TLDR
An audit of 141 product pages from 29 major retailers found 70 per cent lacked all three schema fields required by Google's Universal Commerce Protocol. AI shopping agents in tools like ChatGPT and Gemini cannot complete purchases for products missing price validity dates, delivery times, or return policy data. Australian ecommerce brands competing in a fast-moving agentic commerce landscape face a concrete and fixable gap. Three specific schema fields added to structured data feeds are all it takes to close it.
KEY TAKEAWAYS
A quiet shift in how products get bought
Agentic commerce works simply enough. A shopper asks an AI tool to find and buy a birthday gift within a budget, and the agent handles the rest: browsing, comparing, transacting. The appeal is obvious. What is less obvious, until you look at the data, is how many retailers have already been cut out of that conversation.
An audit published in July 2026 examined 207 top-traffic product pages across 29 major retailers. After removing pages that could not be evaluated, 141 remained, and 70 per cent of those pages lacked all three schema fields required by Google's Universal Commerce Protocol.verifiedVerified Source: searchenginejournal.com[1] For those brands, AI shopping agents simply cannot process their products.
What the Universal Commerce Protocol actually requires
Google's Universal Commerce Protocol is an open standard designed to make products readable by AI agents across Search, AI Mode and the Gemini app. It builds on existing structured data standards but demands three specific fields that most product pages have historically ignored.
Ashish Gupta, Vice President and General Manager of Merchant Shopping at Google, said the intent is direct. "We're excited to work alongside Target and others to co-develop the Universal Commerce Protocol (UCP) and enable customers to buy their favorite items directly on Google Search, including AI Mode, and the Gemini app," Gupta said.[2] Sarah Travis, Target's Executive Vice President and Chief Digital and Revenue Officer, told the announcement: "More people are discovering products and finding inspiration in AI-powered environments, and we see a real opportunity to meet them on their shopping journey."[2]
The three fields most retailers are missing
The audit findings are stark when broken down by field. Only 18 per cent of audited pages carried the priceValidUntil schema field, only 13 per cent included shippingDetails.deliveryTime, and only 11 per cent included hasMerchantReturnPolicy.merchantReturnDays.verifiedVerified Source: searchenginejournal.com[1] These are not obscure technical attributes. They map to three questions any buyer asks before committing: how long is this price good for, when will it arrive, and can I return it?
An AI agent completing a purchase on behalf of a shopper needs those answers in machine-readable form. Without them, the product is effectively invisible, regardless of how competitive the price is or how strong the brand.
Blocked before the agent even reads the page
Of the 207 URLs sampled in the audit, 32 returned HTTP 403 errors, meaning AI agents were blocked from reading those product pages at all.verifiedVerified Source: searchenginejournal.com[1] A 403 error is a server-level access denial, often a side effect of bot-blocking configurations that make no distinction between a malicious crawler and a legitimate shopping agent.
For those retailers, the schema problem is secondary. The agent cannot get through the door at all. Fixing access controls for verified AI agents is a precondition for any structured data work to matter.
Three feed fixes Australian retailers should make now
The three required UCP fields map directly onto information retailers already hold. priceValidUntil tells an agent when a listed price expires. shippingDetails.deliveryTime specifies the delivery window in a structured format. hasMerchantReturnPolicy.merchantReturnDays states the return window in days.[1]
Adding these fields to existing structured data feeds or JSON-LD blocks is a targeted technical task, not a platform rebuild. Australian ecommerce brands already investing in search visibility can act on this now. Retailers who make themselves easiest for an AI agent to read and transact with are the ones who will appear when a shopper hands the buying decision to a machine.
What agentic commerce means for search visibility
The audit data points to a structural shift in ecommerce search that goes beyond any single platform update. As AI agents take on more of the discovery and purchasing work shoppers once did manually, competitive advantage shifts toward feed quality rather than page design or ad spend.
Google's Universal Commerce Protocol, co-developed with Target and announced in June 2026, is the clearest signal yet that the industry is moving in this direction.[2] With only 11 per cent of audited product pages carrying the return policy field and 32 URLs blocked entirely, most retailers are not ready for a shift that is already underway.
SOURCES & CITATIONS
FREQUENTLY ASKED QUESTIONS
What is the Universal Commerce Protocol?
Which three schema fields does the UCP require?
Why do HTTP 403 errors matter for AI commerce?
How can Australian retailers fix the gap?
Helen Ross covers wealth, inheritance and the human side of money for Bushletter. She focuses on how families actually navigate estates.
