According to Adobe Analytics data released in April 2026, AI-driven traffic to U.S. retail sites grew 393 percent year over year in the first quarter of 2026, with March alone up 269 percent. More importantly, AI traffic is now converting 42 percent better than non-AI traffic. Twelve months ago, AI traffic converted 38 percent worse. That is an 80 point swing in conversion rate in one year.
Most coverage frames this as a digital marketing story. It is not. It is the moment the discovery layer of retail moved from search engine results to AI agent answers, and most retailers, including the ones running modern physical stores, have not yet adjusted what they put online to be quoted by those agents.
Your physical store is no longer just a destination. It is a data source that AI agents are quoting, or ignoring, when they answer the shoppers who used to find you on Google. The retailers who make their stores quotable will own the next decade of brick and mortar discovery.
The Shift Underneath the 393 Percent Number
The headline figure is real. However, the structural change underneath it is more important. Adobe tracks more than one trillion visits to U.S. retail sites annually. Furthermore, in March 2026, the company reported that revenue per visit from AI referrals was 37 percent higher than from non-AI traffic. Just twelve months earlier, regular human traffic was worth 128 percent more than AI traffic per visit. The shift is not gradual. It is structural.
The behavior data is equally telling. Shoppers who arrive at a retail site from an AI assistant spend 48 percent longer on the site, view 13 percent more pages per visit, and show a 12 percent higher engagement rate. Specifically, they are not browsing. They are buying. According to Adobe’s companion survey of more than 5,000 U.S. consumers, 39 percent have used AI for online shopping and 85 percent of those say it improved their experience.
What changed is not the customer. What changed is the intermediary. Google built a 200 billion dollar business as the layer between shoppers and brands. By contrast, answer engines like ChatGPT, Gemini, Perplexity, and Claude are becoming the new intermediary. They do not just return a list of blue links. Instead, they make recommendations. They pre-qualify. They compress the funnel into a single answer.
And the physical store, the one most retailers spent the last decade modernizing, is largely invisible to that new layer.
Why Most Retailer Websites Are Failing the Machine Readability Test
Walk through the digital front door of a typical regional retailer in 2026 and you will see a homepage with hero banners, a category page with product grids, and product detail pages with photos and reviews. The site looks fine. The site converts well from paid search. However, the site is not fully readable by the large language models now driving 393 percent year over year traffic growth.
Adobe’s new AI Content Visibility data exposes the gap. The average U.S. retailer scored 75 percent on homepage readability. Product detail pages, where purchase decisions actually happen, scored just 66 percent. As a result, roughly a third of the content on the page where the conversion happens is invisible to the models recommending the product.
From the deployment side, three structural gaps consistently keep retailer storefronts out of the agentic discovery economy.
Gap 1: Content Trapped in JavaScript and Images
A meaningful share of product data on modern retail sites is rendered through JavaScript, embedded in images, or hidden inside interactive widgets. The human eye reads it. The AI agent does not. Specifically, ingredient lists, sizing information, fit details, sustainability credentials, and store availability are often the first casualties. Therefore, the retailer is technically online but functionally invisible to the answer engine.
Gap 2: No Structured Data for Physical Store Attributes
The shopper asking ChatGPT for a tortilla press in Miami at 7 p.m. on a Tuesday is asking for three data points: product availability, store location, and store hours. However, most retailer sites publish those three data points in three separate places, none of which is structured in a format an AI agent can resolve into a single answer. The retailer has the data. The retailer just has not made it quotable.
Gap 3: Product Detail Pages Optimized for Humans, Not Agents
The 2022 product detail page was designed to convert a human shopper who had already arrived through paid search. By contrast, the 2026 product detail page also has to convince an AI agent that this is the right product for the question being asked. That is a different writing job, a different data architecture, and a different testing framework. Most retailers have not even started.
The Adoption Layer: What Has to Change to Become Quotable
This is the part vendors will not put on a slide. Becoming quotable to AI agents is not a website redesign. It is an operating model change that touches digital content, store operations, product data governance, and the role of physical stores in the discovery funnel.
The Content Operating Model Has to Shift
In a pre-agentic world, the marketing team wrote product copy for humans and the search team optimized it for Google. In an agentic world, the content layer has to serve a third audience: the AI agent that will quote it or reject it. As a result, product copy now needs three properties at once. It must be human-readable, search-optimized, and machine-parseable, with structured data attached. Most retailers are running two of those three. The third one is the one that compounds AI-driven revenue.
The Physical Store Has to Publish Itself
Every physical store in 2026 generates real-time data: on-shelf availability, current pricing, planogram state, staffing, and exception alerts. However, almost none of that data is exposed to the digital storefront in a format an AI agent can resolve. As I described in the context of the Walmart Mexico ESL rollout and the GS1 RFID TDS 2.3 standard, the underlying infrastructure to publish that data is finally arriving at scale. The retailers who connect that store data to their digital surface will appear in AI agent answers as if their stores were API endpoints. The ones who keep the store data trapped inside the store will not.
The False Success Mode
The most common failure pattern in 2026 will be adding a schema markup plugin to the website, declaring the site AI-ready, and never connecting the deeper layers of product, store, and inventory data. The site will pass the basic readability test. The AI agents will still ignore it because the content is generic, the store availability is wrong, and the product specifications do not match the shopper’s question. Schema is not a strategy. Plugins are not architecture. Crawlable is not quotable.
Therefore, the retailers who will derive real value from agentic discovery are those who treat AI readability as an operating model change, not a marketing project. The technology has shifted. The job around the technology has to shift with it.
Three Architecture Decisions Every Retailer Has to Make
Before any retailer commits budget to an AI discovery program in 2026, three architectural decisions deserve direct answers.
Product Data as a System of Record
If product attributes live in five different systems (PIM, e-commerce platform, marketing site, store ERP, supplier feed) and none of them is the single source of truth, an AI agent will get inconsistent answers depending on which one it crawls. As a result, the agent learns to distrust the retailer. By contrast, a unified product data layer is the only way to give answer engines a coherent story to quote.
Store Data as a Published Layer
The shopper asking an AI agent, “Is this in stock at the store near me right now?” expects an answer, not a hyperlink. Therefore, store availability must be published in real time as structured data, not buried in a store locator widget that an agent cannot parse. This is where electronic shelf labels, RFID, and inventory robots stop being store technology and start being discovery infrastructure.
The Conversational Layer Continuity
The shopper who started a conversation in ChatGPT and walks into a physical store should not arrive as a stranger. As I described in the context of the conversational POS reframe, the in-store experience now has to consume the upstream conversation. Otherwise, the AI did the work, and the store wasted the lead.
What This Means for LatAm Retailers
Latin America has a structural disadvantage in agentic discovery that most CIOs in the region underestimate. The AI traffic share that Adobe is measuring is dominated by ChatGPT, Gemini, and Perplexity, all of which have stronger product knowledge graphs for U.S. retailers than for regional chains in Mexico, Colombia, Chile, and Brazil.
From the deployment side, I have walked stores in the region where the digital catalog is incomplete, store availability data is not published, and product descriptions are written for human readers in Spanish but not structured for machine consumption. As a result, a shopper in Mexico City asking ChatGPT for a specific product gets recommended a U.S. brand with shipping to Mexico rather than a local retailer with the product in stock five blocks away. The local retailer has the product. The local retailer is not quotable.
For LatAm grocers, drugstore chains, department stores, and specialty retailers, the strategic question is not “should we invest in SEO?” Rather, it is “Is our product catalog and our store data structured well enough that an AI agent answering a question in Spanish about our category will quote us first?” Importantly, that question goes to the CIO, the CMO, and the head of e-commerce at the same time. It is not a single-function decision.
Where to Start: The 18 Month Playbook
The sequencing playbook for retail leaders evaluating an AI discovery strategy over the next 18 months is concrete.
First 30 Days: Run the AI Visibility Audit
First, score your own digital surface for AI readability. Adobe’s scorecard shows the industry average at 75 percent on homepages and 66 percent on product detail pages. Therefore, anywhere below that benchmark is competitive debt. Run the audit category by category. Apparel, electronics, and groceries each have different machine readability gaps.
Two Quarters Out: Publish the Store Layer
Next, publish real-time store availability and pricing as structured data on every product detail page. Connect the ESL system, the inventory system, and the store locator into a single feed that an AI agent can resolve. Most importantly, measure how many AI agent queries return your store as the answer before and after the change.
The 18 Month Horizon: Build for the Agentic Funnel
Finally, rebuild the entire product detail page architecture for two audiences at once: the human shopper and the AI agent. That means structured product specs, machine-readable store availability, transparent pricing, and content written to answer specific questions rather than to rank for keywords. The retailers who treat this as a multi-quarter rebuild will compound advantage. The ones who treat it as a plugin will not.
In conclusion, the retailers who treat AI search as a digital marketing problem will lose share to retailers who treat it as a distribution architecture problem. The intermediary changed. The infrastructure has to change with it.
Google made retailers compete for ten blue links. AI agents make them compete for a single answer. The retailers who make their stores quotable will be the answer. The ones who do not will be the alternative that the agent did not mention.
If you are evaluating an AI discovery strategy or a digital architecture refresh for your retail network, connect with me here or reach me on LinkedIn. I am happy to walk through the deployment framework we use across the U.S. and Latin America.
Adriana Rivas is a retail technology executive and AI strategist, and the founder of a U.S.-based hardware company specializing in self-service kiosks, POS systems, electronic shelf labels, and digital signage deployed across the United States and Latin America. She is the award-winning author of How to Implement Self-Service Without Failing (Amazon #1 Hot New Release, Silver Nonfiction Book Award 2025) and recipient of the Gold Stevie® Award — Thought Leader of the Year 2026. She is also recognized by Thinkers360 as a Top 10 Thought Leader – Retail and a Certified Expert – Retail.