Home AI Surfaces Advertisers Spark Blog AboutContact Us

Glossary

The vocabulary of agentic advertising.

The core terms of advertising on AI surfaces: the protocols, the transaction models, the targeting methods, and the formats. Each entry opens with a working definition, then shows how the concept behaves in practice, with current examples from Adgentek's Agentic Ad Server, AdsMCP, and Spark.

01Agentic advertising AI agents planning, buying, and selling media over open protocols. 02Ad Context Protocol (AdCP) The open standard for how AI agents transact advertising. 03MCP advertising Connecting ad infrastructure to AI apps via the Model Context Protocol. 04Agent-to-agent media buying Buyer agent to seller agent, no dashboard in between. 05Intent-native targeting Matching on what is being said now, not on who was profiled last month. 06Conversational ad format An ad unit built for dialogue instead of a static impression.

New to the category? Start with what conversational advertising is, then read how conversational ads work and how brands advertise in AI chatbots. For the numbers, see the H1 2026 conversational AI advertising data report.

FAQ

The short versions.

What is agentic advertising in one sentence?

Agentic advertising is media planned, bought, and sold by AI agents that transact with each other over open protocols such as the Ad Context Protocol (AdCP), replacing dashboard-driven Programmatic workflows.

What is the difference between AdCP and MCP?

MCP (Model Context Protocol) is the general-purpose transport connecting AI systems to tools. AdCP (Ad Context Protocol) is the advertising-specific open standard that runs on top of it, defining briefs, proposals, media buys, and delivery reporting between agents.

Ready when you are.

Contact Us hello@adgentek.ai