Fewer than four in 10 US e-commerce storefronts can technically support the full agentic AI shopping journey from discovery to payment, according to an InMobi analysis of roughly 3 million digital merchants. The finding lands with more than a month to go before Black Friday, and it frames the practical limits of AI-driven commerce for brands heading into the peak season.

The readiness gap is not evenly distributed. InMobi's scan found 15 percent of storefronts can support cart and checkout but not AI discovery, while 12 percent are machine-readable yet offer no way for an agent to complete a purchase. Another 37 percent of the 3 million sites have not been scanned at all, so the headline figure could still shift.

InMobi has packaged the diagnostic as a free tool, "Ready, Set, Agent," which grades a site against 80 metrics — whether AI agents can find the store, access it, read the catalog and check out.

"The first step is knowing exactly where agents can work with your store today, where they run into problems, and what will have the greatest impact if you fix it," said Abhay Singhal, co-founder of InMobi and CEO of InMobi Advertising.

What do shoppers actually trust AI to do?

Consumer-side data suggests the demand side is even softer than the supply side.

• More than one-third of US consumers are very or extremely willing to use generative AI tools and chatbots for holiday gift recommendations, per a Gartner survey of over 1,000 US consumers.
• 77 percent plan to shop in-store; 75 percent plan to shop on Amazon — traditional channels still dominate.
• Only 22 percent of over 4,000 US shoppers polled by Accenture would let AI decide what to buy, and just 8 percent would let it complete the purchase.

Family and friend recommendations and conventional search still outrank AI for gift ideas, Gartner found. One caveat: several of these surveys predate the release of Meta's Muse, the personal agentic AI that has topped app stores recently. Whether shoppers will hand it their money this holiday season remains untested.

Why structured product data is now a revenue question

Two vendor moves this season speak directly to the plumbing problem.

Algolia, which powers in-house search on brand websites, has acquired New York-based Velou. Velou uses AI to extract evidence-grounded attributes from product text and images against a curated retail library, so listings reflect how shoppers actually phrase queries — "machine-washable navy midi dress for a fall wedding" rather than "blue dress." The stated payoff: search and discovery that work out of the box "without months of manual tagging and merchandising rules."

The implications run beyond human search. AI agents depend on structured product data; if an attribute is missing from the data layer, agent shoppers cannot see it at all, Algolia said in a statement. For brands, incomplete catalog data now translates directly into lost visibility in both conventional and agentic channels.

Prompt-based deal tools and Amazon's cross-border play

Mint, an AI shopping companion, has launched a free "Prompt to Save" tool offering ready-made prompts from 250 partner brands and counting. Users fill in specifics and paste the prompt into Minty Chat, ChatGPT, Claude, Gemini or Grok. A sample Nike prompt asks the model to check recent resale prices for an exact colorway, compare authorized sellers, and return one out-of-pocket number with a buy-or-skip verdict.

Amazon, meanwhile, is targeting a different bottleneck: cross-border reach. Only 30 percent of its sellers operate in more than one country, the company said. Amazon is rolling out an expanded, unified view of expansion opportunities across the US, Europe and Japan, showing sellers where demand sits, per-country economics and required steps.

By year-end, sellers will be able to ship inventory to a single Amazon-operated facility near their factory and reach the US, UK, Japan, Germany, France, Italy, Spain and Canada from one consignment — rather than splitting stock per destination. Amazon says the consolidation cuts storage costs by up to 45 percent.

For sourcing and e-commerce teams, the near-term takeaway is concrete: catalog data hygiene and checkout accessibility for AI agents are now measurable, graded variables — and the retailers that close those gaps first will capture whatever agentic traffic materializes this season.