AI shopping is moving from ‘help me research’ to ‘help me buy’. Instead of a customer clicking through search results, an AI assistant can understand intent, compare options, confirm delivery dates, and complete checkout inside the conversation.
This is not just a new ad format. It is a change in how your products get discovered. When the buyer becomes an agent, your product data becomes the storefront. If your product feed is not optimised, you will be ruled out before you compete.
Google is bringing ‘Buy’ buttons and checkout into Gemini and AI Search, alongside a new standard called the Universal Commerce Protocol (UCP). Google is also introducing new Merchant Center data attributes for conversational discovery, and testing ads in AI Mode with a Direct Offers pilot.
Shopify is backing UCP as an open standard it co-developed with Google, and is encouraging merchants to plug into these AI shopping channels through Agentic Storefronts, managed centrally in Shopify Admin.
PayPal is framing this as a ‘protocol moment’, with consent, trusted identity, reliable payments, and fraud protection as non-negotiables when an AI agent is the one clicking ‘Buy’. For merchants, that means a focus on data structure, integration readiness, monitoring, and ongoing compliance as standards evolve. PayPal has now confirmed it supports UCP and says it will soon be a payment option within Google’s UCP-powered checkout.
Multiple players are competing to define the common language for AI shopping, and that is what is fuelling the ‘protocol wars’. Salesforce has also announced support for UCP, signalling adoption is spreading beyond Google and Shopify into wider commerce platforms.
What to watch: Some commentators suggest AI chat platforms may adopt marketplace-style economics, including charging fees when transactions happen inside the chat. The exact models are still emerging, but the direction is clear: more shopping activity will happen inside AI-led experiences.
Shoppers describe what they want: 'waterproof, recycled materials, under £100, delivered by Friday.' The AI turns that into requirements and finds products that match. This shifts the focus from SEO keywords to structured attributes.
UCP is designed to standardise how AI agents communicate with retailers and service providers, so they can do more than recommend products. The goal is reliable discovery, checkout, and post-purchase support without custom integrations for every new AI channel.
The buyer might not visit your site at all. The AI can guide the purchase through a supported checkout, using approved payment methods and merchant rules. That reduces friction and makes clean data and reliable fulfilment even more important.
The classic customer journey is getting shorter. In many cases, a shopper may see only one recommended option. That raises the stakes on being included, trusted, and chosen based on the data the AI evaluates.
When an AI chooses based on verified facts, your Merchant Center data quality directly shapes whether you are selected, not just whether you get a click.
If your price, stock, delivery promise, or product specs are wrong, the AI cannot confidently recommend you. This is not only a conversion issue, it can become a reputation issue inside agent-led ecosystems.
If purchases happen inside an AI experience, performance signals will not always look like traditional click-through conversions. Be prepared for attribution to evolve, and start tracking performance at product and feed level, not only at campaign level.
In the context of the 'Protocol Wars' and the shift to agent-led shopping, feed optimisation platforms help translate retailer data into a format AI systems can reliably use.
As Google builds the 'answer' through structured protocols like UCP, retailers need feed processes to enrich attributes, validate compliance, and refresh price and stock. Feed optimisation platforms such as Shoptimised are designed to automate those updates across channels, keeping products eligible, accurate, and competitive as AI-led checkout expands.
Focus on the attributes an AI needs to confidently choose a product:
Why it matters: AI shopping interprets intent through attributes and constraints more than SEO keywords.
Increase refresh cadence where possible and set monitoring for:
Why it matters: AI agents need real-time price and stock to buy. Stale data breaks trust and can knock you out of recommendations within the AI ecosystem.
Assume your products will be evaluated in more places than the classic Shopping results page, including conversational journeys where shoppers ask follow-up questions. Make sure your feed can answer those questions with structured facts.
Run controlled experiments on:
Why it matters: when shoppers see fewer options, small improvements in relevance matter. Shoptimised helps you test title and attribute changes across your feed, and scale the winners based on performance.
As AI shopping blends ads, recommendations, and checkout, scrutiny will increase. Make sure you can clearly explain:
Why it matters: protocols like UCP are already specifying stronger consent and verification mechanics, so merchants should expect governance and compliance requirements to become more formal over time.
Google and partners are pushing UCP, while others will push alternatives. You do not need to predict the winner today. You do need to keep your catalogue structured and portable so you can adapt quickly.
AI shopping is turning discovery and checkout into an agent-led conversation. Google is setting the standards for that shift with UCP and in-chat checkout across Search and Gemini.
For Google Shopping Ads users, the winning move is simple: optimise your feed. The brands with the cleanest, most accurate, most current product data will be the ones AI systems trust and recommend. That is where feed optimisation platforms, like Shoptimised, help by keeping your catalogue structured, compliant, and up to date at scale.
AI shopping is moving quickly from 'nice to have' to a real traffic source retailers need to take seriously. New Adobe figures reported in May show UK shoppers arriving from AI sources converted more often than those coming from traditional search, with the UK AI shopping conversion rate up 182% year on year and 543% since January last year.
At the same time, the platforms are building more of the shopping journey directly into the AI experience. Google has been pushing Shopping features in AI Mode, including richer product discovery, virtual try-on using your own photos, and more automated purchase flows.
Amazon is also taking its 'Alexa for Shopping' capability beyond Amazon. It is licensing the underlying tech to other retailers through AWS, with Kate Spade cited as an early customer.
The common thread is simple: as shopping becomes more AI led, clean product data matters even more. If your titles, attributes, pricing and availability are messy, the AI layer has less to work with, and you lose visibility where customers are starting to browse and decide.
If you’re not sure your feed is ready for AI-led shopping, run a free audit or get in touch and we’ll point you at the quickest fixes.
