Google is ushering shopping into the agentic age, deploying A.I. that can dial nearby stores to see what’s in stock — and make purchases for you using Google Pay, if you give it the green light. It’s quite a leap beyond recommendation engines: This system doesn’t just help you find some products — it acts on your intent, across search, chat and checkout.
The company is wrapping that feature with a wider set of retail tools, from conversational product discovery to tracking prices. The vision is convenience, but execution will depend on practical guardrails such as consent safeguards, merchant participation and category limits as the technology scales.
How the Calling Feature Works for Shoppers
When users look up products “near me,” some results list the ability to “Let Google call.” The A.I. will be able to ask you some clarifying questions — model, size, color or based on price range — dial the store and confirm them, then take that information away in a text message or email. Toys, electronics and health and beauty will be among the first available early, with businesses having the option to opt out either through their Business Profile or by using a dedicated hotline.
This draws on Google’s experience with automated calling for restaurant wait times and appointment bookings. The shopping expansion still revolves around high-intent queries where a quick answer leads the user to make a purchase. It’s worth noting that the feature is not yet live in all U.S. states: It will not be available at launch in Indiana, Louisiana, Minnesota, Montana, or Nebraska.
In the wild, in other words, you would ask about a Nintendo Switch OLED White at some local shop. And the AI checks in real time whether an item is in stock and what it costs, and sends you the answer to your phone — no hold music or “let me check in the back.” For people who hate making phone calls or require help with accessibility, that’s a real quality-of-life improvement.
Agentic Checkout Drives the Purchase Experience
Google is also testing out an AI shopping agent that monitors for stock changes and price drops, taking action itself. You can set a target price and list product variations — shoe size, color or pack count — and be notified when the conditions are met. With your approval, the AI can then purchase it via Google Pay and given the shipping information.
The rollout begins in the U.S. with some merchants, like Wayfair, Chewy and Quince, as well as “select” Shopify stores before reaching more locations in the coming months. That curated approach is what keeps the experience consistent even as Google widens the merchant pool. It also addresses an age-old e-commerce pain point directly: checkout abandonment. According to the Baymard Institute, average cart abandonment is around 70%, not all of which is attributable to browsing or price comparison — much can be attributed to friction such as hidden costs and long forms. Drop-off could be significantly reduced if it were a system agent handling timing and form-fill after user confirmation.
Think about the popular serum that goes out of stock momentarily in a particular color or size. Instead of constantly refreshing product pages, you set the alert once and confirm the purchase as soon as stock pops up under your price point. The purchase is completed as a background task; you continue to be in control of what is purchased and where it’s shipped.
Conversational Discovery on Search and Gemini
Even Discovery is getting a renovation. Search’s AI Mode invites plain-language prompts like “cozy sweaters for happy hour in warm autumn colors,” and returns shoppable images, prices, reviews and in-stock signals. Ask it to compare products, and you’ll get an AI-generated table that distills features and user feedback from across the web.
And within the Gemini app, that same model serves as a personal shopping assistant — thinking up ideas for gift lists or tactics to take advantage of sales and surfacing items you might want to buy in-app.
It’s not so much about individual questions and answers as it is about planning — “help me pick practical gifts for a new college student,” with the agent then turning advice into a purchase-ready shortlist.
The Data Engine Behind It and the Built-In Safeguards
Running things under the hood is Google’s Shopping Graph, a real-time index that the company says houses more than 50 billion product listings, many updated hourly. That breadth drives accurate availability checks, pertinent comparisons and timely price alerts. The Graph’s scale also provides Google with signal density — spanning reviews, specs and merchant updates — which smaller assistant players have a hard time matching.
Google makes a big deal of consent at the transaction layer: The AI asks your permission to buy for you and confirms shipping addresses every time. Merchants also hold the reins; businesses can choose not to receive automated calls, and the calling function is limited to certain categories and states. All those guardrails matter as regulators and even consumer advocates watch how agentic buying is executed, and account for what’s being done with stored payment credentials.
The Competitive Stakes and What to Watch Next
The move comes amid increasing competition around AI-powered shopping. Amazon recently filed a lawsuit against Perplexity over an assistant that allowed for one-click purchases through Amazon listings, demonstrating legal and platform risks around automated transactions. Google’s phased rollout with handpicked partners — and no Amazon integration — indicates a more cautious approach.
The big unknowns now are coverage and trust: how soon Google can scale up to categories, states and participating retailers, and whether shoppers will be willing to hand the last mile of a purchase over to an A.I. If such experiences continue to remove all friction while respecting user intent, agentic shopping may graduate from novelty to default behavior for high-intent purchases.
