Google bets on AI to revolutionize your online shopping

Google Bets on AI to Revolutionize Your Online Shopping

At its annual Google I/O conference, the tech giant unveiled its ambitious vision for the future of e-commerce, heavily focused on artificial intelligence. Google is committed to transforming the online shopping experience through innovative tools, even as some of its competitors choose to slow down in this area. Among the new features revealed, the "Universal Cart" stands out. This concept allows users to centralize their purchases across various retailers and Google products, such as Gemini, and soon YouTube and Gmail. Consumers can add items to this universal cart while browsing the web or interacting with Gemini, and finalize their purchases directly through Google. This system also offers advanced features like price tracking, availability notifications, discount suggestions, and alerts about potential issues with selections.

Despite the upheavals that artificial intelligence has already caused in the worlds of work, business, and culture, tech companies still need to convince the general public of its benefits in everyday life. Google believes that the shopping sector could be a key area to demonstrate these advantages. Last November, the company introduced a feature allowing users to request an AI voice to contact physical stores for stock information. Additionally, it began testing a semi-automated method that enables AI agents to make online purchases on behalf of users.

The "Universal Cart" aims to simplify consumers' shopping habits by consolidating them in one place. According to Vidhya Srinivasan, Vice President and General Manager of Ads and Commerce at Google, consumers tend to shop over several days, using different devices and accounts. She explains: "Many of my shopping habits involve opening numerous tabs and syncing profiles. It works, but the cart solves this problem by bringing it all together. It's a cart that will be available wherever I am on Google properties." An icon of the cart will thus be visible next to the user's profile picture.

This universal cart is compatible with various retailers, including Sephora, Target, Wayfair, and Walmart. Eventually, users will be able to add items to their cart from YouTube or when browsing products in Gmail. Once a product is in the cart, users can receive alerts for price drops, check price history, and be notified when an out-of-stock item becomes available again. Srinivasan notes that the cart, operating on Gemini, can also alert a user to potential issues with their planned purchases. For example, if someone is building their first PC and chooses an incompatible motherboard and processor, the cart will flag this inconsistency and warn the buyer of potential problems.

Shoppers also have the option to connect retailer loyalty programs and credit cards via Google Pay. The Universal Cart will then suggest payment methods and potential ways to save money. If a user wants to build a cart but prefers not to finalize the purchase through Google, they can transfer the contents of their cart to a retailer's site and complete the purchase there. Srinivasan explains: "The retailer might have other things to show the person when they arrive on their site, and they can explore other options."

The concept of agentic shopping requires collaboration from various players, such as search engines, retailers, and payment processors. Retailer participation is crucial, as widespread adoption of agentic shopping could mean that customers no longer have a real reason to visit a retailer's website. This is what some refer to as the "Doordash problem." Amazon has even sued the AI company Perplexity for allowing users to purchase products through its Comet AI browser. OpenAI's efforts to integrate payment features into ChatGPT have also been disappointing. As more and more shoppers use AI chatbots to search for products or get recommendations, it becomes urgent for retailers and brands to appear on AI search platforms, which are already adjusting their online presence to be mentioned by chatbots.

Google is aware of its intermediary position between merchants and their customers. Srinivasan states that the company is "very focused" on the value exchange between all parties. She says: "Consumers benefit, but merchants do too, because in the long run, that's the only way it works." Srinivasan describes Google's role in this interaction as that of a "matchmaker." While Google benefits from the availability of billions of products for purchase within Gemini, retailers need something in return. Srinivasan clarifies that Google currently does not take a commission on sales or for purchased products.

When asked about retailers' concerns regarding the idea that Google could become the primary portal for online shopping, Srinivasan reaffirms Google's commitment to facilitating communication between many consumers and merchants, without becoming the go-to merchant. At the infrastructure level, signs indicate that the retail industry is rallying around Google. In January, the company announced the Universal Commerce Protocol (UCP), a new open standard developed with major retailers like Walmart, Shopify, and Target, which enables the entire AI shopping journey: item search, cart addition, purchase, payment, and post-sale customer service. OpenAI has its own competing version. In April, Amazon, Meta, Microsoft, Salesforce, and Stripe joined the committee governing the UCP.

The Universal Cart also provides users with pricing information. Google had already introduced a method allowing shoppers to buy products directly in the AI mode of search and in the Gemini app, and it is now expanding to categories like hotels and local food delivery. By using Gemini Spark, a new "24/7 AI personal agent" announced by Google, users will be able to give more specific directives to AI agents for their purchases, such as their preferred brands, the items they are looking for, and their budget. The shopping agent will then be able to make purchases on their behalf, provided all criteria are met. A buyer might specify the exact model of a pair of boots they want, for example, set a price limit, and ask the AI agent to buy the item when it finds it.

Purchasing uses a technology called Agent Payments Protocol (AP2), which essentially creates a digital trail and an approval process to allow an AI agent to perform a task like making a purchase. However, shopping is complex. What happens, for instance, when the robot buys something below the price specified by a user, but with taxes and shipping fees, it ends up being much more expensive than another option? Would most shoppers trust AI enough to spend their money on their behalf? Srinivasan indicates that Google is currently working on these issues, but generally, a buyer would go to the actual retailer, not Google, to resolve post-purchase problems. This raises another question: will retailers introduce policies regarding purchases made with AI?

Buying things is also an emotional experience. If a rare item from my wish list appears for $4 more than the limit I set for the chatbot, I might go ahead with the purchase even if the robot couldn't. It's hard to imagine a world where shoppers immediately delegate their purchases to a machine: that would represent a radical transformation of what it means to buy things. Most importantly, adoption would require enormous trust from consumers — and that remains a challenge to overcome.