Google brings personalised deals to AI shopping, shifts ads beyond search links

New Delhi: Google is pushing deeper into “agentic commerce” with new AI shopping tools that blend assistance, intent signals and advertising, as the company looks to monetise how people search and shop inside conversational interfaces.

At the centre of the update is a pilot called Direct Offers inside Google’s generative AI search experience, AI Mode, powered by Gemini. The idea is simple: when a user appears close to purchase, brands can surface an exclusive offer inside the AI-led shopping flow, not as a traditional sponsored search link but as a contextual deal insert.

For marketers, that is a meaningful shift. Google’s pitch is that purchase decisions are often lost at the last mile: price hesitation, value doubts, shipping costs, or a moment of indecision. Direct Offers is designed to intervene precisely there, with the system determining when an offer should appear based on user behaviour and shopping intent.

From search ads to “in-conversation” deals

In the familiar search model, ads sit around organic results. Users click out to a retailer, browse, and then decide. With AI Mode, the journey compresses. People describe what they want in natural language, receive curated product recommendations, and can stay within the assisted flow longer than a standard search session.

Direct Offers adds a monetisation layer to that. Retailers pre-configure the offers they are willing to make available, and Google’s AI triggers them only when it believes the timing is right, i.e., when “it matters most” to close the sale, as the company frames it. In the pilot phase, Google is initially focusing on discounts and plans to expand into other value levers such as bundles and free shipping.

A universal protocol for shopping agents

Alongside ad formats, Google is also trying to solve a bigger structural problem: how AI agents talk to retailer systems consistently. That is where the Universal Commerce Protocol (UCP) comes in.

UCP is positioned as an open standard that allows shopping agents to research products and complete purchases without forcing users to hop between apps, websites and fragmented checkout flows. Google said the protocol has been developed with large retailers and platforms, including Walmart, Target and Shopify, and has support from a wider set of commerce and payments players.

In practical terms, Google said UCP will power more seamless in-chat commerce experiences within AI Mode in Search and the Gemini app, including purchases that can be completed within the AI environment for eligible retailers in the US, using supported payment flows.

For the industry, the protocol attempt matters because it signals the direction of travel: AI agents that do more than recommend. They can potentially execute, transact and “finish” shopping tasks, provided the ecosystem agrees on how those agents interact with merchant systems.

Why this matters for brands and marketers

For brands, Google’s update reframes three core levers of digital commerce.

1) Intent becomes the new inventory
In AI Mode, user intent is expressed in richer language. That gives Google more context to decide whether to show a deal, which deal to show, and when. In performance terms, the “moment of decision” becomes the prime placement, not a blue link at the top of a results page.

2) Offers become creative
Traditionally, brands optimise creative, audience targeting and bidding. In AI-led commerce, the offer itself can become the creative unit—discount, free shipping, bundle, or limited-time value. Google is effectively asking brands to think like a retail media player inside a conversational interface.

3) Product data becomes visibility
As shopping moves into AI conversations, structured product information and detailed attributes matter more than keyword stuffing. AI needs to understand what a product is, who it is for, what constraints it solves, and how it compares. Brands that treat product data as a marketing asset, not a backend feed, stand to win discovery.

For marketers tracking media mix and performance, the implication is not that search ads disappear, but that search marketing starts to look like commerce enablement.

• Budgets may shift toward “offer strategy” tied to inventory and margins, not just keywords and creatives.
• Measurement may evolve from clicks to AI-influenced conversions. Did the product get recommended? Did the offer trigger? Did it close?
• Brand safety and transparency questions may move from placements to how AI decides what to show and when, especially as ads are woven into answers.

Google’s announcements land amid a wider race to build commerce features into chatbots and assistants. Across the market, AI platforms are trying to convert free usage into revenue through shopping flows, checkout experiences, and paid placements that feel native to conversation.

That competitive pressure also explains why Google is pushing both ends of the funnel: a new ad format to monetise high-intent moments, and a protocol layer to reduce checkout friction and keep users inside its AI environments longer.

Google is clearly signalling that the next phase of performance marketing is not only about being found but also about being chosen in an AI-mediated moment, and making the purchase frictionless once that choice is made.