Our recent blog post on Universal Commerce Protocol, Google, and What It Means for Enterprise Organizations explored how Google's new open standard, developed in close partnership with Shopify and other industry leaders, established a shared language for AI agents to execute commerce transactions, including fully agentic commerce workflows. With this new standard, agents can now not only shop on behalf of consumers but also make purchases without human involvement. Merkle's own research predicting that 25% of US consumers and 33% of B2B buyers will employ agents to complete online transactions by 2030.

If the agentic era is here and a quarter (or more) of US consumers will leverage some form of agentic capability within the next four years, then the critical question becomes: Is your product data ready for the agentic era?

In an agentic commerce world, catalog readiness is no longer about filling in fields. It is about whether an AI agent can confidently reason about a product, compare it, substitute it, bundle it, and recommend it without human interpretation. The new capabilities enabled by the Universal Commerce Protocol are only as powerful as the data it exposes. Poor data essentially equals poor results in a world where LLMs are informing significant parts of the buying journey. 

AI agents don’t see your beautiful product photography the way humans do. They don’t read marketing copy with emotional intelligence. They read structured data that’s tailored to how they want to consume it (UCP standard), and they make instantaneous decisions based on what they find. 

The implication for enterprise brands: if their data isn’t ready, they will miss sales when they’re competing in a space that supports agentic-selling motions. 

Enterprise organizations need to reframe their thinking about PIM. PIM isn't a back-office data repository. It's part of the core foundation of AI commerce strategy.

After years of working with PIM partners like Inriver, we've identified several key capabilities that differentiate an AI-ready product data infrastructure for agentic commerce use cases:

1. Rich Attribute Modeling
   Agentic commerce demands product data that goes far beyond basic title, description, and price. Effective PIM implementations capture compatibility relationships, use-case scenarios, technical specifications at the variant level, and answers to common buyer questions. All that product data must be structured for machine readability.
2. Dynamic Taxonomy and Relationships
   Remember, AI agents don't just search for products. They explore a variety of relationships around product data, such as which accessories complement a product and which items can substitute for out-of-stock SKUs. This understanding within a broader solution is essential for agent-driven discovery.
3. Real-Time Syndication
   UCP and Model Context Protocol (MCP) work in real time. Agents query current inventory, live pricing, and actual availability. When a PIM platform is strategically built, it integrates seamlessly with enterprise resource planning (ERP), order management systems (OMS), and commerce systems to ensure agents always access accurate, up-to-date data.
4. Multi-Channel Governance
   In an agentic commerce model, governance does not disappear. It becomes more important. While AI agents may query, reason over, and act on product data, human-in-the-loop review, approval workflows, and controlled release mechanisms remain essential to ensure accuracy, compliance, and brand trust across every channel, including AI-driven surfaces.
5. AI-Ready Data Pipelines
   Product data must be accessible programmatically, not just through manual exports or batch feeds. When an agent queries inventory or pricing, it needs current data, not yesterday's export file. This requires an API-first architecture. Consider organizing product catalog data so that every product, across every business unit and geography, follows the same attribute structure. This gives agents predictable responses every time they query your product line.

Organizations that want to thrive with agentic commerce should prepare their data, technology, and teams to implement the capabilities we’ve identified. If you’re responsible for managing your brand’s product catalog, here are five practical steps you can take to get ready:

1. Audit Your Product Data Completeness
2. Establish Data Governance for Agent-First Distribution
3. Invest in Integration Architecture
4. Build for Protocol Interoperability
5. Partner with PIM Vendors Investing in AI Commerce

In an agentic era, your PIM isn't just infrastructure. It's a competitive advantage.

Merkle and Inriver are actively helping enterprise organizations prepare for the agentic commerce era. We’re seeing enterprises shift from asking “is our data complete?” to “can AI confidently interpret and act on our product catalog?” That shift changes everything, from how product data is modeled to how it is governed and released. Whether you're assessing your current catalog readiness, planning a PIM implementation, or building integration architectures for AI-ready data distribution, we're here to help you navigate this transformation.

Reach out to discuss how your organization can build the data foundation for AI commerce success.