If you want to understand what UCP is—and what it isn’t— please don’t start with LinkedIn. Start with the developer documentation or the Github repository. A quick scan of Google’s UCP implementation guide makes the boundaries unusually clear: UCP standardises the exchange of commerce intent and capabilities; it does not magically replace checkout, payment operations, risk, returns, or customer identity. 

That matters because the sneaker example—the one everyone uses—is only “simple” on the surface. The same sneaker (same SKU, same GTIN) can be bought at Target, Zalando, adidas.com, hundreds of specialized sneaker stores, or a local shoe stores. The product is identical. The value creation logic is not. And UCP, by design, cannot neutralise that difference.

The UCP Reality Matrix

To separate narrative from operational reality, two distinctions matter more than anything else: 1) Can the agent execute checkout?; 2) Is the user meaningfully known to the merchant? 

This yields four possible realities:

Which leads to the boundary the protocol quietly enforces: UCP standardises the expression of intent. It does not transfer ownership of commerce.

What is not in the protocol—and why the omission is the point

UCP does not define how to:

• Verify a user’s identity
• Manage fraud and liability
• Execute payments beyond supported wallet credentials
• Handle returns, service, or post-purchase edge cases

The Merchant Center help documentation is explicit about the current payment credential model on Google’s UCP-powered checkout: it uses funding credentials stored in Google Wallet (with expansion over time), and merchants need Payment Service Provider (PSP) capability to accept the relevant payment tokens. 

That is not a footnote. It is the core reason why “UCP = checkout” is a category error. The protocol can standardise how intent is expressed and transmitted; the operational risk stays with the merchant of record—even when the checkout appears to happen on the platform surface.

UCP does not level the playing field. It formalises the interfaces through which intent can be routed—and that tends to advantage businesses whose value is already expressible as price, availability, and logistics. UCP can accelerate a decision. It cannot justify one.

And that is why, for Europe in particular, the first strategic question is rarely “How fast can we implement UCP?” It is: “Which quadrant can we actually get in—and which one could we credibly move to without buying into platform economics we don’t control?”

Once the technical boundaries of UCP are understood, the discussion inevitably shifts from engineering to economics. The decisive question is no longer what the protocol can do, but who benefits when it is widely adopted. Every commerce system, regardless of interface or technology, revolves around three structural roles: decision authority, control points, and risk allocation. 

These roles determine where value is created, where it is captured, and who ultimately carries responsibility when things go wrong. UCP does not change these roles. It redistributes them.

Decision Authority Moves Upstream

In traditional e-commerce, merchants shaped decisions through owned environments: product pages, merchandising logic, bundles, content, and experience design. The moment a customer entered the site, the merchant controlled framing, comparison, and persuasion.

Agentic interfaces invert this logic. When an AI agent curates options, compares prices, and contextualises recommendations before a user ever reaches a merchant, the decisive moment has already passed. The merchant may still process the order, but the decision architecture lives elsewhere.

This is not theoretical. It is the same dynamic that has governed Amazon’s Buy Box (featured offer) for years. Sellers compete fiercely on logistics, price, and compliance, while Amazon controls which offer is presented as the “right” choice. UCP generalises this mechanism beyond a single platform.

Control Points Remain Asymmetrically Distributed

UCP is often described as an “open” standard, but openness does not imply neutrality. Control points still cluster where scale, identity, and data converge.

• Discovery remains controlled by platforms with reach.
• Identity remains controlled by platforms with wallets and login systems.
• Advertising remains controlled by platforms that intermediate demand.
• Merchants can only integrate into this ecosystem; they cannot redefine it.

The sneaker example makes this tangible. A mass retailer like Walmart can expose price and availability in a way that is immediately machine-readable. A curated retailer exposes expertise, selection logic, and community—attributes that do not survive protocol abstraction. UCP does not discriminate; it simply reflects what can be standardised. The following 2 representative examples illustrate this dilemma:

Risk Never Leaves the Merchant

Crucially, while decision authority shifts upstream, risk does not. Payment failures, fraud, returns, customer service, regulatory compliance, and reputational fallout remain with the merchant of record. Even when checkout appears embedded or “agentic,” liability does not migrate. This asymmetry is easy to miss in early-stage narratives. It becomes painfully obvious at scale.

A Familiar Pattern: Open Standards First, Monetisation Later

This dynamic is not new. Google has repeatedly used open standards to lower adoption barriers and accelerate ecosystem alignment. Schema.org is the most prominent example. Introduced as a collaborative effort to structure the web, it simplified participation and improved interoperability. Over time, structured data became deeply entangled with ranking visibility, rich results, and advertising performance. Participation was never mandatory. 

Economically, it became unavoidable.

UCP follows the same logic. The protocol itself is intentionally non-monetary. It creates cleaner intent signals and reduces integration friction. The monetisation opportunity emerges one layer above—in prioritisation, advertising, and access to demand.

Adopting UCP does not automatically increase revenue. It can lower acquisition costs for some, erode differentiation for others, and leave many structurally unchanged. The outcome depends entirely on where a business sits relative to decision authority and control points. Understanding this is what separates strategic adoption from reactive compliance. And it leads directly to the next, unavoidable question: if the protocol itself is free, where is value actually monetised—and by whom?

One of the most misleading aspects of the UCP debate is the assumption that “commerce” is a single category. It is not. The economic impact of agentic commerce differs fundamentally depending on how value is created, captured, and defended.

Looking across real-world commerce setups, five recurring business models emerge. UCP interacts with each of them in materially different ways.

Mass Retail and Price-Driven Platforms

Mass retailers and horizontal platforms—such as Walmart, ASDA, Zalando, or The Home Depot—operate on scale, logistics efficiency, and price competitiveness. Products are highly comparable, margins are thin, and marketing spend is often significant.

If you are a mass retailer like Zalando, the logic is straightforward: you participate as soon as possible. You expose your catalogue, you let intent be routed, and you maximise the probability that demand lands with you rather than a competitor. Agentic surfaces simply become another high-volume entry point into an already-optimised machine. For these players, UCP is not a strategic gamble. It is a rational extension of an existing demand model.

Curated and Specialty Retail

Specialty retailers—often highly local, culturally embedded, and expertise-driven—derive value from curation, selection logic, and trust. Think of nakedcph.com, sneakersnstuff.com, or 43einhalb.com: the slightly higher price (if so) is justified by knowledge, assortment discipline, and community relevance. Here, UCP offers limited upside. While discovery may still happen at the agentic layer, the protocol has no way of expressing why this store is worth choosing over a cheaper alternative. At the same time, opting out entirely risks invisibility. This is the uncomfortable position many curated retailers find themselves in: participating does not help much, but not participating can hurt.

D2C and Configuration-Led Brands

Direct-to-consumer brands that already invest in owned experiences sit somewhere in between. If you are Victorinox (the makers of Swiss Army knife) or Signify (who bring you smart lights with Philips Hue), you may have built D2C capabilities to better understand your customers—even though the same products are still widely sold through retail and marketplaces, such as on Amazon. Where configuration is possible, the equation changes. 

Canyon (the world’s leading D2C bike brand) is the clearest example: the bicycle configurator for their flagship product, the Canyon Aeroad, is not a nice-to-have feature, it is the core of the buying experience. Even Victorinox offers knife customisation that shifts the decision away from pure price comparison. In these cases, UCP can support discovery, but it does not replace the moment where confidence is built. The more configuration and guidance matter, the less relevant agentic checkout becomes.

Luxury and High-Consideration Commerce

Luxury exposes the practical limits of agentic commerce. High price points, emotional value, service expectations, and payment constraints all resist full automation. Anyone who has ever tried to implement credit card split payments in a checkout knows exactly where theory meets reality.

This applies not only to high-end watch retailers like Bucherer (part of Rolex), Breitling, or Chopard but also to so-called “entry luxury” brands such as Pandora, where personalisation and gifting logic already blur the line between configuration and luxury experience. In this segment, efficiency is not the problem to be solved. Trust, reassurance, and human interaction are part of the value proposition.

B2B and Relationship-Based Commerce

In B2B, commerce follows a fundamentally different logic. Pricing is negotiated, availability is contextual, and transactions are embedded in long-term relationships, contracts, and service levels. It is highly unlikely that aircraft spare parts at Airbus-Satair, or precision components from industrial suppliers like Arrow, will be meaningfully transacted through agentic checkout flows any time soon.

Here, intent signals may be useful—but the transaction itself remains deeply tied to account management, compliance, and operational coordination. In this world, agentic commerce is a peripheral interface, not a structural shift. To summarize: Different types of business have a different fit to UCP.

The conclusion remains unchanged: UCP does not create winners. It amplifies the economics that are already in place.

Most of the current excitement around UCP is implicitly US-centric. High wallet penetration, consolidated platforms, and comparatively permissive regulatory environments make agentic checkout experiments feasible at scale. If a user in the US is logged into a platform wallet, has a stored payment method, and accepts platform-mediated flows, a “buy inside the conversation” experience is at least technically plausible (as displayed in the illustration above).

Europe, however, operates under fundamentally different constraints. Europe’s consumer protection law alone introduces friction that agentic checkout narratives rarely acknowledge: mandatory withdrawal rights (typically 14 days), explicit disclosure obligations, regulated returns handling, warranty and liability requirements, and strict rules around who is considered the merchant of record. In practice, this means that a fully embedded checkout inside a third-party agent would require clearly defined contractual frameworks between merchant and platform to handle liability, refunds, guarantees, and customer service responsibilities. These frameworks do not exist at scale today.

Payment infrastructure reinforces this difference. Wallet penetration in Europe is fragmented, with strong local schemes, bank-based payments, and regulatory requirements such as Strong Customer Authentication (SCA) under PSD2. This makes the frictionless, one-click-style flows common in US demos significantly harder to reproduce consistently across markets. Even where wallet usage is growing, compliance requirements often force explicit user interaction at the moment of payment—precisely the step agentic checkout aims to minimise.

The result is a structural mismatch between global narratives and regional reality. The fact that a sneaker can be purchased via an agentic flow at Dick’s Sporting Goods in the US does not imply that a similar experience is available—or even desirable—at Zalando, asos, or JD Sports in Europe. Geography is not an implementation detail. It is a strategic boundary condition. Ignoring it leads to premature investments and misplaced urgency.

One of the loudest claims around agentic commerce is that owned digital experiences no longer matter (a.k.a. the good old “The website is dead.”). If discovery, comparison, and even checkout happen inside AI interfaces, why invest further in websites, content, or service design?

The answer becomes obvious once customer experience is understood correctly. Customer experience is not a page, a funnel, or a UI. Customer experience is the sum of all touchpoints a customer has with a brand. UCP touches exactly one of those touchpoints—the handoff from intent to transaction. To make this tangible, it helps to look at what great customer experience actually means across different business models.

Configurable D2C (e.g. Canyon):

The experience that matters happens before checkout, mainly in the discovery and evaluation phase of the customer journey. Bike configuration, geometry selection, component trade-offs, and delivery timing cannot be reduced to an agentic shortcut. The value lies in exploration and control, not speed.

D2C with Personalisation (e.g. Victorinox):

Product customization—engraving, bundles, gifting—is the experience. A fast checkout is useful, but differentiation comes from making the product personal and meaningful.

Luxury Retail (e.g. Bucherer):

Experience happens mainly in discovery and post-checkout. The business is all about reassurance and trust. Certified pre-owned watches next to new collections, provenance, long-term value, and concierge-style service matter more than transactional efficiency. A €15,000 to €50,000 purchase rarely fails because checkout is slow; it fails because confidence is missing.

Specialty and Curated Retailers (e.g. niche sneaker stores):

Curation is the experience, including discovery and evaluation, but also loyalty and community. Selection, storytelling, and context justify price differences. A purely price-optimizing agent will not replicate that value.

B2B Spare Parts and Industrial Commerce (e.g. Satair–Airbus):

Experience means understanding complex specifications, availability constraints, certifications, and contractual terms. This is not impulse commerce; it is risk-managed procurement.

Across all these examples, UCP does not replace experience. It intersects with it at one narrow point.

UCP extends the system at the moment of transaction, not at the moments where brands differentiate, justify their margins, or build loyalty. This distinction matters economically. In mass retail, Google’s AI mode/ChatGPT plus UCP/ACP may deliver an incremental customer at low acquisition cost. In those scenarios the “small 4% transaction fee” for instant checkouts using ACP, announced by OpenAI and Shopify, can be rational. But whether that customer returns, buys more, or becomes loyal is still determined elsewhere. In simple terms: Efficiency can win a purchase. Experience decides whether it was worth it.

That is why, for most brands and manufacturers, the strategic question is not whether UCP exists, but where experience must remain dominant, and where transactional efficiency is acceptable.

If one actor clearly benefits from the emergence of UCP and agentic commerce, it is not merchants. It is platforms—most notably Google. UCP is often discussed as a commerce innovation. In reality, it is a media infrastructure upgrade. The decisive asset is not the transaction itself, but the pre-qualified purchase intent generated upstream in conversational environments. If a platform already knows what a user wants, when, at what price sensitivity, and with which alternatives considered, the economic value of that intent far exceeds the margin of the product being sold.

This is where the narrative shifts. OpenAI’s early move to test a 4% transaction-based commission is not radical—it is conservative. For a mass retailer like Walmart or Target, 4% is often cheaper than customer acquisition costs via traditional paid search or retail media networks. Losing 4% margin is acceptable if it replaces paid search bids, affiliate fees, and brand spend. For a small, curated sneaker retailer, the same 4% can be existential.

Google’s position is even stronger. With AI Mode embedded into search, UCP becomes a new signal layer for intent monetisation. The logical next step is not hard to predict: sponsored recommendations, paid priority placements, or dynamic bidding on agentic intent. This mirrors Google’s historical playbook—lower friction through open standards first, monetise once adoption is locked in.

What looks like “open commerce” is therefore the foundation of Advertising Model 2.0. Intent moves upstream. Competition intensifies earlier. Media spend shifts from keywords to conversations. And the platforms that own these conversations become the new toll gates. 

In that context, UCP does not primarily redistribute commerce value—it redistributes media power. And history suggests who tends to win those redistributions.

After all the noise, protocols, pilots, and hot takes, the real question is not whether Universal Commerce Protocol will exist, but when it is rational for a business to engage with it. The following decision tree is not a prediction tool. It is a strategic filter. It helps distinguish between companies that should act now, those that should prepare, and those that are better off doing nothing on purpose. Read it left to right. Answer honestly. If the outcome feels uncomfortable, that’s usually a signal, not a flaw.

For most European merchants and manufacturers, the correct short-term strategy is not acceleration, but restraint.

UCP does not remove the need for differentiation. It amplifies the cost of being comparable. If your product competes primarily on price, scale, or availability, agentic commerce may eventually benefit you. If your value lies in curation, configuration, service, expertise, or trust, premature participation risks collapsing your advantage into a line item. This is visible across categories:

• Canyon wins because customers configure bikes—not because they find the cheapest SKU fastest.
• Victorinox benefits from personalisation, engraving, and brand trust—not instant checkout.
• Bucherer sells confidence, certification, and experience—no one impulse-buys a €25,000 watch via chat, not least because many credit cards wouldn’t allow it.
• Pandora sits between worlds: personalisable products, mid-range price points, emotional gifting.
• Specialty sneaker retailers survive through curation and community—not price parity with Zalando.

In these cases, UCP is not a growth lever today. It is a context signal at best. European companies should therefore prioritise visibility without surrendering checkout. From this POV focussing on intent literacy (GEO / AEO) wins over transactional integration (doing a technical UCP project in 2026). Additionally, experience systems that reduce comparability through bundles, services, configuration, or expertise seem to be a safe bet. And finally, waiting is not a failure of innovation; it is often a sign of strategic maturity.

UCP will not eliminate websites. It will not replace customer experience. It will not democratically benefit all merchants.

What it will do is move competition earlier, shift power toward platforms, and rewire how intent is monetised. The biggest structural change is not in commerce flows, but in media economics.

For US mass retail, participation is rational. For platforms, it is strategically brilliant. For European merchants, the smartest move today is often to observe, learn, and invest elsewhere.

Agentic commerce is real. UCP is relevant. But the winners will not be those who integrate first—they will be those who understand where not to compete.