Agents Buy Ads. They Also Build the Shortlist.

Most discussion about agentic advertising focuses on AI agents buying media. That is only half the shift.

An advertiser’s agent can discover inventory and execute a campaign. A consumer’s agent can research a category and narrow the brands a person considers.

One changes how advertising is bought. The other changes whether a brand enters the decision at all.

The customer is still human. But increasingly, an agent reads the sources, compares the claims and assembles the shortlist before that person ever reaches a brand’s website. For advertisers, that means looking at what the consumer’s agent reads, not only at what the advertiser’s agent can purchase.

The shortlist comes before the sale

Imagine someone asking an assistant to compare insurance for a move abroad. The assistant returns three options and explains how they differ. The person then compares prices, asks a friend or contacts a provider.

The assistant did not make the final decision. But it decided which brands got a closer look.

A brand can have an excellent website and the right product and still miss that comparison. Its landing page only persuades once someone arrives. Brand familiarity still matters, but it now works alongside whatever the assistant finds while researching the question.

That creates a specific problem for advertisers: does the information available to the assistant make clear where our product fits?

Accurate product information, strong owned content and independent editorial coverage all play a part. The opportunity here is more specific: when an agent reads a relevant publisher, can a sponsored placement add useful product information at that moment?

A placement in the research, not beside it

That is the job Context Ads are designed to do.

When an AI agent retrieves content from a participating publisher, Oasy can serve publisher-approved, clearly labelled sponsored information within the AI-accessible version of that page. The publisher controls the placement and approves its claims; the ordinary human-facing article remains unchanged.

The advertiser buys a placement in the material the agent is reading, rather than a display unit beside the finished answer, and the publisher shares in the revenue.

A relevant insurance placement might explain a product’s eligibility requirements or its cover for a particular situation. Its value depends on whether that information answers the consumer’s question, not on whether the brand name appears. The agent will weigh it against other sources, so the creative has to be substantive: something useful and supportable about the product.

What is AdCP?

AdCP, the Advertising Context Protocol, is an open standard that gives buying and selling agents a shared language for advertising. It supports workflows such as discovering media products, creating media buys, exchanging creative and requesting delivery reports across platforms.

AdCP is not an advertising format or a marketplace. It standardises how agents work together; Context Ads are one product those workflows can support.

Oasy’s sales agent is verified on Apostra, making Context Ads discoverable and purchasable within an agentic buying environment. A buying agent can evaluate the placement against its brief: what information the consumer’s agent will receive, in which publisher context, on what terms, and what counts as delivery.

The test is consideration, not connectivity

Measuring Context Ads means separating three questions.

Delivery: Did the agreed placement run, in the agreed context, and does the delivery report reconcile with the buy?

Answer visibility: Did the brand or its product information appear more often in answers to the questions the campaign targeted? That needs a baseline and consistent monitoring, not favourable answers selected afterwards.

Business outcomes: Did that presence contribute to enquiries, purchases or another meaningful customer action, even when the person continues their research elsewhere?

An agent-to-agent transaction answers the first question. The other two are what matter to the business, and they need to be measured on their own terms.

So the decision is not whether to “buy AI”. It is whether there is a specific customer question, a relevant placement and a credible way to assess the effect.

Oasy works with advertisers and publishers to build Context Ads around specific customer questions, relevant editorial environments and measurable outcomes.

If your brand needs to enter consideration during AI-assisted research, not simply appear beside the finished answer, get in touch.