Optimizing for AI? Your Customers Aren’t Robots

If you tell a bookstore clerk you want a mystery novel, something not too long and not too gory, they’ll usually come back with two or three paperbacks that fit.

So you turn each one over and read the back. You check which authors are quoted praising it, and maybe you read the first page. One has a cover you’d happily leave on the coffee table as decor (but not read). One is by a writer your friend won’t stop talking about. And the other just sounds like your kind of book.

And yes, while the clerk did find you a couple of books that may interest you, ultimately you make the call.

Shopping online is starting to work the same way. Instead of typing a few keywords into Google and clicking through ten websites, more shoppers describe what they need to an AI assistant like ChatGPT or Gemini, in a full sentence, the way you’d talk to that clerk. The AI comes back with a few products that fit.

Right now, almost every ecommerce brand is optimizing for AI product discovery, which puts their products on the counter right next to yours. So what will it take for your product to be the one that goes home?

On Episode 3 of the DTC Revenue Rocket Podcast, our CEO, Zak, talked through this with a panel that included leaders from Shopify and Klaviyo. He recalled Shopify’s Dwayne Doshier saying he doesn’t see AI agents taking over shopping, and added his own take: AI will “help make that discovery and that decision-making process easier, but it’s gonna need that person to be there to take it across the finish line.”

Getting recommended depends on what AI knows about your product. Getting chosen depends on how a person feels about it: whether they trust it, whether they want it, and whether it seems made for someone like them.

What AI Needs to Know About Your Products

An AI assistant can only recommend a product if it has the details to back it up. When a shopper asks for something specific, the AI needs details that match each part of the request: the price, the size, the use. If you sell on Shopify, your product details are already being shared with AI assistants. Shopify makes your products discoverable in ChatGPT by default and can send the same product information to Microsoft Copilot, Google’s AI Mode and the Gemini app. Shopify also says that the quality of that product information affects whether a brand gets recommended.

Dwayne Doshier made the same point on the panel. For brands showing up in AI shopping, he said, it’s “super important… to have trustworthy content,” covering “not only product, attribute, size, color… price, but how are people using it in the real world?” He went further: “How’s it being reviewed? How safe is it? …Is it safe for my child?”

Take a parent who asks for a stroller that folds small enough for a sedan trunk, works from six months old and costs under $400. One brand’s product page describes its stroller as “lightweight, premium and built for life on the go.” Another lists the folded dimensions, the age and weight range, and the price. Both strollers might meet every requirement, but only the second page gives the AI the facts to prove it.

This is where a lot of AI optimization goes wrong. Adding keywords or making a description sound better doesn’t help if it still leaves out the facts a shopper asked about. As Dwayne put it, the brand that provides this information gets “positioned as more relevant than the brand who isn’t.”

Check Whether Your Descriptions Answer Specific Requests

Pick your best-selling product and write down the most specific request a real customer might make for it, with a budget, a use and a dealbreaker. Your support inbox, your reviews and the terms people type into your site search are good places to find one. We covered how to read those search terms in our post on product discovery

Then go through your product page and see whether it answers every part of that request with a specific fact, like a measurement, an age range or a material. If an answer only shows up in a review or a support reply, add it to the description. Shopify recommends covering key features, technical specs and common use cases.

Check Your Product Attributes

In Shopify, product details can live in two places. One is the description, the paragraph a shopper reads. The other is a set of labeled fields: the product category, the size and color options, and custom fields Shopify calls metafields, which hold details like material or dimensions. Shopify says AI channels favor details stored this way. A field that reads “Material: cotton” leaves no room for misreading the way a sentence can.

So check whether your most important details have their own fields or only appear in paragraphs. Shopify also recommends product and review schema, a type of code that labels the details on your page for search engines and AI tools. This is the same work good SEO for ecommerce product pages has always asked for, and a Shopify developer can set it up across your whole catalog.

Check That Your Details Match Everywhere

Your products probably appear in more places than your own site: a marketplace listing, a Google Shopping feed, maybe a retailer’s page. Shopify keeps price and stock in sync across the AI channels it connects to. The listings it doesn’t manage are yours to keep consistent.

Look up your top product in each of those places and compare the price, dimensions, and materials. If they don’t match, the AI is working from conflicting details. And a shopper who clicks through and sees a different price than the one they were quoted has a reason to doubt you before they’ve read a word.

None of this guarantees a recommendation. Every platform chooses products its own way, and none of them publish exactly how. But complete, consistent details give the AI what it needs to match your product to a shopper’s request, and the shopper who clicks through sees the same facts on your page.

What Shoppers Still Need Before They Buy

An AI assistant rarely answers with just one product. The shopper gets a few that all meet the request, and the facts that got each one there can’t choose between them. The shopper does, even when the checkout happens right inside the chat. 

Gina Potenza, who works on platform and AI partnerships at Klaviyo, doesn’t see that changing. “People aren’t going to stop shopping,” she said on the panel. “We’re evolved monkeys, right? It’s not like the robots are going to totally take over for us. We all want the dopamine of shopping.”

Where that decision happens matters too. In Shopify’s Q1 2026 data, shoppers who came from an AI assistant and landed on a product page converted at nearly 50 percent higher rates than shoppers who arrived from organic search. Shopify calls these early numbers, and they only cover visits that started on a product page. Shopify describes them as shoppers who have already done their research. So your product page has a different job with them. They already know what the product is. They’re deciding whether to buy it.

That decision usually comes down to three questions.

Can I Trust It?

A shopper who comes from an AI recommendation has seen your product described in someone else’s words. Your page is where they check whether that description holds up.

Reviews do most of that work, but only if they say something. Twelve reviews that say “love it!” tell a shopper less than three that mention sizing, how the product held up after a month or what the customer would change. Customer photos help for the same reason, because they show the product outside a studio. We covered where reviews should appear before someone even reaches the product page in our post on product discovery

Then make the basics easy to find: your return policy, shipping times and how to reach you. A shopper who has to hunt for your return policy starts to wonder why.

Do I Want It?

Specs make a product the right one. They don’t make anyone want it.

The parent shopping for a stroller already knows yours fits in a sedan trunk. What makes them want it is watching someone fold it with one hand while holding a coffee, or seeing it next to a car seat for scale. So show the product in use, at its real size, in the kind of place your customers live. Add a line or two about what it’s like to use, not only what it’s made of.

A shopper who can picture using your product is much closer to buying it.

Why This One?

Sometimes two products match on every spec and cost about the same. Then the choice comes down to the brand: who made it, what they stand for, how it looks and whether people like the shopper already buy from them.

Gina flagged a risk here. Talking about brands using AI in their marketing, she warned that the results can feel “the same as all your competitors,” because “they are likely using the same AI models.” Product pages are no exception. If yours were written with the same tools as everyone else’s, they’ll read like everyone else’s, and a shopper has nothing to pick you for.

Try this: open your product page next to two competitors’ pages and cover the logos. If you can’t tell which page is yours, a shopper can’t either. That’s a sign your website design and copy need to show more of what sets your brand apart: the story behind the product, the choices you made that others didn’t, and the customers who already love it.

Will AI Shopping Replace Google

Not anytime soon. In Shopify’s Q1 2026 data, organic search still sent its merchants more visits than every AI platform Shopify tracked, combined. Shopify notes that AI’s share is probably a little higher than it looks, since some visits from Google’s AI Overviews get counted as organic search. For all the attention on artificial intelligence in ecommerce, search is still where most shoppers start. 

That means you don’t have to split your effort between the two. The specific, consistent product details that help AI match your product also help Google rank your product pages, so the same work pays off in both places. And paid search still puts you in front of people looking for exactly what you sell, so pulling back from it now would mean giving up shoppers you can already reach.

To know when AI deserves more of your time and budget, watch the numbers you can see. Orders that come from ChatGPT show up in Shopify labeled with where they came from. If that share keeps climbing, put more into your product information. If it stays small, keep improving your product information alongside your other channels instead of chasing every new AI feature as it launches.

The first order is where AI can help most. The second order shouldn’t depend on an AI assistant choosing you again. Once someone buys, bringing them back is the job of your email and SMS: a thank-you that sounds like your brand, a reorder reminder timed to when they’re likely to run out, recommendations based on what they bought. The AI may have made the introduction, but the relationship is with the person, and keeping it is up to you.