Tech titans are locked in a race to dominate AI powered ecommerce. Google I/O gave us a sneak peek at AI Mode in Search, a search experience that allows agents to recommend products, populate panels and complete purchases.
Apple’s WWDC is next, where it is expected to give an update on Apple Intelligence. However, the company is believed to take a more gradual approach to AI, in contrast to other companies that are racing to define the future ecommerce. Originally aimed at insiders and developers, these developer conferences are now a mainstream event because they are shaping the future of online shopping in real-time.
Underneath the product demonstrations and flashy interfaces is a problem: Today’s internet has been built for humans not machines.
Cofounder of New Generation.
AI Search is Running on Outdated Technology
Today’s web is a patchwork, with visual cues and layouts that are meant to guide humans, not machines, in a shopping journey. AI agents do not browse like humans, and they require structured metadata that includes real-time inventory, pricing and product attributes.
When data is inconsistent, unstructured or breadcrumbed between interactions, AI agents have difficulty extracting meaning or simply skip it. This means that brands and shoppers could find it harder to find products with AI, even when they are the best option.
As AI search platforms increasingly mediate the discovery of products, brands lose visibility, traffic and the ability to control how they appear in the customer journey. The internet is not being rebuilt to support AI, but retrofitted. Many new interfaces appear to be advanced, but they are built on top of outdated infrastructure that machines have trouble understanding.
Discovery is disappearing
Already, we’re seeing the first signals. Traffic from generative AI increased by 1,200% from July 2024 to February 2025. This reflects the increased interest of consumers in using AI tools for product discoveries. The wave has arrived, but most brands haven’t been able to capitalize on it yet because their websites weren’t designed for the AI user journey. Interfaces and data are not always optimized for LLM workflows or structured for agent interaction.
Google AI Overviews could siphon up to 64% organic traffic depending on the industry. This is a major shift in the way that discovery occurs. Brands will have fewer clicks and engagement opportunities, as well as less control over the way they are presented during the shopping journey.
As more consumers use AI agents to shop and make product recommendations, their range of brands and products will be narrower. AI models will prefer data sources that provide clean and well-structured commerce-ready information, such as real-time inventory, pricing, and agentic check-out capabilities.
Without this data, AI agents could surface outdated or missing production information, forcing customers to return to traditional, cumbersome checkout processes. Brands that become AI-friendly proactively will benefit significantly, paving the way for their customers.
Some platform are beginning to recognize the issue. Shopify’s new Catalog API allows agents to access structured product data. This makes it easier for agents to surface listings in agent led environments like Perplexity. The API increases visibility but not interaction. Agents can access product data such as descriptions and prices, but interaction is limited.
With two-way systems, brands can proactively influence customer experience. They may offer a discount or suggest related products, or even offer free shipping, depending on the interaction. Without two-way system, brands will be deprived of the context and control they are used to.
What Brands Stand To Lose
AI innovations move too quickly for brands relying on incremental website updates. Brands risk falling behind if they don’t have a flexible foundation that allows for constant adaptation.
Search is the foundation of brands’ visibility strategy. Google’s AI overviews, which launched a year back, accounted for 80% of all website traffic. With AI Mode, agents change how information is retrieved, displayed, and accessed, threatening to disrupt not only traffic, but also the entire infrastructure for how brands reach out, understand and convert consumers.
It’s more than just a visibility issue. As AI agents take on more of the customer journeys, brands lose the direct connections that they have spent years building, and the rich data associated with them: No more behavioral indicators, preference data, or own loyalty loops. The relationship changes when agents become the interface.
Without traffic on their own websites they lose first-party analytics and personalized engagement as well as long-term insight in customer behavior. Without clear data connections they cannot optimize experiences, measure ROI or retain relevance. Without direct visibility, brand affinity can be at risk. In an AI-mediated Internet, the consumer’s choice is consolidated into a single output. Brands that are not structurally positioned so they appear in this output might as well be non-existent.
A programmatic commerce layer
It requires intelligent infrastructure. Brands need to think about how to present product information for two important audiences, people and machines. Structured real-time data does not equal optimization. It’s a requirement for visibility, growth, and participation in an AI-first eco-system.
New subdomains such as ai.brandname.com are intelligent storefronts in the AI internet that can serve human customers and AI agents with a unified experience. AI storefronts, unlike traditional websites that are updated piecemeal, and are designed for human browsing, are built with speed, natural language and agent-friendly architecture.
It’s time to rebuild
While brands are aware that they are losing clicks, the bigger picture is that they will lose the ability to participate the next era of commerce. AI agents are rewriting how discovery and conversion happens. Brands that are not structurally visible will not be outcompeted; they will be invisible. Visibility is engineered in the AI internet. This begins with rebuilding digital storefronts both for humans and machines.
List of the best website monitoring software.
The article was created as part of TechRadarPro’s Expert Insights Channel where we feature some of the brightest minds working in the technology industry. The views expressed are those of the writer and not necessarily those TechRadarPro, Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
Cofounder of New Generation.
