AI shopping agents will struggle to succeed until they have solved this one big problem
Tech giants are locked in an arms race to dominate AI-powered ecommerce. At Google I/O, we saw a preview of AI Mode in Search – a search experience where agents recommend products, populate visual panels, and complete purchases.
Next up is Apple’s WWDC, where the company is expected to provide an update on Apple Intelligence, but is believed to be taking a more incremental approach to AI in contrast to the rapid-fire rollouts from other companies racing to define the future of ecommerce.
These developer conferences, previously intended for insiders and engineers, are now mainstream moments because they’re shaping the future of shopping in real time.
Beneath the product demos and flashy interfaces, there’s a problem: today’s internet was counterintuitively built for humans, not machines.
Co-Founder of New Generation.
AI Search Is Running on Outdated Infrastructure
The web we use today is a patchwork of skimmable layouts and visual cues meant to guide people, not machines, through a shopping experience. AI agents don’t browse like people do, and they need structured metadata with real-time pricing, inventory and clear product attributes.
When the data is inconsistent, unstructured, or breadcrumbed across interactions, AI agents struggle to extract meaning or skip over it entirely. This means that for brands and shoppers products could become harder to find with AI, even if they are the best choice.
As AI-driven search platforms increasingly mediate product discovery, brands are losing visibility, traffic, and the ability to influence how they show up in the customer journey. The internet isn’t being rebuilt for AI, it’s being retrofitted. Many new interfaces look advanced on the surface but are layered over brittle, outdated infrastructure that machines struggle to understand.
Discovery Is Disappearing
We’re already seeing the early signals. Traffic from generative AI sources increased by 1,200% between July 2024 and February 2025, reflecting increased interest from consumers turning to AI tools for product discovery. The wave is arriving, but most brands aren’t yet positioned to take advantage of it because their websites aren’t designed to continue the AI user journey. Interfaces and product data often aren’t structured for agent interactions or optimized for LLM workflows.
Google’s AI Overviews can siphon off up to 64% of organic traffic, depending on the industry. It’s a dramatic shift in how discovery happens. For brands, it means fewer clicks, fewer opportunities for engagement, and far less control over how they’re presented in the shopping journey.
As consumers increasingly use AI agents for shopping and product recommendations, they’ll discover a narrower range of brands and products. Those that are optimized will be more easily found because AI models prefer sources that provide clean, well-structured, commerce-ready data like real-time pricing, inventory, and agentic checkout capabilities.
Without that data, AI agents may surface outdated or absent production information, forcing shoppers back into traditional, clunky checkout processes. The brands that proactively become AI-friendly will significantly benefit, making the path forward clear for their shoppers.
Some platforms are starting to recognize the problem. Shopify’s new Catalog API gives agents access to structured product data, making it easier to surface listings in agent-led environments like Perplexity. The API improves visibility, but not interactivity. One way infrastructure allows agents to access existing product data like descriptions and pricing, but the interaction ends there.
Two-way systems enable brands to proactively influence the experience, maybe by offering a discount, surfacing related products or offering free shipping depending on the customer interaction. Without two-way systems, brands will lose out on the control and context they’re accustomed to having.
What Brands Stand to Lose
AI innovation moves too quickly for brands to rely on incremental website updates. New model capabilities and consumer expectations emerge weekly, and without a flexible foundation built for constant adaptation, brands risk permanently falling behind.
Brands depend on search as the backbone of their visibility strategy to reach shoppers. Organic and paid search drove up to 80% of website traffic until Google’s AI overviews launched a year ago. Now, with AI Mode, agents are changing how information is retrieved and displayed, threatening not just traffic but the entire infrastructure of how brands reach, understand, and convert consumers.
This amounts to more than a visibility problem. As AI agents handle more of the customer journey, brands are losing the direct connections they’ve spent years building and the rich data that comes along with it: No more behavioral signals, preference data, or owned loyalty loops. When agents become the interface, the relationship gets rewritten.
Without traffic to their own websites, they forfeit first-party analytics, personalized engagement, and long-term insight into customer behavior. Without clear data connections, they can’t optimize experiences, measure ROI, or retain relevance. And without direct visibility, even brand affinity is at risk of erosion. In an AI-mediated internet, consumer choice gets collapsed into a single output. Unless a brand is structurally positioned to appear in that output, it might as well not exist.
A Programmatic Commerce Layer
This demands intelligent infrastructure. Brands should already be thinking about how they present their product information to make it legible to two important audiences: people and machines. Structured, real-time data is not optimization. It’s the baseline requirement for visibility, participation, and growth in an AI-first ecosystem.
In the AI internet, new subdomains like ai.brandname.com serve as intelligent storefronts that can serve both human customers and AI agents in one unified experience. Unlike traditional websites, which are updated piecemeal and built for human browsing, AI storefronts are built for speed, natural language, and agent-friendly architecture.
It’s time to rebuild now
Brands know they’re losing clicks, but the big picture is that they’re losing the ability to participate in the next era of commerce. AI agents are rewriting the script for how discovery and conversion happen; brands that aren’t structurally visible won’t be outcompeted, they’ll be invisible. In the AI internet, visibility is engineered. This starts with rebuilding digital storefronts for humans and machines.
We list the best website monitoring software.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
