Media Buy Protocol

AdCP Media Buy Protocol - Unified interface for AI-powered advertising automation across all platforms.

AI is buying ads. The Media Buy protocol ensures it can buy yours.

This is AdCP’s core advertising automation interface, providing 8 standardized tasks for managing the complete advertising lifecycle—from inventory discovery through campaign optimization.

Protocol Access

Media Buy tasks are accessible through multiple protocols:

• MCP (Model Context Protocol): Direct integration with AI assistants like Claude
• A2A (Agent-to-Agent): Complex agent workflows and collaboration

All protocols provide identical functionality - choose based on your integration needs. See Protocol Comparison for guidance.

The 8 Core Media Buy Tasks

The Media Buy protocol provides these essential operations:

Discovery & Planning

• get_products: Discover advertising inventory using natural language briefs
• list_creative_formats: Understand creative requirements and specifications
• list_authorized_properties: Verify publisher authorization and available properties

Campaign Execution

• create_media_buy: Launch advertising campaigns with complete lifecycle management
• update_media_buy: Modify budgets, targeting, and campaign settings

Creative Management

• list_creatives: Browse and filter creative asset libraries
• sync_creatives: Upload and synchronize creative assets across platforms

Performance Optimization

• get_media_buy_delivery: Track performance metrics and campaign delivery

Key Design Principles

1. Protocol-Agnostic Design: Access through MCP, A2A, or future protocols with identical functionality.

2. Asynchronous by Design: Operations may take seconds to days to complete. The protocol embraces pending states as normal workflow elements. This is not a real-time protocol - response times range from 1 second for simple lookups to days for operations requiring human approval.

3. Human-in-the-Loop: Publishers can require manual approval for any operation. The protocol includes comprehensive task management for human intervention.

4. Multi-Platform Abstraction: A unified interface that works across Google Ad Manager, Kevel, Triton Digital, and more.

5. AI-Optimized: Designed for AI agents to discover, negotiate, and optimize media buys autonomously.

Key Features

• Natural Language Discovery: Find advertising inventory using plain English briefs
• Unified Targeting: Consistent targeting dimensions across all platforms
• Standard Formats: Pre-defined creative specifications powered by the Standard Creative Agent
• Creative Agents: AI-powered agents for building, validating, and previewing creatives
• Creative Flexibility: Support for standard IAB and custom publisher formats
• Real-time Optimization: Continuous performance monitoring and adjustment
• Human-in-the-Loop: Optional manual approval workflows where needed
• Performance Accountability: Built-in feedback loop that tracks publisher delivery against promises

Response Time Expectations

The Media Buy protocol is designed as a timely but not real-time protocol. Response times fall into four categories:

• Simple database lookups (~1 second): Format and creative listings
• Inference/RAG operations (~60 seconds): Product discovery, signal discovery with AI/LLM processing
• Reporting queries (~60 seconds): Delivery metrics with data aggregation
• Asynchronous operations (minutes to days): Campaign creation/updates, creative sync, signal activation with potential human-in-the-loop approval

Implementers should design for asynchronous operation and provide appropriate user feedback during processing.

Separation of Concerns: A Collaborative Model

The Media Buy protocol is built on the principle that optimizing media campaigns is a collaborative process where each party focuses on what they do best. This separation of concerns creates efficiency and better outcomes for all participants.

The Three Roles

1. Publisher Role

Publishers bring expertise and data to optimize campaign delivery. Their needs are simple:

• Money: Budget to work with
• Brief: Clear understanding of campaign goals
• Feedback: Performance signals to know if it’s working

Publishers say: “Give me money, tell me what you’re trying to do, and tell me if it’s working.”

2. Principal (Buyer) Role

Principals maintain control over their brand and campaign strategy:

Upfront Controls:

• Campaign brief and objectives
• Budget allocation
• Targeting overlay (e.g., “must run in California”, “near our stores”)
• Creative approval

Real-time Signals:

• Audience data
• Brand safety requirements
• Frequency capping rules
• Performance feedback

The principal focuses on high-level campaign goals while giving publishers flexibility to optimize delivery.

3. Orchestrator Role

The orchestrator handles the technical mechanics, similar to a DSP in digital advertising:

• Information synchronization between parties
• Creative asset management
• Frequency capping enforcement
• Real-time signal processing (AXE)
• Campaign state management

The orchestrator enables principals to stay focused on strategy rather than implementation details.

Why This Model Works

1. Expertise Alignment: Each party focuses on their strengths
2. Clear Boundaries: Well-defined responsibilities prevent conflicts
3. Flexibility: Publishers can optimize within principal constraints
4. Scalability: Orchestrators handle complexity behind the scenes
5. Transparency: Clear signals and feedback loops

This collaborative approach optimizes outcomes by letting each participant do what they do best, creating a more efficient and effective advertising ecosystem.

Accountability & Trust Framework

AdCP creates a built-in feedback loop that improves marketplace quality over time through measurable performance tracking and accountability mechanisms.

The Performance Promise Model

When publishers respond to product discovery requests, they make implicit performance promises:

• Delivery Estimates: “I can deliver 50K impressions in this package at a $3 CPM”
• Audience Quality: Products targeting specific demographics or behaviors
• Minimum Exposure Commitments: Guaranteed minimum impression delivery per user
• Format Compatibility: Supported creative specifications and requirements

These aren’t guarantees, but they represent realistic expectations that can be measured and tracked.

Measurable Accountability

The protocol enables systematic tracking of publisher performance against their promises:

Delivery Tracking

• Compare actual vs. estimated impression delivery
• Monitor CPM accuracy against quoted prices
• Track completion rates and audience quality metrics
• Measure adherence to minimum exposure requirements

Quality Metrics

• Audience alignment with targeting promises
• Creative format compatibility and performance
• Brand safety compliance and policy adherence
• Response time and operational reliability

Historical Performance

• Track publisher accuracy over time across multiple campaigns
• Identify consistent over-performers and under-performers
• Build reputation scores based on promise fulfillment
• Enable data-driven publisher selection for future campaigns

The Feedback Loop

This creates a self-improving marketplace where performance data influences future opportunities:

1. Discovery Quality: Publishers who consistently deliver what they promise receive higher visibility in product discovery results
2. Allocation Decisions: Buyers can factor historical performance into budget allocation decisions
3. Price Efficiency: Accurate delivery estimates help buyers plan budgets more effectively
4. Marketplace Evolution: Publishers are incentivized to provide realistic estimates and deliver quality results

Trust Building Benefits

For Buyers:

• Reduced risk through performance-based selection
• More accurate campaign planning and budgeting
• Higher confidence in publisher promises
• Data-driven optimization opportunities

For Publishers:

• Competitive advantage through consistent performance
• Higher allocation from satisfied buyers
• Reputation-based pricing power
• Clear incentives for operational excellence

For the Ecosystem:

• Self-regulating quality improvement
• Reduced fraud and misrepresentation
• More efficient allocation of advertising spend
• Long-term relationship building based on performance

This accountability framework transforms the advertising marketplace from a series of one-off transactions into a trust-building system that rewards performance and reliability.

Media Buying Lifecycle

The following diagram illustrates the complete lifecycle of a media buy in AdCP:

Documentation Structure

Task Reference 🔗

Complete API reference for all 8 media buying operations, from product discovery and creative management to campaign creation and optimization.

Capability Discovery 🔍

Foundation concepts including creative format specifications and property authorization. Learn about preventing unauthorized resale and understanding format requirements.

Product Discovery 📋

Natural language approach to finding inventory, including brief structure, product models, and real-world examples.

Media Buys 🎯

Complete campaign lifecycle management from creation through optimization, including asynchronous operations, human-in-the-loop workflows, performance monitoring, and data-driven campaign optimization.

Creatives 🎨

Creative asset management including library management, asset lifecycle, and cross-platform synchronization. Works in conjunction with the Creative Protocol for building and managing creative content.

Advanced Topics 🛠️

Advanced features including targeting dimensions, security models, design rationale, and development tools.

Getting Started

Choose your path based on your role and needs:

For AI Agent Developers

1. Start with Protocol Selection - Choose MCP or A2A based on your use case
2. Learn Capability Discovery - Understand creative formats and property authorization
3. Try Product Discovery - See how natural language briefs work
4. Reference Task Reference - Implement the 8 core tasks

For Campaign Managers

1. Understand the Media Buy Lifecycle - Learn the complete workflow
2. Review Product Discovery - See how to find inventory with briefs
3. Study Policy Compliance - Understand approval requirements
4. Explore Optimization & Reporting - Learn performance management

For Publishers/Sales Agents

1. Learn Authorized Properties - Understand authorization requirements
2. Review Creative Formats - See supported creative specifications
3. Study Advanced Topics - Deep dive into technical implementation

For Technical Implementers

1. Choose your Protocol - MCP vs A2A comparison
2. Study Task Reference - Complete API documentation
3. Review Advanced Topics - Security, testing, and architecture
4. Explore Creative Management - Asset lifecycle and synchronization

The Media Buy protocol makes advertising automation accessible to AI agents while maintaining the human expertise and approval workflows that ensure quality and compliance.