UCP is a communication standard designed to allow AI agents to search for, evaluate, and purchase products and services by querying APIs, structured data, and a company’s business rules directly.
Unlike a marketplace or an e-commerce platform, it doesn’t introduce a new sales channel. Instead, it defines how commerce systems interact with AI—enabling machine-to-machine transactions in the era of the agentic economy.
In B2B, this is a fundamental shift: buyers’ AIs won’t browse web pages. They will analyse data models, contractual terms, and automatable processes to make decisions and trigger orders.
Unveiled by Sundar Pichai at NRF 2026 in New York, the Universal Commerce Protocol was created to make commerce globally interoperable in the age of intelligent agents.
Developed with partners such as Shopify, Etsy, Wayfair, Target, and Walmart, it is compatible with leading agent protocols and payment systems.
Like HTTP for the web, UCP establishes a shared standard. Its goal is to make products and services purchasable directly within AI-driven conversational experiences—through standardised, machine-readable integrations.
This protocol can apply to B2B too – with important differences – because business-to-business transactions are business-critical, recurring, highly complex, and governed by customer-specific terms. What’s certain is that more and more B2B buyers will use intelligent assistants to shortlist suppliers, compare offers, and manage automated reordering. The implication is clear: if you’re not readable by AI, you won’t make it into the decision-making process.
The risk isn’t selling less because the market changes- it’s not being evaluated at all by the AI tools buyers use. It’s a silent risk: it won’t show up in your KPIs, it won’t trigger complaints, but it will erode market share over time. A buyer won’t call to tell you their AI is excluding you because your product data isn’t machine-readable. They’ll simply buy elsewhere.
The winners will be the companies able to expose data, rules, and processes in a structured, interoperable, and governable way. UCP is the enabling infrastructure for this scenario, while integrated, process-oriented platforms become the strategic asset to stay relevant, chosen, and competitive in the age of the agentic economy.
Today, B2B companies invest in e-commerce portals, customer areas, and apps. These tools are human-facing interfaces and will remain essential assets for managing human-to-human commercial relationships. The difference is that these interfaces won’t be used only by people anymore.
AI doesn’t browse pages: it queries processes, interprets relationships, and scans data tables to find the information it needs to achieve a goal.
Interface usability will still be necessary – but it will no longer be enough to ensure process effectiveness. It must be matched by strong data-model design and architectural quality. In other words: how easy is it for an AI to understand your offer and buy it?
In this new landscape, the winners are those who expose, clearly and reliably, complete and standardised product data, codified commercial terms, explicit sales rules, and order processes that can be automated.
This means that PIM, CRM and B2B e-commerce systems are no longer just essential operational tools – they become strategic assets for staying relevant in an agentic economy.
A PIM becomes the place where attributes, variants, certifications and logistics constraints are made machine-readable. The critical constraint is simple: if product data isn’t structured, AI won’t consider it.
In B2B, CRM has traditionally been the memory of relationships. With UCP, it also becomes the memory of rules, exceptions and customer priorities. The constraint here is just as clear: AI buys by applying procedures that must be defined upstream, codified, and applied to properly segmented customers.
Sales teams won’t disappear, but they will evolve: less order entry and manual follow-ups, more rule configuration and exception management. The shift in perspective is substantial: value no longer lies in manual execution, but in the governance of automated processes.
In this scenario, .one comes with a structural advantage. As an integrated platform that brings together CRM, PIM, B2B e-commerce, processes, data and rules within a single architecture, it enables unified and structured data, codifiable business logic, end-to-end traceable B2B processes, native integration across product, customer and sales, and APIs designed for interoperability.
The .one architecture was designed around 3 guiding principles:
These three characteristics make .one UCP-ready because they go straight to the heart of what matters when the “buyer” becomes an AI agent: integration, machine-readability, and API-driven execution.
A broad, flexible API library means an agent can not only read data, but also perform concrete actions (transactions, CRUD operations, reporting) without relying on human-facing interfaces.
JSON:API compliance adds a decisive advantage: data and relationships are exposed in a standard, predictable way, with consistent error handling—reducing ambiguity and making automations, filtering, pagination and queries more reliable.
Finally, an OpenAPI foundation makes the system discoverable and self-describing: an agent (or a developer) can quickly understand what it can do, how to do it, and under which constraints – speeding up integration and reducing friction.
To be truly “UCP-ready”, APIs and an excellent data model aren’t enough. Data and commercial rules must also be semantically clear and machine-executable (customer terms, policies, logistics constraints, priorities, exceptions), and there must be a layer of higher-level actions that turns 1,000 endpoints into simple operations for an agent (e.g. “create order”, “apply contract”, “check availability”).
This is also part of what we do: guiding customers in designing and implementing the right solution, supported by an in-house delivery team that understands segment- and market-specific commercial best practices, and is involved from the very first phase of analysis and functional requirements definition.
UCP doesn’t describe a distant future – it describes how B2B purchasing decisions will be made over the next few years. Companies investing today in processes, data and integrated platforms are building the infrastructure that will make them visible and competitive in a market where data clarity matters as much as the quality of the offer.
