Your next customer might be an AI—are you ready?

5 min read

(July 27)

For decades, retail operated under an unspoken pact. If you had a good store—whether brick-and-mortar or online—customers would come. Foot traffic for brick-and-mortar stores, Google Shopping clicks for online stores, and marketplace algorithms for those who found both of those too much of a hassle. Just provide a decent experience, and the system would send you buyers.

That model is now under pressure—from two sides at once.

The old balance has been disrupted

On the one hand, over the past decade, e-commerce has trained consumers to compare everything. Price, delivery time, availability—all visible at a glance. Anyone selling generic products at the standard price loses the battle before the customer even walks into the store or opens the product page.

On the other hand, the way people search is fundamentally changing. In 2025, 58.5% of all U.S. Google searches ended without a single click to a website. AI-powered searches provide immediate answers. The consumer asks, “What are the best pants for a casual office look under €100?” and receives a summary, recommendations, and perhaps even a direct purchase link.

There’s no clicking. There’s just a conclusion.

What’s becoming commoditized is generic product information: standard specifications, basic descriptions, and lists with no added value—exactly what most retail product pages consist of.

But that’s just the beginning.

The agent-driven commerce era has begun

Imagine: a customer asks his AI assistant where he can pick up size 43 hiking boots today near Ghent. The agent doesn’t search on Google. It consults multiple sources simultaneously, checks real-time availability by location, compares prices, and provides a single answer with a direct reservation link.

Will your store be mentioned? Or does it not even exist for that agent?

This isn’t just a pipe dream. Visa and Mastercard have already developed agentic commerce protocols: In April 2025, Mastercard launched Agent Pay, which allows AI assistants like ChatGPT or Microsoft Copilot to complete a purchase without ever showing the customer a checkout page. Visa followed with the Trusted Agent Protocol and predicts that millions of consumers will already be using it during the 2026 holiday season. . Platforms like Perplexity and OpenAI are rolling out shopping agents that make purchases directly on behalf of the user. Between January 2025 and January 2026, the number of orders placed via agentic chats already grew by a factor of 14.

The traditional marketing funnel—awareness, consideration, conversion—is losing its meaning as consideration and conversion take place within a model that humans never see.

The question for every retailer is thus fundamentally shifting:

“Does the agent trust me enough to make purchases through me?”

What retains value in this new landscape

First, product information has value when it contains original, non-synthetizable input. Real inventory data per location, up-to-date prices, product descriptions enriched with in-store expertise, and customer reviews with specific use cases. These are the signals an AI system needs to make a reliable recommendation. Generic catalog data from the supplier, without its own context, isn’t cited. It’s ignored.

Second, product information has value as source material, not just as a destination page. A product page can create value even if no human ever reads it line by line—as long as it’s the page that an agent references, compares, or uses to complete a transaction. AI traffic to retail sites grew by 393% year-over-year in the first three months of 2026 alone. Being accessible to the agent increasingly means being accessible to the buyer.

Third, channel experience retains value when it is linked to trust and action. When a customer buys an expensive device, wants to exchange a gift card, or needs to arrange a return, they want more than just a summary. They want a person, a store, a story. AI can condense information, but it cannot replicate an in-store experience.

From Customer Experience to Agent Experience

We’ve invested years in CX—customer experience. A tidy store, a fast online store, a smooth checkout process. Now a second audience has emerged: the agent.

Agent experience is how easily an AI system can interpret your store, trust it, and act through you. In retail, this boils down to four concrete criteria:

Readable: Is your product data structured enough—with the right attributes, consistent taxonomy, and a clean feed—for a machine to work with it? Poor data foundations are the biggest blind spot in traditional retail. Research by Adobe shows that retail sites lag significantly behind in AI search visibility, precisely because their product data isn’t set up to be machine-readable.

Citable: Does your store have a clear identity, real-time availability data, and information that an agent can cite? “In stock in Ghent, available for pickup today before 5 p.m.” is citable. “View our collection” is not.

Trustworthy: Does your brand have enough of a digital presence—reviews, consistent product information, reliable omnichannel behavior—for an agent to recognize you as a trustworthy source? Trust isn’t something you build in a single campaign.

Actionable: Can an agent complete a task with you—check inventory, make a reservation, place an order—using standardized protocols? This is the criterion where the bar is set the highest, but also where the difference is the greatest.

The Omnichannel Foundation as a Strategic Weapon

Here lies the paradox facing many retailers: those who have invested in a strong omnichannel foundation in recent years—a single, up-to-date inventory, centralized customer data, consistent product information across all channels—have unwittingly built the best starting position for the agentic era.

After all, the four elements of agent experience aren’t new. They’re simply a new name for what omnichannel already required:

A single data source for products and inventory → transparent
Real-time availability by location → quotable
Consistent prices and promotions across channels → trustworthy
A point-of-sale system connected to e-commerce, loyalty programs, and reservations → actionable

Retailers who have this in order are well-positioned not only when a human visits the online store, but also when an agent does.

The question is whether your infrastructure can handle it yet.

That’s exactly what Tilroy helps retail customers build: a single platform that connects inventory, point-of-sale, online store, and customer data across all channels. Structured product feeds that are machine-readable, real-time inventory information per location that you can access via API today, and a point-of-sale environment that supports integrations with the platforms where agentic commerce is evolving. No vaporware. These are the building blocks that retailers already use every day—and that will form the foundation for agent-ready retail tomorrow.

Is your store already agent-ready?

Check the four criteria against your own situation:

Readable — Is your product data consistent, complete, and synchronized across channels?
Citable — Does your online store or platforms display real-time inventory by location?
Trustworthy — Are your prices, promotions, and availability the same everywhere?
Actionable — Can external systems check your inventory or place a reservation via an API?

More than two “no” answers means your foundation needs attention—not as a future project, but right now. Agents are already starting to make their choices.