Retailers Spent 20 Years Competing for the Customer’s Attention. Now They Also Have to Compete for the Agent’s.

On August 3, Scott Thompson at the Retail Technology Innovation Hub published a piece anchored by an observation from Dr. Astha Purohit at Walmart, where she serves as Director of Product Management, Customer Data and Identity. The piece drew on the Fynd whitepaper on agentic commerce. That whitepaper documented a platform processing 1.6 million daily orders and an AI agent driving 40 percent revenue uplift. The most important line, however, was not about Fynd. It was a single observation from Purohit about what retail competition now looks like:

“For 20 years, retailers competed for the customer’s attention, the search result, the shelf, the homepage. Increasingly, you’re also competing for the agent’s attention, because the agent decides which handful of products the customer ever sees.”

That is a precise description of a shift that most retail marketing teams have not yet priced into their 2026 budgets. The customer still makes the purchase decision. The agent now decides what the customer gets to choose from. Appearing in that handful is not an SEO problem. It is a GEO problem. Generative Engine Optimization. It is the most consequential visibility challenge retail has faced since Google became the default discovery layer twenty years ago.

GEO is not a new name for SEO. It is a different discipline, with different inputs, different measurement, and a different competitive dynamic. The retailers who understand that distinction in 2026 will own product visibility in the agent layer. The ones who treat GEO as a rebranding exercise will lose transactions they never knew they were competing for.

The Framework That Changes Where You Invest

Purohit offered a three-segment framework for thinking about where AI agents change the retail experience and where they do not. The framework is the most operationally useful retail technology analysis published this week. It tells you directly where to invest in GEO and where to invest in something else.

Luxury and High-Consideration Purchases: Experience Over Agents

For luxury purchases, where the customer is spending significant money, Purohit expects a shift toward in-person, analog experiences. The purchase becomes more experiential and shopper-centric. Brands will lean hard into that. An AI agent recommending a luxury watch or a high-end handbag does not deliver the experience that drives conversion in that category. The physical environment, the human interaction, and the brand story are doing that work. Consequently, for luxury retailers, GEO is a secondary concern. The investment belongs in the physical experience and the human layer.

Considered Purchases: AI Does the Research, Humans Make the Call

For considered purchases, a laptop, a television, an expensive appliance, AI agent involvement is already high and growing. The agent handles comparison research, pulls reviews, and narrows the options to a short list. However, the final decision belongs to the customer. It depends on whether they saw an influencer they follow using the product, whether they trust the brand, and how they feel about the price. The agent shapes the consideration set. The brand shapes the conversion.

For this segment, GEO is critical at the top of the funnel. If a retailer’s products do not appear in the agent’s consideration set, the brand never gets the opportunity to influence the human decision. Furthermore, the product data needs natural language query structure, not keyword matching. A shopper asking an AI agent to find a laptop for video editing under two thousand dollars is not typing keywords. The agent needs product attributes, use-case descriptions, and structured comparisons to surface the right result. As I described in the AI shopping cart and ChatGPT transactable commerce analysis, presence in the catalog is necessary but not sufficient. The quality of the product data determines the quality of the recommendation.

Everyday Shopping: The Agent Builds the Basket

This is where the structural shift is most visible and most immediate. Purohit’s description of everyday shopping is worth quoting directly. Agents build the basket, assemble the grocery list, push it to the app for a quick tap review, and reorder the staples automatically. It moves toward a subscription rhythm where the customer is barely thinking about it.

For grocery and everyday retail, the agent does not narrow a consideration set. The agent makes the decision. The customer approves it with a tap. In this segment, appearing in the agent’s default selection is not a top-of-funnel visibility problem. It is a retention and share-of-basket problem. A brand that gets into the agent’s default basket stays there through inertia. Any brand that does not get in faces a basket already built without it.

What GEO Actually Requires

Structured Product Data for Natural Language Queries

SEO is built around keywords. GEO is built around context. A shopper asking an AI agent for oat milk that is barista-grade and comes in a one-liter carton is not typing a keyword. The agent searches for products where the description, attributes, and category tags match the contextual request. Retailers whose product data uses keyword structure will appear in fewer agent recommendations. Retailers who describe use cases, contexts, and specific attributes in natural language will appear in more. The investment is in catalog restructuring. The skill set is in product taxonomy and data quality, not in keyword bidding.

Authority Signals That AI Models Can Read

Traditional SEO built authority through backlinks. GEO builds authority through a different set of signals. These include editorial coverage from sources that informed AI model training, reviews across structured platforms, expert endorsements in formats agents can parse, and brand mentions in authoritative contexts. A retailer with strong editorial coverage in trade publications, verified review profiles, and structured product certifications will surface more consistently in agent recommendations than a retailer with the same search ranking but weaker structured authority signals. As I described in the AI-referred traffic conversion analysis, AI-referred shoppers convert at higher rates than visitors from any other channel. That means the authority investment in GEO has a higher conversion payoff than the same investment in traditional search.

Transactability Through the Agent Layer

Visibility without transactability is incomplete GEO. An agent that can surface a product but cannot complete the purchase inside the conversation will recommend a competitor whose checkout infrastructure is connected. As I described in the shopper agent analysis, retailers who grew sales 59 percent faster were running agents that completed transactions, not just recommended them. GEO therefore requires both the discoverability layer and the transactability layer. Optimizing only for discoverability while leaving checkout infrastructure unconnected to agent networks produces a misleading result. The agent surfaces the product. The customer does not buy because the friction of leaving the conversation is too high. That outcome is easy to measure and easy to mistake for progress.

What This Means for LatAm Retailers

The GEO conversation is currently centered on English-language AI models and U.S. retail. However, the same dynamic is already active in Spanish-language and Portuguese-language AI interactions through Google, Meta AI, and WhatsApp. A shopper in Mexico City asking WhatsApp’s AI assistant for a recommendation in a consumer electronics category triggers the same agent-mediated discovery process that Purohit described. The product that appears is the one whose data is structured for that AI model to read and surface. The product that does not appear loses the transaction before the customer ever knew it was available.

For LatAm retailers, the GEO investment starts with the everyday shopping segment, because that is where agent penetration is already happening through WhatsApp commerce and AI-assisted grocery ordering. Product data quality, catalog structure, and review presence all determine agent visibility in that segment. Those decisions can be made now, before agent-mediated everyday shopping reaches in LatAm the scale it has already reached in the U.S. and UK markets.

Three Questions Worth Asking Before the Next Planning Cycle

Purohit concluded her remarks to RTIH with a statement that most retail marketing teams should print out and put on the wall: getting this right has quietly become table stakes. It is not a future concern. It is a current competitive condition.

The Segment Question

Which of the three segments, luxury and experiential, considered purchase, or everyday shopping, represents the majority of your transaction volume? The answer determines where your GEO investment belongs and whether the agent layer is your primary visibility problem or a secondary one. Most retailers serve more than one segment and need a different GEO strategy for each.

The Visibility Test

If a customer asked an AI agent to find and buy the top three products in your primary category right now, would your products appear? Not in a hypothetical future deployment. Today, using the AI assistants your customers already have access to. The answer is a current state assessment, not a strategy discussion. If you do not know the answer, that is itself the answer.

The Strategy Gap

Does your marketing team have a GEO strategy, or do they have an SEO strategy with AI mentioned in the executive summary? Those are different documents, built on different inputs, optimizing for different outcomes. The one you have determines which competitive layer you are investing in.

For 20 years, the search result was the shelf. The agent is the new search result. And unlike the search result, the agent does not show the customer a page of options. It shows them a handful. The retailers in that handful will define the next decade of retail share. The retailers outside it will be competing for the attention of a customer who already decided.

If you are building your GEO strategy or evaluating your product data architecture for agent-layer visibility, connect with me here or reach me on LinkedIn. I am happy to walk through the 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 the #7 Global Thought Leader in Retail and a Certified Master Expert in Retail.