Google Cloud's Invisible Shelf Strategy: How CPG Brands Can Be Chosen by AI Agents

Google Cloud Announces Retail Strategy for the Agentic Era

The invisible shelf: How CPGs can win agentic commerce

Google Cloud explains how CPG companies can prepare for the agentic commerce era with the Invisible Shelf strategy.

Google Cloud published an article in October 2025 explaining strategies for retail and CPG (Consumer Packaged Goods) companies in the agentic commerce era. In January 2026, they further introduced the concept of the "Invisible Shelf" for CPG companies.

Google Cloud states: "The next big shift is not just a minor change, but a fundamental transformation in how consumers discover, research, and purchase products. We are entering the era of agentic commerce."

This paradigm shift heralds the arrival of the "Invisible Shelf" era, where products are chosen by AI agent decisions rather than physical store shelves or e-commerce search results.

Industry Background and Trends

In traditional e-commerce, consumers themselves visited websites, searched, compared, and made purchase decisions. In agentic commerce, this entire process is delegated to AI agents.

Google Cloud outlines the evolution as follows:

• Predictive AI Era: Improved demand forecasting accuracy
• Generative AI Era: Automated content generation
• Agentic AI Era: Executing actual purchasing actions on behalf of users

The key point is that consumers' personal AI agents automatically interact with multiple retailers' AI agents to discover, compare, negotiate, and complete purchases for optimal products. Winning in this "invisible" space becomes the challenge for CPG and retail companies.

Two Interaction Models

Google Cloud presents two major models in agentic commerce.

Consumer-to-Merchant (C2M) Model: Consumer personal agents interact with merchant agents from multiple stores. For example, when asked "I'm going to the Canadian Rockies in August, suggest 2 outfit sets that match my style," the agent learns the consumer's preferences, budget, and needs, then automatically selects optimal combinations from multiple stores.

Merchant-to-Merchant (M2M) Model: A model where store AI agents cooperate with each other. When a product not in stock is requested, the store's agent negotiates with other stores' agents to procure the product and complete the transaction. This prevents lost sales and captures additional revenue.

Three Requirements for Being Chosen by AI Agents

Google Cloud outlines three strategies needed to win on the "Invisible Shelf."

1. Agentic Optimization of Product Data

Structured data is essential for AI agents to correctly recognize and recommend products. Not just descriptions, but attributes like color, size, use case, price range, and inventory status must be prepared in machine-interpretable formats.

Google Cloud recommends "reliable, structured data with rich metadata mapped to user intent."

2. Building Agent-Ready Infrastructure

Infrastructure enabling agent-to-agent dialogue is necessary. Google Cloud provides the following protocols and tools:

• Agent-to-Agent (A2A) Protocol: An open-source framework establishing common language between agents
• Agent Payments Protocol (AP2): A payment framework enabling agent-driven payments
• Agent Development Kit (ADK): A toolkit simplifying the building and deployment of multi-agent systems

3. Human Oversight and Governance

To build trust in full automation, human intervention for critical decisions, anomaly detection, audit logs, and transparency are essential.

Impact and Practical Applications for E-commerce Businesses

The "Invisible Shelf" strategy demands concrete actions from e-commerce businesses.

Actions to Start Now:

1. Product Data Inventory and Structuring: Evaluate whether current product data is interpretable by AI agents and enrich attribute information

2. API Development: Build APIs that can provide inventory, pricing, and product information to external systems in real-time

3. Standard Protocol Compliance Planning: Develop a roadmap for compliance with standards like A2A, AP2, and ACP

PayPal Collaboration: Google Cloud and PayPal jointly announced an agentic commerce solution for merchants. Combining PayPal's payment infrastructure with Google Cloud's AI technology, they're creating an environment where even small and medium businesses can easily enter agentic commerce.

Conclusion

Google Cloud's "Invisible Shelf" concept demands a fundamental mindset shift from CPG and retail companies.

Previously, competition was about "how to secure good shelf positions" or "how to rank high in search results." In agentic commerce, however, winning depends on "being recognized as the optimal choice by AI agents."

Google Cloud states: "The future of commerce is being built through agent collaboration. Now is the time to build." Preparation for product data structuring and standard protocol compliance should begin now.

References

• Google Cloud: Agentic commerce is here - How retailers can prepare (October 2025)
• Google Cloud: The invisible shelf - How CPGs can win agentic commerce (January 2026)
• Google Cloud: New Retail Solutions for the Agentic AI Era
• Google Cloud: Agentic commerce solution from PayPal and Google Cloud
• Google Cloud: AI Agent Trends in Retail & CPG 2026
• Google Cloud: The agentic era in retail and CPG