AgenticAdvertising.org

AdCP Academy study guide

Everything you need to know to earn your AdCP credential. Three levels from foundational literacy to certified specialist — start free, go as deep as your role requires.

All training is delivered interactively by Addie, AgenticAdvertising.org's AI instructor, with hands-on exercises against live partner sandbox agents.

Credentials

1
Level 1
AdCP Basics
Access
Free and public
Modules
A1 + A2 + A3, ~45 min
Audience
Anyone in advertising

Understand what agentic advertising is, why it matters, and how AdCP works at a conceptual level. No technical background required.

Earns: AdCP Basics credential
Topics covered
  • What broke in programmatic and why AI agents are the next paradigm
  • Agentic buying vs. agentic bidding — the core distinction
  • The principal hierarchy: brands, agencies, platforms, publishers
  • AdCP architecture: MCP transport, tool discovery, agent manifests
  • How agents find each other, negotiate, and transact
  • The role of .well-known/brand.json and brand identity
  • Where AdCP fits relative to OpenRTB, AAMP, and existing standards
Access
AgenticAdvertising.org members
Modules
A1-A3 + specialization track (B, C, or D), ~2 hours
Audience
Professionals who will work with AdCP

Deep, role-specific knowledge with interactive exercises against live sandbox agents. Choose your specialization based on your role.

Earns: AdCP Practitioner credential
Specialization tracks
Track B: Publisher / seller
  • Building and configuring a sales agent
  • Product catalog design and management
  • Creative format specifications
  • get_products and list_creative_formats in depth
  • Measurement and reporting
  • Agent guardrails and safety constraints
Track C: Buyer / brand
  • Multi-agent orchestration
  • Portfolio allocation strategies
  • Brand Protocol
  • Brand identity and compliance
  • Creative workflows
  • Sponsored Intelligence
Track D: Platform / intermediary
  • MCP server architecture
  • Tool discoverability
  • Cryptographic signatures and verification
  • Agent identity and authentication
  • ads.txt and sellers.json integration
  • Migration strategies from existing systems
  • Performance benchmarking
Access
AgenticAdvertising.org members + capstone
Modules
Practitioner + capstone lab (E1-E4), ~3 hours total
Audience
Technical leaders, solutions architects, senior ad ops

The top credential. Requires completing a hands-on capstone lab combining live sandbox exercises and an adaptive exam administered by Addie. Choose your protocol specialization.

Earns: AdCP Specialist credential (protocol-specific)
Protocol specializations
E1: Media buy
  • Transaction lifecycle management
  • Multi-agent orchestration patterns
  • Pricing models and negotiation protocols
  • Delivery reconciliation and discrepancy resolution
E2: Creative
  • Creative lifecycle from brief to delivery
  • Format compliance and validation
  • Cross-platform adaptation strategies
  • Brand consistency enforcement
E3: Signals
  • Signals framework architecture
  • Measurement activation workflows
  • Attribution and optimization models
  • Cross-publisher measurement coordination
E4: Governance
  • Property lists and classification
  • Content standards definition
  • Content calibration workflows
  • Compliance automation and enforcement
Evaluation dimensions
Conceptual understanding
Can you explain why the protocol works this way?
Practical knowledge
Can you apply concepts to realistic scenarios?
Protocol fluency
Do you use correct terminology and understand spec details?
Problem-solving
Can you diagnose issues when things go wrong?
Communication clarity
Can you explain AdCP concepts to non-technical stakeholders?

Key terms and glossary

AdCP
Ad Context Protocol. An open standard built on MCP that provides a unified interface for AI-powered advertising workflows across diverse platforms.
MCP
Model Context Protocol. The underlying transport and tool-discovery layer that AdCP extends for advertising-specific operations.
Sales agent
An AI agent operated by a publisher or seller that exposes advertising inventory through AdCP tools like get_products and list_creative_formats.
Buyer agent
An AI agent that acts on behalf of a brand or agency to discover, evaluate, and purchase advertising opportunities from sales agents.
Brand agent
An AI agent that represents the brand layer in the principal hierarchy, enforcing brand standards and policies across buying decisions.
Agentic buying
The practice of using AI agents to discover, negotiate, and transact advertising at a strategic level — as opposed to agentic bidding, which optimizes within predefined parameters.
Agentic bidding
AI-optimized bid management within existing programmatic frameworks. Operates at the tactical level, adjusting bids and targeting within guardrails set by humans or buying agents.
.well-known/brand.json
A standardized file hosted at a brand's domain that declares brand identity, policies, and agent authorization — enabling agent-to-agent trust verification.
Portfolio allocation
The strategic distribution of advertising budget across multiple publishers and channels, managed by buyer agents using AdCP's multi-agent orchestration capabilities.
AdCP-Training-Mode
A protocol header that signals a sandbox environment for training and certification exercises, preventing any real transactions or spend.
Hosted sales agent
A sales agent hosted and managed by AgenticAdvertising.org on behalf of a publisher, lowering the technical barrier to AdCP adoption.
Sponsored Intelligence
A native advertising format within AdCP where brands sponsor contextually relevant information delivered through AI agents, transparently labeled as sponsored.
OpenRTB
Open Real-Time Bidding. The existing IAB Tech Lab standard for programmatic auctions. AdCP operates at a higher strategic layer, complementing rather than replacing OpenRTB.
AgenticAdvertising.org
The member organization that stewards the AdCP specification, provides certification and training, and supports adoption across the advertising industry.

Recommended preparation

AdCP specification
adcontextprotocol.org
AgenticAdvertising.org
agenticadvertising.org
BokOnAds
bokonads.com
MCP documentation
modelcontextprotocol.io

Frequently asked questions

Do I need technical skills for AdCP Basics?

No. Level 1 is designed for anyone in advertising — media planners, account managers, brand marketers, and executives. It covers concepts and context, not code.

Can I earn multiple specialist credentials?

Yes. You can complete multiple capstone labs (E1-E4) over time. Each earns a separate protocol-specific specialist credential.

How is the exam administered?

By Addie via Socratic conversation. There are no multiple-choice questions. Addie adapts its questioning based on your responses and evaluates across five dimensions.

Do credentials appear on LinkedIn?

Yes. Credentials are issued via Certifier with one-click LinkedIn add, so they appear in your certifications section.

How long are credentials valid?

AdCP Basics does not expire. Practitioner and Specialist credentials are valid for 2 years, after which a renewal assessment keeps you current with protocol changes.

What are the sandbox exercises like?

Live interactions with real AdCP agents running in training mode. Addie guides you through scenarios — discovering products, creating media buys, syncing creatives — against partner sandbox environments.

Is there a cost?

Level 1 (AdCP Basics) is free and open to everyone. Levels 2 and 3 require an AgenticAdvertising.org membership.