Agentic Commerce & AI Checkout: What Google & Shopify’s UCP Means For Ecommerce Retailers

AI-led discovery is already changing how shoppers find products – with recommendations increasingly being surfaced directly within AI experiences. We explored this in a detailed recent article on how discovery is happening before the click, exploring what this means for performance marketing in particular.

Now, in a move that could shift the customer journey even further upstream, Google and Shopify are laying the groundwork for customers to buy inside AI experiences too – removing the traditional ‘click through to site → browse → add to cart → checkout’ flow.

At the centre of this is the launch of the Universal Commerce Protocol (UCP) – a new open standard to support agentic commerce – alongside a range of supporting AI tools designed to help retailers connect with high-intent shoppers and drive sales.

In this article, we’ll cover:

• what the UCP is and what it enables;
• the other agentic commerce tech announced by Google and Shopify;
• practical use cases for retailers;
• what this could mean for ecommerce brands as AI-led checkout becomes more viable; and
• how to prepare effectively, without overreacting.

What is the UCP?

In January 2026, Google and Shopify, in collaboration with an esteemed ecosystem of commerce platforms, retailers and payment partners, announced the launch of the UCP – representing a significant evolution in agentic commerce.

At its core, the UCP is a new open standard that provides a common ‘commerce language’ to enable native commerce across all major AI channels. With this new publicly available technology, AI agents will essentially be able to connect and transact with any retailer, representing checkout flows in any platform.

In practice, that means an AI agent could, where enabled, guide a shopper from search to purchase within the same conversational interface – handling the operational details that typically introduce friction. This includes actions such as applying discount codes, recognising loyalty credentials, supporting subscriptions and confirming delivery options, without forcing the shopper to jump between multiple pages and hand-offs.

Importantly, this doesn’t change the fundamentals of retail operations. The retailer remains the seller of record, with orders being created, managed and fulfilled through the brand’s ecommerce platform. However, the transaction takes place without the customer visiting a traditional frontend website.

The UCP will power native shopping on Google surfaces – enabling direct selling in AI Mode in Google Search and the Gemini app.  And for Shopify retailers, a new embedded checkout experience is also being rolled out to Shopify’s Microsoft Copilot integration thanks to the UCP. AI channels are managed directly from Agentic Storefronts in the Shopify Admin, with the ability to toggle selling on and off, and even exclude specific products from direct selling per channel.

What other agentic commerce tools is Google launching?

Alongside the UCP, Google also announced several additional initiatives designed to help retailers show up more effectively in AI-led shopping journeys – and, where appropriate, reduce friction between discovery and purchase.

Business Agent

Consider this a virtual sales associate that can answer shoppers’ questions, in a brand’s voice, within Search. It will enable retailers to connect with consumers during critical moments in the customer journey, helping to reduce hesitation and encourage sales.

In time, Google’s direction suggests richer capabilities too: training the agent on data, generating customer insights, surfacing offers for related products, and supporting transactional flows as agentic checkout matures.

New data attributes in Merchant Center

Designed to help retailers optimise conversational discovery across surfaces like AI Mode, Gemini and Business Agent, Google will be rolling out new Merchant Center data attributes to complement existing data feeds. These will go beyond traditional keywords, including richer product context such as frequently asked product questions, compatible accessories, and substitutes/alternatives.

For retailers, this puts even greater emphasis on structured product knowledge. The more clearly you can define what a product is, who it’s for and what it’s compatible with, the easier it becomes for AI experiences to recommend it confidently.

Direct Offers

With this new Google Ads pilot, advertisers will be able to present exclusive, context-aware offers (like discounts, bundles and free shipping), directly in AI Mode for shoppers who appear close to purchase.

Google will leverage AI to determine when an offer is relevant to display, which should help brands use incentives more efficiently.

Agentic plan: Agentic commerce for non-Shopify brands

As part of its ‘Agentic plan’, Shopify has also announced the opening up of the Shopify Catalog to non-Shopify brands – meaning retailers on any platform can now use Shopify’s infrastructure to sell on AI channels without needing to have a Shopify online store. Catalog uses specialised LLMs to standardise, structure and enrich product data, helping products to get discovered across AI platforms.

Rather than managing platform-by-platform integrations, brands simply add products to Shopify Catalog once, and these are syndicated to Google, Copilot, ChatGPT, Shop app and every future AI partner.

This is an important development for retailers who are not currently using Shopify and who are concerned about falling behind the pack. In some instances this will likely serve as a first use case for Shopify, ahead of more substantial commerce replatforming projects. We anticipate that this will be of particular appeal to larger retailers whose enterprise architecture and scale cannot support a rapid platform migration.

Agentic commerce use cases: Where might there be value for brands?

Here are some examples of practical use cases for this new agentic commerce tech that we’re discussing in our team right now.

1. UCP for high-intent buy-in-chat journeys

Scenario: A customer asks an AI surface: “What’s the best night-time moisturiser that’s suitable for sensitive skin and can be delivered by Saturday?”

With the UCP in place, the AI agent can move beyond recommendations and instead support the full transaction – confirming product suitability, checking availability and delivery promises, and then completing checkout within the AI experience. Rather than linking the shopper out to the brand’s ecommerce site, the journey becomes simpler and more decisive.

Where it brings most value: For high-intent, time-sensitive searches and repeat purchase categories where shoppers want the fastest route from confidence to checkout – for example, replenishment skincare, personal care and household staples.

2. Direct Offers for competitive categories

Scenario: Within Google AI Mode, a shopper asks: “I’m looking for a colourful, lightweight running shoe designed for overpronation. I mainly do road-running, so I want something that’s suitable for that sort of terrain. I’m a UK size 10.”

Google can already narrow this down into a shortlist that matches the shopper’s criteria. Direct Offers adds another lever at the decision point, with eligible retailers able to surface a targeted incentive within AI Mode, directly alongside the recommendation. Done well, this can help a shopper to choose between otherwise similar options without forcing an extra click, search or code-hunt.

Where it brings most value: In highly competitive categories where products are often comparable on features and price, and shoppers naturally evaluate several options before committing – for example footwear, beauty, consumer electronics accessories and gifting-led ranges.

3. Business Agent for recommending the right variant

Scenario: A shopper searches for a makeup brand on Google, activates the AI Chat experience in Search and asks for assistance choosing a foundation.

The agent can act like a digital beauty advisor, asking a small set of clarifying questions (for example, around skin type, undertone, preferred coverage and budget) and then recommending the most suitable shade and formula. Crucially, it can reduce guesswork at the moment it typically creeps in, helping shoppers feel confident they’re selecting the right option before they click through or buy.

Where it brings most value: In categories with high variant complexity and low tolerance for getting it wrong – where customers would benefit from guided decision-making that mimics in-store support. As well as beauty, other categories include apparel, footwear, supplements, furniture, and any range where choice overload contributes to product page drop-off and abandoned baskets.

What does this mean for online retailers & how can you prepare for AI checkout?

It’s important to remember that this isn’t about AI replacing ecommerce storefronts. Your site remains the centre of your brand experience, the place where merchandising is richest, storytelling is strongest, and a significant portion of customers will continue to browse, validate and convert in the traditional fashion.

Also, several of the new launches are being positioned as US-first or in pilot, which means they won’t be available to all retailers immediately, particularly in Europe.

Google and Shopify’s latest move, however, is a strong signal that transactional surfaces are expanding. Conversational interfaces are increasingly becoming places to both discover and buy from brands.

Over time, we should therefore expect more journeys where a customer:

1. discovers products within an AI experience;
2. narrows choices in conversation; then
3. completes checkout without ever clicking through to a product page.

The importance of commerce hygiene

For retailers, this means that the fundamentals that typically make your store convert (like data quality, clear policies, strong pricing logic and reliable fulfilment) need to be portable into these new surfaces. Rather than overhauling everything to be prepared for AI checkout, commerce hygiene will be key. This includes:

• Catalogue and data quality
  • Consistent product titles, attributes and variant logic
  • Clear taxonomy (collections, categories and filters that reflect how customers shop)
  • Content that answers real objections (around materials, sizing, compatibility, usage, care etc.)

• Inventory, fulfilment and promise accuracy
  • Stock accuracy that prevents “recommended but unavailable” moments
  • Lead times that reflect real operational capacity
  • Clear regional logic (where you ship, what you can’t ship, and why)

• Pricing and promotion governance
  • Fewer, clearer offers with documented eligibility rules
  • Pricing logic that remains consistent across channels and surfaces

• Policies that can be understood and executed
  • Return/exchange policies written in plain language
  • Operationally realistic ‘change window’ logic
  • Customer service escalation rules for edge cases where human support is required

What we’d recommend next

Moving forwards, the most valuable first step is to build internal awareness and understanding of the new agentic commerce initiatives emerging across Google, Shopify and the wider ecosystem. Whilst there’s certainly no need to panic and take rash action, it’s important to be aware of significant shifts like these and how they could impact the customer journey in the future.

Teams that take the time now to pressure-test their product data, promises and policies will be better placed to adapt as new AI-led buying journeys become more widely available and adopted.

We also anticipate that the launch of the UCP and the associated ability to transact within such agents and LLMs will have a material impact on performance marketing strategy. Early indications are that a ~4% fee is likely to be levied at the point of checkout. Building a clear understanding of contribution margin, experimenting thoughtfully with new tools, and being open to experimentation will all be key. 

Ultimately the mission for retailers remains the same: serve customers brilliantly, in every context.

How can Swanky help you prepare for agentic commerce?

If you’re a brand exploring what the UCP and other agentic commerce tools could mean for your growth strategy, we can provide support in a number of practical ways.

A few examples would include:

• Readiness & opportunity review – Identifying opportunities around agentic commerce and what needs refining to support them.

• Data & merchandising foundations – Ensuring taxonomy, collection logic, catalogue structure etc. is optimised to improve conversion both on and off-site – supporting AI-checkout-readiness.

• AI-led discovery & performance marketing – Optimising paid media for better performance in AI environments, focusing on ads as a validation layer in the new-look customer journey.

To discuss your requirements, please contact our solutions team today.