Introduction
AI Mode, AI Max, agentic checkout and a wave of new feed attributes all landed within months of each other, and together they change what a Google Shopping campaign competes on. This article covers what Google announced, what the early traffic data shows, and why product feed quality now determines whether AI recommends your product or your competitor's.
The shopping surface is expanding
This year, Google unveiled a wave of AI-powered commerce tools that together represent a major shift in how products are discovered and compared across shopping surfaces. The announcements included the Universal Commerce Protocol (UCP), an open standard co-developed with Shopify, Walmart, Target, and Wayfair that allows AI agents to handle entire shopping journeys, from discovery through checkout. Google also introduced Business Agent, a branded chat feature connecting shoppers with retailers directly in Search, and Direct Offers, a pilot that lets advertisers serve exclusive discounts inside AI Mode results.
In short, Google is building an end-to-end commerce layer inside its AI surfaces where shoppers discover, compare, and buy products without ever visiting a merchant's website.
Since the announcement, Ulta Beauty is already live with agentic commerce in AI Mode and the Gemini app, where shoppers receive product recommendations, compare options, and complete checkout within the conversational interface. Walmart has integrated its full assortment into Gemini through UCP. These new announcements are live consumer experiences backed by some of the largest retailers in the world.
For PPC managers and e-commerce teams, this means the surface where Shopping ads compete is expanding from a carousel of product tiles into a conversational AI layer that reads, reasons about, and recommends products based on structured data.
AI Max for Shopping, expanding automation
On 30 April 2026, Google announced AI Max for Shopping campaigns, extending its AI-powered ad automation layer from Search into Shopping. The feature generates ad copy from Merchant Center feed data, matches ads to conversational queries, and uses Final URL Expansion to route shoppers to the most relevant landing page for each search. Optimal Format Selection then chooses between a text ad and a Shopping ad depending on what fits the query.
Google positions this as a one-click upgrade for existing Shopping campaigns. Your product targeting and bidding settings carry over, and Final URL Expansion can be switched off at any time to restrict delivery to Shopping ads. For advertisers running Performance Max, URL expansion and format selection are already active there, with text customisation for Shopping ads arriving as an addition.
What this means in practice: as Shopping ads start appearing against complex, conversational questions, a static product title pulled from a feed matches fewer of those queries than a headline generated from the language of the search. AI Max builds that headline from your Merchant Center attributes, material, fit, use case, durability. The quality of what it generates is capped by the quality of what it reads. Advertisers with rich, structured feed data and clear text guidelines give the system something to work with. Advertisers upgrading a thin feed with one click will see the automation, not the results. Before opting in, treat your feed as the campaign asset it now is: complete the attribute fields, review the generated copy, and keep FUE under evaluation rather than assumption.
The zero-click paradox
AI Mode has a 93% zero-click rate, according to Semrush data. A randomised field experiment by researchers at the Indian School of Business and Carnegie Mellon University found that AI Overviews reduced organic clicks by 38% on queries where they appeared.
Although that sounds like a big shift, the full picture is more nuanced:
The clicks that do come through from AI surfaces convert dramatically better. Cross-site analysis shows AI-referred traffic converting at 7 to 16%, compared to 1.7 to 2.8% for traditional organic search. Shopify merchants saw 14x more orders from AI-driven chats in January 2026 compared to a year earlier. AI was credited with influencing 20% of all retail sales during the 2025 holiday season.
The traffic model is transforming. Volume is shifting from high-quantity, low-intent clicks to lower-quantity, high-intent interactions where the AI has already pre-qualified the shopper. For merchants, this means the relevant metric is shifting from impressions and click-through rate to citation visibility (how often your products appear in AI answers) and conversion quality.
What this means for product feeds
In traditional Google Shopping, a product feed needed to be "good enough." Accurate titles, clean images, correct pricing, valid GTINs. The feed got your product into the auction. Your bid strategy and account structure did the rest.
In AI-powered discovery, the feed does most of the heavy lifting. When a shopper asks AI Mode something like "What are the best high-quality clothes for lounging?", the AI doesn't match that query to a keyword. It evaluates product attributes: fabric composition, fit descriptions, care instructions, customer review sentiment, imagery, shipping speed, return policies. Products with richer, more complete structured data get recommended. Products with sparse data get skipped.
Google has made this explicit. Alongside the January announcements, they introduced dozens of new Merchant Center data attributes designed specifically for conversational commerce, including answers to common product questions, compatible accessories, and substitute products. These attributes go beyond traditional keyword-driven fields. They're designed to help AI agents answer the kinds of questions real shoppers ask: "Will this fade in sunlight?" "Is it machine washable?" "Does it work with my existing setup?"
Industry practitioners are already quantifying the gap. Stores with near-complete attribute coverage (what some in the industry call a "Golden Record" at 99.9% completion) are reporting 3 to 4x higher visibility in AI recommendations compared to stores with sparse data. That's not a marginal improvement. It's a different league.
The practical checklist for merchants looks something like this: context-rich product titles that describe use cases, not just product names. Descriptions that run 500+ characters with specific material, dimension, and performance details. At least three high-quality images per product. GTINs on every eligible product. Shipping windows and return policies populated in the feed, not just on your website. And increasingly, the new conversational commerce attributes that Google is rolling out through Merchant Center.
Want to get ahead on feed quality today? We wrote an actionable eBook that’s free to download on how to do it: Product feed fundamentals: the Google Shopping checklist.
What this means for CSS partnerships
For merchants advertising through a Comparison Shopping Service in Europe, the structural CPC advantage (up to 20% lower costs in the auction) remains intact. That hasn't changed.
What has changed is the surface area where that advantage plays out. Shopping ads now appear in AI Mode, Gemini, and conversational commerce surfaces that didn't exist a year ago. The DMA-driven changes to CSS PLAs are introducing direct competition between CSS providers in the auction. And the new Merchant Center attributes designed for agentic discovery mean that feed quality has a direct, measurable impact on which products AI recommends.
This shifts the value proposition of a CSS partnership. A partner that simply submits your existing feed and passes through the auction advantage is leaving performance on the table. The CSS partners that will deliver outsized returns are the ones actively enriching feeds for AI discovery: filling the new conversational attributes, optimising product data for the Shopping Graph, and ensuring merchants are visible not just in the traditional Shopping carousel but in the AI surfaces where an increasing share of high-intent shopping is happening.
At Shoparize, we manage feeds and campaigns for 25,000+ merchants across 21 EU countries. Feed optimisation isn't a separate service. It's built into how we run Managed Ads, because in the agentic commerce era, the quality of your product data directly determines whether AI recommends your product or your competitor's.
The bottom line
The merchants who win in this environment aren't the ones with the biggest budgets. They're the ones with the best-structured data. When AI is the intermediary between a shopper's intent and a merchant's product, the feed becomes the primary customer touchpoint. Not the homepage. Not the brand story. Not the lifestyle photography. The structured data.
This shift is already happening, and already separating the merchants who are adapting from the ones who are still optimising for a version of Google Shopping that no longer exists.
