Blog Post

Agents, Content, and Commerce: Preparing for Google's Universal Commerce Protocol (UCP)

By Steve Duran & Scott Rowland, 03.05.2026


 

In our recent article on Universal Commerce Protocol, Google, and What It Means for Enterprise Organizations, we explored how Google's new open standard enables AI agents to execute commerce transactions on behalf of consumers. We have formally entered a new era of fully autonomous robots buying things on behalf of buyers—seemingly sooner than many organizations expected. These new protocols are the missing links that establish a common, shared framework that allows AI agents to actively engage in everything from search and product discovery, through checkout and payment, and into post-purchase support. Merkle’s research predicts that 25% of US consumers will employ agents to complete online transactions by 2030, with that number climbing to 33% for B2B buyers.

With nearly 1 in 4 US buyers set to leverage agentic commerce capabilities, many enterprise organizations are asking: How do we ensure our products are the ones these agents recommend?

It starts with optimizing for agentic consumption instead of exclusively publishing for human consumption. Enterprise brands must think beyond product data. While structured catalog information remains essential, agents also need the contextual content that surrounds, explains, and differentiates a company’s products and brand. If a quarter of US consumers use agents to complete online transactions by 2030, digital experiences will start to be judged, at least in part, by how quickly agents can ingest content and contextualize it for online consumers. These factors will shape how organizations think about creating content that’s truly agent-ready.

Most current enterprise content strategies are designed for human consumption, not machine comprehension. Brands that are sluggish to adapt will hemorrhage attention. If agents can’t confidently source reliable content from you, then those agents will turn to third-party interpretations, outdated documentation, or worse, competitor alternatives.

What "Agent-Ready Content" Actually Means

Agent-ready content architecture has four defining characteristics:

  1. Semantic Structure Over Visual Design
    AI agents don't experience your content the way humans do. They need content that's semantically structured and marked up in ways that communicate meaning, relationships, and context. Your content management systems (CMS) need to output structured data that connects content assets to products, solutions, use cases, and customer segments. For instance, when an agent encounters your buying guide, it should programmatically understand which products it references, what problems it addresses, and what type of buyer it's written for. In short, structure your content for machine comprehension, not just human browsing.

  2. Connected Content Ecosystems
    In many enterprise organizations, content lives in silos and is static. Marketing content sits in the CMS (ie., Adobe, Contentstack, Sitecore, etc.); technical documentation lives in a separate system (ie., Bynder, Adobe, etc.); training materials exist somewhere else entirely; and product data resides in PIM (ie., Inriver). Having fresh, accurate, dynamic content will be key for success in an agentic era. Investments in a connected ecosystem must be accelerated to prepare for an influx of agentic search and commerce activity.

  3. Real-Time Content Delivery
    Just as UCP requires real-time product data, agents require dynamic, updated content as well. If your application guide references a discontinued product, or your compatibility matrix hasn't been updated to reflect new offerings, agents will surface inaccurate information. Or worse, they'll recommend competitors with better-maintained content. Traditional publish cycles measured in days or weeks don't work when agents are making recommendations in milliseconds. Your content supply chain needs to support continuous updates with immediate propagation to agent-facing channels.

  4. Governance and Claim Control
    Trust matters for your content strategy. AI agents prefer content that is attributable and governed because it reduces the risk of recommending the wrong thing. Enterprise organizations should treat key product claims, guidance, and performance statements as governed content objects with clear audit trails. This enables agent-facing answers that are relevant and defensible. 

Practical Steps for Content Leaders

If you're responsible for content strategy, content operations, or digital experience platforms at an enterprise organization, here are your five immediate action items to get your content agent-ready:

  1. Audit your content for machine readability. 
  2. Assess your CMS capabilities. 
  3. Map your content ecosystem connections. 
  4. Evaluate your content supply chain velocity. 
  5. Build the business case for investment. 

What's Next

Merkle is actively helping enterprise organizations build agent-ready content architectures.

Our teams bring deep expertise in Adobe Experience Cloud, content operations at scale, and the emerging requirements of agentic commerce. We understand that this isn't just a technology challenge. It's a fundamental shift in how enterprise organizations create, manage, and deliver content. We offer distinct capabilities like Adobe GenStudio dentsu+, our first-to-market content supply chain solution, combining Adobe's GenStudio capabilities with dentsu's services, technology, and audience intelligence. 

The organizations that will win in agentic commerce won't be the ones just producing more content. They'll be the ones producing content that's simultaneously optimized for human engagement and machine comprehension. Reach out if you’re ready to learn more on how you can thrive in this era of agentic commerce.

You might also like:

Agents, Content, and Commerce: Preparing for Google's Universal Commerce Protocol (UCP)