What Agentic Commerce Protocols Mean for Order Management Systems

The Urgency No One Can Ignore

If you're leading operations at a mid-market or enterprise retailer, you've felt it: a measurable, material hole in your revenue that correlates directly to declining search traffic. Not because you changed anything. Because AI is now answering the questions your customers used to ask Google—and those answers increasingly route them to your competitors or to agents bidding on behalf of your own products.

Beauty and apparel brands are down 30% in SEO traffic over the last six months. Not hype. Real money.

This is why agentic commerce isn't a "2027 problem" anymore. It's a 2026 urgency. And that urgency has massive implications for how order management systems need to work. 

Three Truths About Agentic Commerce and What They Actually Mean

Agentic commerce is easy to overhype because the phrase sounds futuristic. But the near-term reality is much more practical and much more urgent. The biggest changes are not happening at the final “buy” button. They are happening earlier, in the messy middle where customers discover products, compare options, build confidence, and decide who to trust. For OMS leaders, that distinction matters. Agentic commerce is not one big shift. It is three connected shifts that change how commerce systems need to behave.

This Isn't About Robots Buying Toilet Paper...Yet

There's a lot of noise about "zero-click shopping," which is where autonomous agents are placing orders without human intervention. That's the headline. That's also not the primary threat or opportunity happening right now.

The real shift is how consumers discover products and make purchase decisions. Instead of navigating to your website and hunting through filtered search results, they ask an AI agent. Instead of visiting Amazon's homepage, they chat with ChatGPT. Instead of Googling "hiking boots for wide feet," they get Claude or Gemini to research options and present recommendations.

What this means for OMS: The commerce platform's role is fundamentally changing. Your website is no longer the primary interface between consumer and product. Agents are. And agents need real-time, authoritative data to do their job: inventory availability, accurate delivery dates, pricing, inventory position across all nodes in your network, and detailed product context that no static feed has ever captured.

Your OMS has to be the backbone that makes this possible.

The Web Is Dying (As a Shopping Interface)

If you've noticed that e-commerce websites have become increasingly hostile—cookie banners, "save 15% off" popups, retail media networks hiding products behind sponsored listings—you're not imagining it. The web, as a shopping experience, has degraded significantly. And that degradation is accelerating migration to AI-powered discovery.

Meanwhile, consumers still spend $30 trillion annually on B2C transactions. They haven't stopped buying. They've just shifted where they buy.

What this means for OMS: Commerce platforms that cling to web-only experiences are building for a shrinking market. The requirement is now omni-interface: web, chatbots, voice assistants, embedded agents. And all of those channels need to pull from the same source of truth for inventory, orders, and fulfillment. If your OMS is tightly coupled to your website platform, you have a problem. If it's composable and API-first, you have options.

Product Data Is Your New Competitive Moat

In the Google-search era, a retailer could get away with minimal product attributes. Five keywords in a search box meant you only needed five attributes in your product feed. The shopper's intent was narrow.

Now, the opposite is true. Agents have enormous context windows and sophisticated reasoning. They want to know everything about your products: materials, origins, certifications, fit guidance, comparative advantages, customer reviews in contextual slices, Q&A that tells a story. Brands that can articulate their products compellingly at this level will win. Brands that treat product data as a checkbox exercise will lose.

This is your moment to reclaim the role of product experts. You're not competing on lowest price or fastest shipping anymore (though those matter). You're competing on whether you can tell agents why your hiking boot is the right choice for someone with a specific foot shape, activity profile, and past buying history.

What this means for OMS: Product data flows are no longer separate from order management. The same detailed product context that feeds agents also needs to feed your OMS, so that when an agent finds the right product and initiates a transaction, your system has complete visibility into what was promised and what needs to be fulfilled

The Agentic Protocol Landscape: Control Your Destiny, or Have It Controlled For You

In the endless alphabet soup of emerging agentic commerce protocols, it's understandable to feel lost. Let's separate signal from noise.

The Selling Channel: ACP, UCP, and Beyond

Agentic Commerce Protocol (ACP)—led by Stripe and OpenAI—governs how orders are captured and passed between AI agents and commerce platforms.

Universal Commerce Protocol (UCP)—from Google and Shopify—aims to standardize how product data and transactions flow through agent-to-ecommerce interfaces.

Google just announced four new additions to UCP and is rolling out a merchant center (similar to Google Shopping) where brands can provide supplemental product information. This is moving fast. UCP is already gaining traction because Google has the distribution to make it matter.

The strategic decision: Most brands don't need to "pick a protocol." That's an implementation detail. Your job is to ensure your product data and commerce platform can speak both languages. By mid-2026, supporting both ACP and UCP should be table stakes—not a competitive differentiator.

The Fulfillment Channel: onX (Order Network eXchange)

Here's where the OMS community has taken control of its own destiny.

Order Network eXchange (onX) is an MCP (Model Context Protocol) interface that sits on top of your OMS, 3PL systems, WMS, and ERPs.

It exposes:

• Resources: Orders, inventory, customers, fulfillment capacity
• Tools: Capture order, exchange, return, update inventory

By providing a uniform interface across 86+ vendors (every major OMS player, every major 3PL), Onyx allows AI agents to access the fulfillment layer with the same level of abstraction that ACP and UCP provide on the selling side.

Why this matters: ACP and UCP are "sufficient" for capturing an order. But an agent recommending a product needs to know: Can you actually fulfill it? When? From where? With what shipping speed? onX is the answer layer that says, "Yes, we can deliver your hiking boot by Saturday morning because our California DC has it in stock and you're in California."

Without onX (or something equivalent), you're either manually syncing data between layers (fragile and slow) or you're at the mercy of proprietary vendor lock-in.

The Product Data Layer: Still Being Defined

One gap that remains: how agents consume rich product data (attributes, reviews, Q&A, storytelling) isn't yet standardized across protocols. ChatGPT has one format. Google's UCP is still evolving. Stripe has another approach.

Implication: Brands need to invest now in product data infrastructure that's flexible enough to support multiple formats. Store your source of truth in a way that allows you to export to any protocol standard that emerges.

What Do the Agentic Protocols Mean for Your OMS?

Four structural changes. For OMS teams, the impact of agentic protocols isn't theoretical. These protocols only work if agents can access accurate, current, decision-ready fulfillment data before the order is placed. That means the OMS can no longer sit downstream waiting for completed orders. It has to participate earlier in the buying journey, powering availability, delivery promises, routing decisions, and customer-specific recommendations in real time.

Real-Time Inventory Visibility Becomes Mandatory (Not Optional)

If an agent is going to promise "delivery by Saturday," your OMS needs to know, in real-time, what inventory is available in each node of your network. Not batch updates. Not daily syncs. Real-time.

This changes how you architect inventory allocation, capacity reservations, and demand signaling. Your OMS needs to be fast enough to handle millions of concurrent availability checks from agents.

Fulfillment Context Has to Travel Upstream

Your OMS currently gets an order, picks it, ships it, and reports back. New model: agents need to know fulfillment constraints before they recommend a product.

• Can you fulfill this size/color combination from a store near the customer?
• What's the actual delivery date (not an estimate)?
• Does this customer have loyalty data we should factor in?

Your OMS needs to expose this context upward through APIs so agents can make informed recommendations.

Multi-Channel, Multi-Node Orchestration Becomes the Core Competency

Agents will route orders based on cost-to-serve, delivery speed, and inventory position. Your OMS has to orchestrate across stores-as-fulfillment-nodes, DCs, 3PLs, drop-ship partners, and potentially even marketplace fulfillment (Amazon MCF, etc.).

This isn't new conceptually, but the speed and complexity at which it needs to happen is orders of magnitude higher.

Data Silos Between Product, Commerce and Fulfillment Must Collapse

Your product team owns product attributes and storytelling. Your ecommerce team owns product feed formats for your website. Your OMS/fulfillment team owns inventory and shipment data.

In an agentic world, these are the same system. An agent needs to correlate:

• "This customer bought hiking boots from us before and rated them 5 stars"
• "We have this exact boot available in a warehouse 50 miles from the customer"
• "Our delivery window for this origin-destination pair is 2 days"

If your data is siloed, you can't compete.

A Practical Path Forward: What to do Today, Not Tomorrow

Waiting for protocols to stabilize or for your vendor to build out features is a luxury you don't have. Here's what winners are doing now:

For Retailers & Brands:

1. Audit Your Product Data: Start with Sid's mandate: make sure your product information is comprehensive, searchable, and ready to be consumed by AI systems. Not 5 attributes. Hundreds. Include reviews, Q&A, material specs, fit guidance, sustainability claims—everything a bot would need to recommend confidently.
2. Invest in API Quality: Your OMS and ecommerce platform need world-class APIs. Not for "future-proofing." For now. Agents are already trying to consume your inventory data. Are you making it easy or hard?
3. Test Agent Channels Early: Don't wait for ChatGPT Shopper or Google's agent to ship in full commercial form. Start experimenting with ACP and UCP now. Work with your technology partners to prototype how agents would discover and fulfill orders from your business.
4. Build a "Chief Data Officer" Mindset: Product data isn't a fulfillment concern anymore. It's a strategic asset. Organize your team around data quality, enrichment, and distribution—not around channels or platforms.
   
   

For Systems Integrators & OMS Vendors:

1. Become Fluent in These Protocols: ACP, UCP, onX. If your customers ask, "How do we connect to agents?" and you can't give a crisp answer, you've lost their confidence. Make this a standard part of your pitch and implementation methodology.
2. Expose onX (or Equivalent): If you haven't yet, add support for MCP interfaces that expose your OMS data in a protocol-agnostic way. This is how you future-proof your customers and your own business.
3. Help Customers Understand the Product Data Gap: Most OMS implementations are good at order-to-shipment workflows. Few are set up to handle product-level context that flows into and out of the fulfillment layer. This is a huge opportunity to expand your value.
4. Join the Standards Community: Order Network Exchange, Commerce Ops Foundation, and other industry bodies are where this is being decided. Don't be a passive observer. Contribute.
   
   

The Real Question: When?

Jason Goldberg (speaking virtually, and disagreeing with Scott, as always) posed the right question: not "will this happen?" but "when?"

Our read:

• 2026: Agents are a growing fraction of commerce discovery (10-20%). Most transactions still have human friction (human clicks the buy button). But the revenue impact is material enough that brands can't ignore it.
• 2027-2028: Autonomous transaction capabilities move from edge cases to mainstream. Some categories and use cases are fully agent-driven. Others are still mixed.
• 2029+: Agentic commerce is normalized. The question isn't "should we support agents?" It's "which agents are we optimizing for?"
  
  

By then, if your OMS isn't ready, you're not just behind. You're out.

The Bottom Line: This Isn't About Technology, It's About Market Structure

Agentic commerce isn't reshaping retail because LLMs got smarter (though they did). It's reshaping retail because the economic structure of discovery is changing.

For a century, retailers controlled the storefront. For 25 years, they fought for dominance on digital storefronts. Now, a third party (AI) is the storefront. Brands don't get to control that experience. They get to feed it.

This is uncomfortable. It's also inevitable.

The OMS is the critical infrastructure that either enables brands to compete in this new world or prevents them from competing at all. Not because of features. Because of data quality, real-time visibility, and the ability to fulfill complex, agent-driven orders efficiently across a network of fulfillment nodes.

The brands and systems integrators investing in this now will find that the protocols, standards, and capabilities they're building are exactly what the market demands by 2027. The ones waiting for clarity will be in catch-up mode.

What Retailers Should Do Next

Agentic commerce is still taking shape, but the OMS implications are already clear. Retailers do not need to predict which protocol wins. They need to make sure their inventory, product, order, fulfillment and customer promise data can be exposed accurately, securely, and in real time across whatever agent-driven channels emerge next.

That starts with a practical readiness assessment:

• Can your OMS expose real-time inventory availability across stores, DCs, vendors, and 3PL nodes?

• Can your systems provide accurate delivery promises before the order is placed?

• Can fulfillment capacity, sourcing rules, and delivery options be accessed through APIs?

• Can product, commerce, and fulfillment data work together instead of living in separate silos

• Can your architecture support ACP, UCP, onX, MCP, or whatever protocol layer becomes dominant?

The retailers that act now will not just be “ready for agents.” They will have stronger inventory accuracy, better customer promises, more flexible fulfillment, and a cleaner foundation for every digital commerce channel.

Is Your OMS Ready for Agentic Commerce?

Agentic commerce will put new pressure on the systems behind every customer promise: inventory visibility, order promising, fulfillment orchestration, product data and APIs.

Nextuple helps retailers assess, modernize, and integrate the OMS, inventory, and fulfillment capabilities needed for AI-driven commerce channels.

This article is based on a recent Order Management Gurus panel featuring Scot Wingo, Founder of ReFiBuy; Kelly Goetsch, President of Pipe17 and Founder of Commerce Ops Foundation; Sid Ariduri, VP of IT at 5.11 Tactical; and insights from Jason Goldberg, Chief Strategist at Publicis Groupe.

Join the Conversation

If you've got perspectives, challenges, or different takes on how agents will reshape your OMS, bring them. That's what this community is for.

Have you already implemented ACP or UCP? Are you supporting onX? Are you struggling with product data quality as you prepare for agentic channels? Share your experience. The conversations we have now become the battle scars and playbooks that the industry learns from.

The future of commerce is being written right now. Your OMS is the plot that matters most.

About the Order Management Gurus

The Order Management Gurus is a vendor-neutral community of 3,400+ product leaders, architects, technologists, and operations professionals discussing the real challenges reshaping order management, fulfillment, inventory, and commerce operations. The community is powered by Nextuple.