords, queries or strings) to make sure matching requests are considered, and exclude terms to block contexts they never want. Excluded terms always win. There are two cut-offs: a request scoring 0.72 or above can be shown, and an include term can admit a request scoring between 0.60 and 0.72; nothing below 0.60 is shown, and caps and category rules still apply.
Run it with Particular Audience
Sponsored Prompts runs on Adaptive Transformer Search and the open-source ATS MCP server. Advertisers pick products or a collection to target, while the retailer's AdOps team sets the budget, rate, dates, eligibility, approval and reporting; eligible items can then appear in chats, prompts and AI searches on the retailer's own assistants where they score highly, and the retailer's interface must show or speak the sponsored label, holding back any result that arrives without it.
Sources
- Adaptive Transformer Search MCP server (GitHub), checked 2026-10-25
How Particular Audience does it
Frequently asked questions
Can advertisers still add keywords?
Keyword targeting remains available in [Sponsored Search](/sponsored-search); Sponsored Prompts does not need it.
Who sets the threshold?
The retailer.
Related guides
- Sponsored products in agentic and conversational AI: How retailers run sponsored products inside the AI assistants they operate: relevance-first ranking, labelling, CPM and CPC pricing, pacing and holdout measurement — and why ChatGPT and Claude apps are discovery-only.
- Retail MCP product search with Adaptive Transformer Search: Connect Adaptive Transformer Search to ChatGPT, Claude and onsite assistants through PA's open-source MCP server, with retailer-controlled ranking and checkout.
- How to monetise an AI shopping assistant: To monetise an AI shopping assistant, connect it to a product search that returns relevant sponsored products, label them in answers, cap how many appear,.
- How to launch Sponsored Prompts: Launch Sponsored Prompts by choosing one AI surface, inviting advertisers to pick products or collections, setting the relevance threshold and sponsored.