Amazon & Meta Join the UCP Council: Here Is What That Means for Your Brand

Five names changed agentic commerce on Friday. Amazon. Meta. Microsoft. Salesforce. Stripe. All five took seats on the Universal Commerce Protocol Technical Council, joining Google, Shopify, Etsy, Target and Wayfair around a single governance table.

That table now decides how AI agents talk to retailers, how baskets get built, how payments clear, and how returns get processed across every meaningful surface in commerce. The implications for brands are immediate.

The Quick Picture of UCP

Think of UCP as the common language AI agents and retailers use to do business with each other. Launched by Google early this year, it has emerged as the dominant protocol for the era of agentic commerce, pulling decisively ahead of the Agentic Commerce Protocol that OpenAI launched a few months prior.

The mechanic is simple. An agent asks a retailer what it can do, sees the catalogue, builds a basket, processes the payment and handles the return. All of it through one shared standard rather than a hundred different bespoke integrations.

Under the hood it covers the things that actually matter: how agents discover what a retailer supports, how baskets and orders get managed, how payments clear, and how identity, loyalty, local inventory and returns get handled. New extensions are being added every month.

The latest release shipped over 60 updates, the biggest yet. Retailers and AI surfaces are already building against this infrastructure today.

Why Amazon Joining Matters Most

Amazon spent the back half of 2025 and the start of 2026 running the opposite playbook. They blocked third-party AI crawlers. They updated seller terms to force agent identification. They took Perplexity to court over Comet's automated checkouts. They poured engineering resources into Rufus, Buy For Me and Shop Direct as Amazon-owned, Amazon-controlled, Amazon-only surfaces.

The message was clear. Agentic commerce was happening on Amazon's terms or not at all.

Then last Friday, Amazon walked into the room and took a seat on the council that governs everyone else's terms.

There are two ways to read this, and we think both are partially correct.

The pragmatic read

UCP was going to ship with or without Amazon. Once Target and Wayfair signed on, every other major retailer faced a forcing function. A protocol with Google's distribution behind it, Shopify's merchant base inside it, and Stripe's payments infrastructure powering it does not need Amazon to succeed. Amazon staying outside meant accepting whatever rules got written by the people inside. Joining gives them voting weight on payment handlers, on identity standards, on the capability surface every retailer has to expose. It is a defensive move dressed as cooperation.

The strategic read

Amazon's flywheel runs on catalogue breadth. The single biggest threat to Amazon is a world where shoppers can access non-Amazon inventory through a single conversational interface that feels as good as Amazon's own. UCP makes that world possible. By sitting inside UCP rather than outside it, Amazon gets to influence what that interface can and cannot do. They also get the option, down the line, to consume UCP inventory into Amazon's own surfaces while keeping their first-party data protected from competing agents. Read that way, joining looks less like a concession and more like calculated positioning.

Rufus as a horizontal agent

The strategic read leads to a more interesting possibility. With Amazon now able to ingest the wider UCP catalogue, Rufus could become a surface through which every UCP-enabled retailer's inventory is shoppable. Rufus today only transacts against Amazon's own catalogue, plus Buy For Me and Shop Direct. There is nothing technically stopping it from extending across the full UCP surface.

Rufus is uniquely positioned to win that race. It already has the largest installed base of any agentic shopping assistant. It just launched agentic features including automated re-ordering and personalised shopping guides. Plug those capabilities into UCP-mediated catalogues, and Rufus becomes the most powerful horizontal commerce agent in the market.

➡️ If you want to go deeper on Rufus and how to optimise for it, our whitepaper with the Digital Shelf Institute is the most thorough guide currently available.

Why Meta's adoption is incredibly significant

Meta's agentic commerce story until now has been a frustrated one. Four billion users across WhatsApp, Instagram and Facebook. A persistent inability to convert that scale into transactional revenue at the level the audience would suggest. Shop tabs, in-app checkout experiments, ad-funded discovery loops. None of it has produced the agentic shopping experience the platforms are clearly built for.

UCP gives Meta the missing piece. Standardised access to retailer catalogues. Standardised checkout. Standardised post-purchase flows. Plug Meta AI into UCP and the entire commerce graph becomes addressable from inside a WhatsApp conversation, an Instagram DM, a Messenger thread.

Meta also had a foot in OpenAI's ACP camp until last week. Their move to UCP is a directional vote. With Stripe handling payments and Microsoft bringing Copilot into every workplace, the coalition now has every category of player it needs.

What You Can Do to Prepare for This New Reality

Step back and the picture is clear. Amazon, Meta, Microsoft, Salesforce and Stripe joining UCP means agentic commerce is about to scale across every surface a shopper actually uses. The Amazon app. WhatsApp. Instagram. ChatGPT. Gemini. Copilot. Every one of them, querying retailer catalogues through the same protocol, applying the same logic to recommend, compare and transact.

That changes what a winning PDP looks like. Persuasive copy, lifestyle imagery and brand storytelling continue to do useful work for the human shopper who clicks through. For a UCP-compliant agent making a recommendation, none of it lands. Agents read structured signals. They reward clarity, completeness and consistency. They penalise ambiguity ruthlessly.

Here are five concrete actions for any brand selling on marketplaces or DTC.

1. Treat backend attributes as crucial to your listings

Most brands fill the required attribute fields and ignore the rest. That is a discoverability tax. UCP will use backend attributes to construct the relational graphs that agents query against. Function, intended audience, occasion, location of use, dietary or material specifications, complementary products. Every empty field is a missing edge in the graph. The brands winning agentic search right now treat backend attributes with the same priority they used to reserve for hero images.

2. Build one canonical product record across every channel

UCP is going to expose every inconsistency in your data. If your Amazon listing says one pack size and your Walmart listing says another, agents querying both will deprioritise you. The brands that win agentic commerce will be the ones running a single source of truth that feeds Amazon, Walmart, Target, Shopify and direct-to-consumer surfaces with identical canonical attributes

3. Convert spec dumps into feature-to-outcome copy

"Padded shoulder straps" tells an agent nothing about who this product is for. "Padded shoulder straps that distribute weight for commuters carrying laptops over 2kg" gives the agent everything it needs to recommend confidently. Product descriptions should connect a concrete feature to a defined customer outcome and, where possible, to a specific use scenario.

4. Plan for capability-aware listings

UCP supports retailer-specific capabilities like subscription checkout, real-time inventory and complex variation handling. Make sure your product data exposes the attributes those capabilities depend on. Subscription-eligible products need replenishment frequency. Variations need clear option types. Future-proof your listings now, before agents start filtering for capabilities your data does not declare.

5. Get checkout-readiness signals that an agent can trust

Agentic checkout systems do not retry. If an agent cannot confidently execute, it picks the next product. Pricing must be stable. Stock must be accurate. Variants must be cleanly differentiated. Fulfilment options must be explicit. Claims must be consistent between title, bullets, attributes and imagery. Three places on a PDP saying three slightly different things about pack size will lose you the transaction every time.

Beyond UCP: The Agentic Merchant Protocol

UCP solves the retailer-to-agent connection. It standardises how Amazon, Walmart, Target, Shopify and the rest expose their catalogues to AI agents. It does not solve the brand problem.

The brand problem is this. Your product appears across hundreds of surfaces. Retailer listings. Off-site reviews. Editorial content. AI training data. Conversational agents on every platform. Each surface holds a different version of your product story, with different attributes, different claims, different imagery. UCP will accelerate this fragmentation, not contain it. Without a coordination layer, every agent ends up reasoning over a slightly different version of your brand.

That is why we built the Agentic Merchant Protocol (AMP).

AMP sits above UCP and ACP. It gives enterprise brands a single system to define, distribute and govern how their product catalogues are understood by AI agents across the entire commerce ecosystem. One canonical source of truth for every product. Programmatic distribution of that intelligence across the open web and agent ecosystems where AI reasons and decides. Enterprise-grade brand governance over messaging, compliance and product representation on every AI surface. Agent-agnostic by design, so you stay independent of any single platform.

Early adopters include L'Oréal, Unilever, Mars, Beiersdorf and Reckitt. These are the brands that have decided not to leave their representation in the agent ecosystem to chance.

If the past 12 months of UCP, ACP and the rest of the agentic commerce stack have made one thing obvious, it is that the protocol layer is moving fast and the brands without a coordination strategy are going to be reasoned over in ways they cannot influence. AMP is how you take that control back. Apply to join the early adopter programme here.

The Through Line

UCP is the moment agentic commerce stopped being a feature each platform could implement differently and became infrastructure every platform shares. That is good for shoppers. It is good for retailers who do the data work. It is going to be brutal for brands that have treated the digital shelf as a creative exercise.

The brands ready for this shift have already turned their PDPs into machine-readable assets. Clean attributes. Conversational copy. Consistent data across every channel. Multimodal content that holds up under OCR. They will see compounding visibility across every UCP-connected surface as the protocol rolls out.

The brands not ready are about to find out that the rules of agentic discoverability got written while they were optimising for last year's search algorithm.