In November, Google encouraged users to “let AI do the hard parts of your holiday shopping” thanks to some new AI technology it was rolling out. Not sure what to buy a loved one? Search in AI Mode and the assistant will guide you all the way from discovery to checkout.
The search giant is referring to this new concept as “agentic commerce,” and last month it built on its initial offering with new tools designed to facilitate AI-powered eCommerce journeys—and bring paid ads directly into those journeys.
In this blog, we’ll explain “agentic commerce”, how Google is bringing paid ads into the fold, and highlight other recently-announced technologies that will contribute to Google’s overall vision to remake eCommerce with AI.
What is Agentic Commerce?
Google first introduced agentic commerce as a feature within its burgeoning AI Mode, the search engine results page (SERP) version of Google’s Gemini AI. Users have already been relying on chatbots like ChatGPT to research products, but Google’s countless advertising and eCommerce resources allow it to take that process a bit further. Plus, “agentic commerce” sounds a lot better than “chatbot shopping.”
In AI Mode, users can chat with Gemini to home in on products that fit their needs. It can reference over 50 billion product listings within Shopping Graph to identify options and provide product comparisons before eventually pointing them towards a purchase page. And for those looking to buy products in person, Gemini app could also help users inquire about product availability by emailing or calling local stores on their behalf.
Google’s “Direct Offers” Announcement
For advertisers, Google to introducing paid aids into AI Mode was always a matter of when, not if. As far back as 2023, Google has been indicating that (what it now calls) AI Overviews would someday integrate with Shopping ads. Now, a pilot program has finally arrived—what Google’s calling “Direct Offers.”
Direct Offers utilize promotional deals (like discounts or special pricing) from existing Shopping or Performance Max campaigns, which Google’s systems can surface inside AI Mode. This won’t be a case where advertisers can pay to “influence” the AI to recommend their products over others, however, and placements aren’t triggered by keywords or governed by a conventional auction. Instead, Google decides when an offer is relevant enough to show based on context and perceived purchase intent and then surfaces the deal alongside the AI chat.
Google Ads product liaison Ginny Marvin took to X, the social network formerly known as Twitter, to clarify three points about Direct Offers:
- “𝗜𝘁 𝗰𝗼𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝘀 𝘆𝗼𝘂𝗿 𝗲𝘅𝗶𝘀𝘁𝗶𝗻𝗴 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝘆: Direct Offers works alongside PMax and Standard Shopping campaigns. It currently serves in a specific product view to drive incremental conversions.
- 𝗜𝘁 𝗮𝗰𝘁𝘀 𝗼𝗻 𝗿𝗲𝗮𝗹-𝘁𝗶𝗺𝗲 𝗶𝗻𝘁𝗲𝗻𝘁: You don’t need to guess who needs a discount with rules-based offers. Google will use AI to determine when it’s best to display the offer based on shopper intent and market context, helping you close the sale with high-intent shoppers.
- 𝗬𝗼𝘂 𝗺𝗮𝗶𝗻𝘁𝗮𝗶𝗻 𝗳𝘂𝗹𝗹 𝗰𝗼𝗻𝘁𝗿𝗼𝗹: The AI will never “invent” a deal. The system relies strictly on your verified Merchant Center feed and the specific promo codes you upload.”
What This Means for Advertisers
On the surface, this seems like just another tool in your digital marketing toolbox. Performance Max, for instance, already displays ad content across Shopping, Search, GMail, Maps, Youtube, and other Google properties—what’s one more placement option added to that list? But the feature does raise some key questions for advertisers to consider.
We know users are already comfortable turning to AI for shopping assistance: They did it en masse prior to the 2025 holidays. Recent surveys suggest a majority of shoppers have used tools like ChatGPT or Gemini for product research, price comparison, and narrowing options. In some studies, nearly 60% of respondents report having used AI at some point in their shopping process.
But while many consumers are willing to consult AI, far fewer are willing to act on its recommendations without additional validation. Some studies point to an “AI trust gap,” where shoppers use AI as a starting point but still cross-check reviews, retailer sites, or other sources before purchasing.
That tension matters, because as Google introduces paid promotions like Direct Offers into AI Mode, some users may reasonably question whether recommendations are being influenced by advertising—even if the AI isn’t “inventing” deals, it may be hard to counter that perception.
Currently, the Direct Offers pilot is still in alpha—meaning that even Google Premier Partner agencies don’t yet have access to Direct Offers. Google says it is working with select retailers to refine the technology “into a faster, smarter way to help retailers connect with the next generation of shoppers.” In the meantime, advertisers will have to wait just a bit more to begin testing this long-expected feature.
Agentic Commerce at Scale
Google sees agentic commerce as the future, and one that it plans to dominate the field well beyond just ad features within its existing AI chats—and Google also recently announced a number of infrastructural pieces and tools for retailers
Universal Commerce Protocol
The goal of Google’s new Universal Commerce Protocol (UCP) is to bridge the gap between AI agents and merchant systems. Rather than building one-off integrations for each retailer or platform, Google is attempting to standardize how AI systems interact in service of eCommerce. According to Google’s announcement, the UCP will “soon power a new checkout feature on eligible Google product listings in AI Mode in Search and the Gemini app,” meaning that users will be able to complete purchases without even having to navigate away from the chat.
Business Agent
Alongside this infrastructure push, Google also announced new merchant-side tools designed to make it easier for retailers to get their piece of the agentic commerce pie. One of those is Business Agent—a chat assistant that displays to users viewing your Google Business Profile. Brands can input conversation openers and prompts that reflect their own voice, and the tech Business Agent will also draw from your website and data within Google Merchant Center to provide thorough answers about your company and products.
A version of this was first introduced in 2025, but customizability is the newest evolution—and Google claims it will continue to evolve, with enhanced training capabilities, performance insights, and agentic checkout integrations are “coming soon.”
Gemini Enterprise for Customer Experience (CX)
Google rounded out its agentic commerce announcements on the enterprise side of the equation, introducing Gemini Enterprise for Customer Experience (CX). The AI is integrated into a retailer’s own website, app, or support channels, where it can answer customer questions using information pulled from product catalogs, order systems, and support content. From the shopper’s perspective, it appears as a brand-owned assistant, though it does not seem intended to replace full-featured CX platforms that are already on the market—at least, not at this point. Google claims these agents can be deployed “in days, not months,” depending on a retailer’s infrastructure.
The Bigger Picture
It’s an understatement to say big tech is betting on AI. It’s pouring everything it has into it, and Google is no exception. Google’s “agentic commerce” concept isn’t exactly reinventing the wheel, but all of these initiatives in combination might be akin to gathering four wheels and putting an engine behind them.
On its own, Direct Offers is an inevitable culmination of the new AI-powered, Shopping-focused SERP that Google has been building towards over the past few years. In the greater context of this agentic commerce push, it’s just the opening salvo of consumer-facing tools that will continue to improve with all of the new infrastructure Google is developing on the backend.
Taken together, AI Mode, Universal Commerce Protocol, Business Agent, and Gemini Enterprise CX reflect a deliberate effort to standardize how commerce moves through Google’s systems. Advertisers and retailers are currently reliant on Google for helping consumers find their brands and products: Typically, users might search for multiple products on Google, follow either paid and organic listings to brand or retailer websites, compare reviews on YouTube or elsewhere, and then make a purchase on one of those websites.
This push aims to collapse those processes—and make them even more reliant on Google. Google is envisioning a future where all of that exploration, comparison, and even final purchases are made entirely in conversation with its own AI.
That means the emphasis right now should be on readiness—clean, consistent product data and brand identities will be essential. A willingness to jump in and experiment with these new tools and features will also be essential. For our part, our team will be ready to embrace each of these new tools as they become available to help our eCommerce clients compete on a whole new AI-powered playing field.
