A first-of-its-kind industry initiative

The industry standard for evaluating AI agents in retail media

Your retail media, run intelligently.

The Agentic Retail Media Council is defining what "good AI" actually means in advertising, and we're inviting the brands, agencies, and platforms shaping the future of retail media to help build it.

Black background with evenly spaced small white dots arranged in a grid pattern.Soft gradient background blending warm orange, pink, and blue hues on a dark backdrop.

A new operating model for retail media

AI agents are already running your campaigns. But how do you know they're any good?

Today, AI agents are making bidding decisions, generating recommendations, and shaping campaign strategy across retail media. But, the industry has never defined what "correct" actually looks like. No shared benchmarks. No audit standards. No way to compare one system to another.

That's the gap the Agentic Retail Media Council exists to close.

Help shape the standard. Don't get standardized.
  • Define the rules before they're defined for you. Shape the evaluation criteria, scoring rubrics, and audit standards the entire industry will adopt; instead of inheriting someone else's framework.

  • Establish category leadership in responsible AI. Founding members are publicly recognized as the brands, agencies, and platforms setting the bar for transparent, trustworthy AI in advertising.

  • Influence procurement before everyone else does. Get early access to the EVAL Framework, scorecards, and the annual State of Agentic Advertising report. These are the tools your peers will use to evaluate AI partners.

A glass box, not a black box.


•The EVAL Framework v1.0, the first standardized scoring system for AI agents in advertising
• EVAL Scorecards for evaluating platforms, partners, and internal systems
• The annual State of Agentic Advertising benchmark report
• Open audit and evidence standards built for advertisers, by advertisers

A small ask. An outsized impact.


1. One executive sponsor, and one working lead from your organization
2. Quarterly 60-minute working sessions
3. Input on the evaluation framework as it evolves
4. Optional public recognition as a founding member

Founding seats are limited. Make yours count.


Submit your interest below. We'll be in touch to discuss founding member participation.

Put Me on the List