Building the Operational Backbone for Agentic Commerce
A new industry standard aims to give commerce systems, AI agents, and fulfillment networks a common way to move orders with accuracy and speed.
Artificial intelligence has made it remarkably simple for customers to shop wherever they choose. A discovery moment might start on a brand site, continue in a marketplace, surface in a social feed, or unfold inside a conversational assistant. But while the front end of commerce has diversified and accelerated, the systems behind it have not.
There are more ways than ever for customers to engage with a brand and place an order, but once received, those orders still have to move through fulfillment networks that depend on one-off integrations, brittle connections, and outdated logic. And as networks expand, the strain grows.
This widening gap is what pushed Kelly Goetsch, President of Pipe17 and President of the Commerce Operations Foundation, to act. He saw an industry that had invented new ways to create demand, but not new ways to carry that demand through the operational maze that turns an order into a delivered package. His conclusion was clear: commerce needed a shared operational standard.
The result is the Order Network eXchange. It launched on Tuesday under the Commerce Operations Foundation, a nonprofit created to govern the standard and keep it neutral.The announcement brings together a wide cross-section of the commerce ecosystem alongside Pipe17, including OMS leaders like Manhattan Associates and IBM Sterling, 3PL and logistics providers such as Radial and Ryder, composable commerce platforms like commercetools and system integrators like Orium, and major retailers and brands including Allbirds and The Vitamin Shoppe. Together, these organizations represent more than a trillion dollars in annual GMV.
For an industry long reliant on proprietary integrations, the level of early alignment signals a pivotal shift toward a more connected future.
A Growing Landscape Without Shared Language
Goetsch entered the order management space as the number of selling channels was expanding. “I saw more selling channels than we’ve ever had before,” he said. “Historically you had your brand dotcom. Now you’ve got social channels, marketplaces, and you’re starting to see agentic commerce rise with AI selling channels. At the same time, the fulfillment side has exploded with more 3PLs, more WMSs, more ERPs. I looked around and realized there was no unified network or standard connecting any of it.”
A connection between an OMS and a 3PL is often written as a one-off integration that requires attention every time a process changes— what Goetsch described as “hand-welded solutions”. With a single channel, this has been a manageable nuisance. As channels proliferate and introduce their own integration needs, the burden quickly becomes untenable.
The absence of a shared model means simple activities—such as inventory checks—often pass through several layers of systems before reaching the source of truth. In a perfect world, according to Goetsch, an AI selling channel would connect directly with a fulfillment provider to retrieve accurate data rather than relying on a chain of copies. But until now, there has been no way to create that direct connection.
That gap led Goetsch to gather more than sixty companies across OMS, WMS, ERP, commerce platforms, and logistics with the goal of aligning on a common interface that any system could adopt.
A Simple Interface for a Complex Network
At its core, Order Network eXchange offers a consistent interface that defines how systems communicate about orders, inventory, and fulfillment.
“We’re agreeing to adopt the Order Network eXchange protocol, which defines sixteen tools and nine resources” Goetsch said of the 60+ organizations that have signed on. “The tools are the verbs, the things you can do. The resources are static copies of key data. We’re saying, ‘This is the interface. This is the front door for our systems,’ and we’re going to roll it out to our collective thousands of customers.”
The consistency of that interface means a commerce platform, marketplace, or AI agent can connect directly to fulfillment systems without new custom work each time. It supports real-time communication across the network and reduces the complexity required to bring new channels online.
“Imagine going from ChatGPT to a 3PL directly,” he added. “Right now that doesn’t exist. This makes it possible.”
Why This Standard Is Possible Now
Goetsch isn’t the first to see this gap. Several earlier attempts to create unified models for retail and commerce struggled because of implementation challenges. What makes this moment different is the arrival of the Model Context Protocol, or MCP.
MCP gives AI systems a structured, simple way to interact with servers and tools, and removes many of the challenges associated with traditional API approaches.
“We’ve never had something like MCP before,” explained Goetsch. “It’s uniquely enabling for this kind of technology. Historically you had to do this with REST APIs, and doing order-related data with REST is doable, but hard. We’ve seen standardized retail models attempted before, but they failed because they were too onerous. What I love about MCP is that it’s AI-native. You publish, ‘This is my server, these are my tools, these are my resources,’ and the AI figures it out.”
As more selling channels incorporate AI agents, the need for a clean way to route orders becomes more important. MCP provides a direct path for that communication, which makes a standard like Order Network eXchange feasible today.
A Foundation Designed for Neutrality and Longevity
To support the standard, the Commerce Operations Foundation was formed as an independent nonprofit. It holds the trademarks, patents, and copyrights related to onX and provides long-term oversight.
“Somebody has to oversee it,” Goetsch said. “And that somebody can’t be a random collection of people meeting once a quarter. There’s IP involved, so you need vendor neutrality.” He compared it to how the Cloud Native Computing Foundation oversees Kubernetes, providing the level of governance required for real adoption.
Those familiar with Goetsch’s history may wonder whether this is meant to echo the MACH Alliance. It is not. The Foundation represents the community, but its scope is narrower. It exists as a standards body, focused on supporting and maintaining one technical specification.
Adoption and Early Momentum
Although it only launched this week, early signs point to strong adoption. Support for the standard already includes major vendors across logistics, retail, and commerce technology.
Part of the reason for this early momentum is that many of these companies are already preparing to implement MCP, which makes the Order Network eXchange a natural extension of work already underway. When the finalized version of the standard—currently expected in early January—goes live, we should see a swift rollout among vendors.
“We’re largely there,” Goetsch said. “We wanted this to be a community thing. People need something they can adopt. They need it to be reviewed, stress-tested, and signed off by others, not just handed over.”
That shared ownership increases the likelihood of adoption and provides a clearer path for vendors already planning to support AI-driven workflows.
A More Connected Future
Goetsch sees the standard enabling a future where adding new selling channels becomes far easier. Today, each channel requires careful coordination around inventory, order capture, and fulfillment. Order Network eXchange simplifies that work by providing consistent expectations and processes.
“I think it enables the proliferation of selling channels,” he said. “Right now, every time you add a channel, you have to figure out what to do with inventory and fulfillment. There’s a lot of back and forth. What’s great about this is that it standardizes that. You don’t have to write a custom driver for each one. You just plug it in.”
If adoption continues as expected, Order Network eXchange could become a foundational part of agentic commerce, helping AI-driven selling channels connect directly into the physical world of fulfillment and logistics. Brands would gain a cleaner operational model. Consumers would get more consistent experiences. And the industry would have a standard built to support the next wave of commerce innovation.
Leigh Bryant
Editorial Director, Composable.com
Leigh Bryant is a seasoned content and brand strategist with over a decade of experience in digital storytelling. Starting in retail before shifting to the technology space, she has spent the past ten years crafting compelling narratives as a writer, editor, and strategist.