AI is changing how shoppers find your products

AI is already reshaping how shoppers discover, compare, and purchase products. Here’s what’s happening, why it matters, and what WooCommerce merchants should know right now.

If you sell online, the way your customers find you is changing.

Half of all consumers now use AI when searching the internet, according to McKinsey. Shoppers are asking ChatGPT for gift ideas, using Google’s AI summaries to compare products, and turning to recommendation engines instead of scrolling through search results.

For now, most of this activity is about discovery — people using AI to research what to buy. Google is exploring discovery and agentic commerce capabilities in Gemini. Visa, Mastercard, and American Express are all actively developing ways for AI agents to transact on behalf of consumers.

McKinsey projects that AI-driven commerce could influence up to $1 trillion in US retail revenue and $3–5 trillion globally by 2030. Whether that number proves accurate or not, the direction is clear: AI is becoming part of the path to purchase, not just a research tool on the side. And the earlier you understand how it works, the better positioned your store will be.

AI in commerce isn’t one thing. It ranges from features you might already use, like AI-generated product descriptions or smarter search, all the way to autonomous agents that can find, compare, and buy products on behalf of a shopper.

Here’s a practical way to think about this spectrum:

AI-assisted tools

These are what most merchants encounter first. ChatGPT or Claude helps you write product descriptions. AI-powered search on your store helps shoppers find what they want faster. Email platforms use AI to personalize subject lines and send times. These tools make existing workflows more efficient. You’re still doing the work, but faster.

AI-enhanced browsing

This is what shoppers increasingly experience on the other end of shopping. Search engines now offer AI-generated summaries alongside traditional results. 

When someone searches for “best ceramic coffee mug for a gift,” they may see an AI-curated answer before they see any individual store. Social shopping platforms are using AI to make product discovery feel more conversational and personalized.

AI-optimized processes

Behind the scenes, AI is also transforming operations. Retailers are using it to forecast inventory demand, automate customer service workflows, and optimize pricing. AI assistants are helping merchants run their businesses as much as they’re helping shoppers find products.

AI agents

AI agents are the next stage, and the one generating the most attention right now. An agent can do more than suggest an answer. It can take action: search across stores, compare options based on specific criteria like price and shipping speed, and in some cases, complete a purchase on behalf of the shopper. 

If that sounds futuristic, consider that voice assistants like Alexa have been completing Amazon orders for years. The difference now is that AI agents are becoming capable of agentic commerce: making purchases across any store, not just within one platform’s ecosystem.

Agentic commerce is what happens when AI agents do more than help people shop. They shop for them. It’s the shift from humans browsing stores to AI agents shopping across them, finding products, comparing prices and policies, and increasingly completing transactions on a shopper’s behalf.

In practice, it works like this: a shopper tells their AI assistant, “Find me a fair-trade ceramic mug under $30 with free shipping to NY.” The agent searches across multiple stores, evaluates options based on the shopper’s criteria, and presents the best matches. The checkout still happens on the store’s website, but the AI assistant helps with a targeted discovery process and soon, it might complete the transaction on the shopper’s behalf.

This changes the fundamental mechanics of how products are found. 

Does this impact your business? Where AI shopping will take hold first, and where it won’t

Consumers aren’t ready for big-ticket agentic commerce 

Not all purchases are equally suited for AI agents. The type of product, the price, and the shopper’s existing habits all shape whether an AI assistant adds real value or gets in the way of a purchase.

At the high end, shoppers aren’t ready to hand over big purchases. Research from Checkout.com found that U.S. consumers are comfortable letting AI spend an average of $233 on their behalf — and that’s among people already open to the idea. 

For expensive, complex purchases, shoppers want AI to research and compare, but they want to make the final call themselves. The trust isn’t there yet, and the stakes around returns and refunds are too high.

At the low end, routine replenishment is already locked up. If you reorder the same coffee pods every month, Amazon likely already knows your brand, your schedule, and your payment method. An AI agent offering to handle that is competing against a one-click system with years of purchase history. There’s no friction left to remove.

Comparison shopping presents a major opportunity

The real opportunity is in the middle: the considered purchase. This includes purchases where shoppers know they need something, but haven’t decided exactly what. The real opportunity is in considered purchases or products where shoppers know they need something but haven’t decided what. Think trail running shoes under $120 or a standing desk that actually fits a small apartment. These are the categories where specs, reviews, and comparisons drive the decision, and where an AI agent can do useful work.

The potential multiplies when the agent combines products across stores. Imagine a shopper asking an AI assistant to put together a complete backpacking kit for a March trip in the Pacific Northwest. Tent, layers, cookware, pack, all within a budget, all arriving within two weeks, all suited for cold rain. No single search query handles that today. 

An AI agent can pull from multiple merchants, match specifications across products, and assemble a complete recommendation that works as a set. Every specialty retailer with strong product data becomes a potential part of that answer.

Structured product data will provide a clear advantage

For independent stores, the opening here is significant. Amazon has blocked OpenAI’s crawlers from accessing its product listings, meaning hundreds of millions of Amazon products are invisible to ChatGPT. 

Meanwhile, ChatGPT product results are organic and unsponsored, ranked on relevance to the shopper’s query, not ad spend. A specialty retailer with clean, structured product data can surface ahead of mass-market competitors in a way that traditional search has never allowed.

The pattern is already visible: AI agents add the most value where there’s real comparison to do, where the shopper doesn’t already have a default, and where the price justifies research but doesn’t require a leap of faith. That’s the sweet spot for the types of products many WooCommerce merchants already sell.

As AI agents become capable of searching, comparing, and, eventually, purchasing on behalf of customers, a practical question emerges: How do these systems actually interact with real stores?

AI agents need several things to work with a store, including:

• Access to structured product data.
• Permission to take actions like adding items to a cart.
• A secure way to process payment.
• A shared set of rules for how to communicate with your store’s systems.

A shared set of rules comes in the form of protocols: standards that define how systems exchange information, what actions are allowed, and how trust is maintained. Just as payment APIs made online checkout scalable, AI-driven commerce depends on new coordination layers between agents and stores.

Model Context Protocol (MCP): Giving AI real-time access to your store

Model Context Protocol, introduced by Anthropic, defines a standardized way for AI models to securely access external systems like product catalogs, inventory databases, and order management tools.

Large language models are powerful, but by default they generate responses based on training data. They don’t automatically know your current inventory levels, updated pricing, or whether a product is in stock. 

MCP creates a bridge between the AI model and your live store data so that instead of guessing, the AI can pull real information before making a recommendation.

Without structured access to live data, AI assistants approximate. With it, they can check product availability, specifications, pricing, and order status before generating a response.

MCP is quickly becoming a shared standard across major AI providers because it reduces the need for custom connectors between every model and every tool. It turns your store from a static website into something an AI can reliably read and work with.

Agentic Commerce Protocol (ACP): Connecting merchants to AI-driven discovery

ACP, developed in collaboration between OpenAI and Stripe, defines how agents can surface products, build carts, and connect shoppers to merchants through ChatGPT.

OpenAI shifted its commerce strategy away from checkout directly in ChatGPT (Instant Checkout) and toward a discovery-first model. ACP remains the underlying structure connecting ChatGPT to merchant catalogs and payment systems, but in a narrower, more practical scope. ChatGPT will focus on product search, comparison, and recommendation, with purchases handled through connected retailer apps or redirected to merchant websites directly. OpenAI is relying on platforms like WooCommerce to help deliver the best experience for shoppers.

For stores, this means ChatGPT is already a discovery channel today. Products can be found, compared, and recommended inside a chat session. The transaction completes in the store’s environment and checkout. Supporting ACP means your products will be visible within the ChatGPT ecosystem. 

Universal Commerce Protocol (UCP): Enabling purchases across Google and beyond

Universal Commerce Protocol, introduced by Google, takes a broader approach to the same opportunity.

UCP provides a standardized way for AI agents, including Google’s Gemini and AI Mode in Search, to discover what a merchant offers and complete purchases accordingly. Merchants expose structured product and store capabilities through the Google Merchant Center. Agents read that information and can surface products to shoppers during conversational research, with checkout available directly on eligible Google listings..

Where ACP is narrowing toward discovery and app-based handoffs, UCP is actively expanding agentic commerce on Google’s own channels. Both protocols treat the merchant as the seller of record.

How these protocols relate and how to prepare for the future

These protocols are not mutually exclusive. They represent different AI ecosystems building different models for how agents interact with commerce, much like different payment networks represent different financial rails. ACP is evolving toward discovery and then handoff. UCP is building toward discovery and checkout directly on Google Search. Both are early, and both will continue to change.

Supporting one enables participation in one environment. Supporting another extends your reach to a different set of shoppers. For now, each protocol represents a distinct channel with a distinct approach to where the transaction happens.

Together with MCP, these three layers form the emerging infrastructure for AI-driven commerce. MCP enables visibility so AI can see your store. ACP and UCP enable reach so AI can surface your products to shoppers across these distinct ecosystems. Merchants who are accessible through these protocols become discoverable by AI shopping agents. Those who aren’t will be harder to find as these channels grow.

AI has been part of commerce for years. Recommendation engines and predictive analytics aren’t new. What’s different now is the convergence of three things at once:

1. AI that can reason and act, not just recommend. Today’s models can hold context across a complex shopping query, compare options on multiple criteria, and complete multi-step tasks.
2. Protocols that connect AI agents to real stores. MCP, ACP, and UCP are turning theoretical agentic commerce into working infrastructure.
3. Major platforms building AI directly into shopping. OpenAI, Google, Stripe, Visa, Mastercard, and others are investing heavily in making AI-assisted purchasing work at scale.

The window to prepare is now: infrastructure is being built, consumer behavior is shifting, and the potential is mounting. This combination of capable AI, real protocols, and platform-level investment is what makes this moment different from previous waves of AI hype. 

Woo’s AI priorities are clear: Merchants should benefit from this shift, not lose control because of it.

AI should help surface your products to more shoppers, automate the tasks that eat up your time, and open new sales channels. It should do all of that without centralizing power away from you and your store.

How you participate in AI-driven commerce matters. 

To connect your business to AI agents, two models are emerging:

Centralized catalogs are where you upload your product data to a third-party platform that serves it to AI agents on your behalf. This can be fast to set up, but it means someone else controls how your products are described, presented, and sold.

Direct access is where AI agents connect to your actual store, read your live data, and transact through your checkout. You keep control of your product information, your customer relationships, and your margins.

WooCommerce helps you build the foundation for rapid change

WooCommerce is built on WordPress open-source infrastructure that powers millions of stores globally. That openness is directly relevant in the AI era because it means your store can be made readable and actionable by any AI system through open standards, not through a single platform’s proprietary integration.

Woo is actively building on the emerging protocols that make this possible. This includes MCP integration so AI tools can connect directly to WooCommerce stores. It also includes ACP and UCP development so Woo stores are discoverable through Google’s and OpenAI’s surfaces.

We’ll share more on these integrations as they ship. 

In the meantime, the single most impactful thing you can do right now is make your product data as clean, complete, and structured as possible. 

That’s the foundation every protocol depends on, and it’s work that pays off regardless of which standards ultimately win.

Whether AI agents find your products through ChatGPT, Gemini, or a tool that doesn’t exist yet, what they’re looking for is the same: accurate product information, clear policies, and a store they can reliably transact with.

The merchants who get that right now will be the ones AI recommends first.

Ready to get started? 

Try this right now: Go to your preferred chat assistant, like ChatGPT or Claude, and search for one of your top products by describing it the way a shopper would. Does it recommend your business, or a competitor’s?

About

Julia Callicrate

Julia Callicrate leads Product Marketing at Woo, helping global businesses discover new ways to scale their stores with the WooCommerce platform. A storyteller at heart, Julia bridges the gap between products and customers. Prior to her work at Woo, Julia had spent over a decade as a product manager and marketer in healthcare, banking, and SaaS industries. She currently lives in Virginia with her family. Julia also has two cats that always seem to pop up on her video conferencing.