By: Kavan Sohal Apr 29/2026
AI is changing how people research and buy products, and it’s happening faster than most marketers or businesses can keep up with. Whether you’re a fan or not, platforms like ChatGPT, Perplexity, Claude, and Gemini are shifting from simple answer engines into something that looks a lot more like a shopping experience. Users can now ask an AI for product ideas, compare options, and, in some cases, check out without ever visiting a traditional website.
This shift is part of a much bigger trend of AI in retail. From product discovery to personalized recommendations to checkout, AI is working its way into every stage of the buying journey. But is AI shopping actually the future, or is it just hype? Let’s break it down.
Will AI Replace Google or Amazon for Shopping?
We’re not there yet, and we may not fully get there. What’s more likely is that AI gets embedded inside major retail and search platforms rather than replacing them outright. Google is already doing this with AI Mode. Walmart tested it. Shopify is building toward it. The direction we’re seeing right now is integration, not replacement.
That said, AI is genuinely useful during the discovery phase. If you’re looking for “a maroon blazer under $200 for a winter wedding” and Google isn’t quite pointing you in the right direction, ChatGPT and similar platforms are where things start to shine. They’re great at helping you figure out what you want before you commit to where you buy it.
The current limitations? AI platforms are still not visual enough. You cannot scan a grid of product images the way you can on Amazon, a retailer’s site, or even through Google Shopping. And when it comes to properly researching a product, specs, reviews, Q&A, and return policies still matter, and most people want to see that information on the actual product page. User-generated content, from customer photos and videos to honest reviews, still plays a major role in building trust and driving purchases. That is why Reddit, YouTube, TikTok, and other social platforms remain so important: people want real opinions from real users, not just corporate speak.
How AI Shopping Assistants Are Changing Product Discovery
If you know exactly what you want (a specific model, a repeat purchase), you’re going straight to Google, Amazon, or the brand’s site, or performing a quick price check. That’s not changing anytime soon, and why would it?
But if you’re exploring, if you’re not sure about the exact product and you need help narrowing things down (gift ideas, outfit planning, niche hobby gear, buying a new TV), that’s where AI shopping assistants are starting to pull ahead. They can clarify your requirements through conversation and find products from niche or lesser-known retailers based on what actually matches your needs, not just who has the biggest ad budget or most recognizable brand. As we know, the larger brands aren’t always the best brands for our needs or the products we are interested in.
This is why AI-powered shopping could end up being a real advantage for smaller retailers whose products match specific intents, even if they don’t dominate traditional search results.
ChatGPT Shopping: What Walmart and Shopify Tell Us
This is where things get interesting, and where we can see just how early we still are in AI for shopping.
Walmart Pulled Back After 3x Worse Conversion
Walmart plugged a large catalogue of products into OpenAI’s Instant Checkout, letting users discover and buy items directly inside ChatGPT without visiting walmart.com. The results were not great. According to Search Engine Land, purchases completed entirely inside ChatGPT converted at roughly one-third the rate of transactions where users clicked out to Walmart’s own site. Daniel Danker, Walmart’s EVP of Product and Design, called the in-chat experience “unsatisfying,” and Walmart is stepping away from that model, at least for the time being.
That doesn’t mean AI shopping is dead. What it tells us is that moving checkout into the chat window hurts conversion when the user experience can’t match the visuals, trust signals, and familiarity of a mature ecommerce site. Notably, Walmart isn’t abandoning AI altogether. They’re keeping their AI assistant Sparky but routing checkout back to walmart.com, which tells you exactly where the industry is landing: AI for discovery, owned properties for conversion. For most brands, skipping straight to full in-chat checkout is likely premature. The smarter move is to start where Walmart ended up.
Shopify is Betting on AI Distribution
Shopify is taking almost the opposite approach. They’ve partnered with OpenAI so that merchants can sell through ChatGPT conversations. Shopify feeds real-time product data (pricing, inventory, images, variants) into ChatGPT, making hundreds of millions of products discoverable in a format AI can reason about.
To be clear, this doesn’t mean every transaction happens entirely in the chat window. When someone asks ChatGPT for a product recommendation, AI can pull from Shopify merchants and present purchase options, but the actual checkout is still powered by Shopify’s infrastructure. Orders flow back into the merchant’s Shopify admin, with Shopify acting as a layer between AI agents and store owners.
So the picture right now: Walmart pulled back from pure in-chat checkout after seeing poor conversion, while Shopify is betting that getting millions of merchants into AI conversations will pay off over time. Both experiments tell us something valuable about where ChatGPT shopping is headed.
Google Shopping AI: Direct Offers, Gemini, and What it Means for SEO
Google is also moving fast to blend AI assistance with shopping and ads, and this is where it gets especially relevant for SEO and performance marketing.
AI Mode and Direct Offers. Google is piloting “Direct Offers,” a new ad format that lets retailers surface exclusive promotions, such as a percentage discount, inside AI Mode when Google detects strong purchase intent. AI Mode is part of Google’s broader Gemini-powered search experience, and these shopping features show where Google is heading using AI not just to answer questions, but to shape purchase decisions in real time. Rather than functioning like a traditional search ad in a standard results page, Direct Offers appear to be triggered contextually within the AI shopping experience.
The retailer opportunity. This creates some interesting dynamics for brands. Retailers can define promotions and discounts, then observe when AI-assisted offers actually help move users toward conversion. Over time, that could give brands better insight into when discounting is necessary, when shoppers may convert without it, and how promotional strategy can shift based on intent signals. It adds a new layer to offer strategy and performance analysis.
And this is not just a paid story. Google is also pushing deeper into agentic commerce through its Universal Commerce Protocol, an open standard built to let AI surfaces like AI Mode and Gemini connect directly to retailer systems for product discovery, cart actions, and checkout. That means stronger product feeds, cleaner structured data, better content, and stronger review signals become even more important. If your product data is thin or untrustworthy, AI has less to work with, which can weaken both organic visibility and participation in emerging AI-driven shopping experiences.
The SEO Playbook for AI Shopping
This is the part most people are missing. Everything I described above runs on data, and that data comes from how well your products are represented on the web. If AI can’t understand your products, it can’t recommend them. Period.
Schema markup matters more than ever. Structured data is how AI engines understand what your product actually is, what it costs, whether it’s in stock, and how it compares to alternatives. If you’re not implementing product schema with detailed attributes, you risk not being found by these systems and losing conversions simply because the AI didn’t have enough detail to recommend you versus a competitor’s products.
Product descriptions need to do real work. Generic copy won’t cut it when an AI is trying to match a user’s specific request. Your descriptions need to include the details that actually help someone make a decision: the brand name, the specific colours, the stitching, the fabrics or materials used, the weight, and the dimensions. If someone asks an AI for “a hand-stitched Italian leather crossbody bag in cognac,” the products with those exact details in their descriptions are the ones that get found. Intricate details make all the difference.
Alt tags on images need to be unique and descriptive. AI models are increasingly pulling from image metadata to understand products. A generic alt tag like “brown bag” does nothing. “Cognac Italian leather crossbody bag with brass hardware and adjustable strap” gives the AI something to work with. And this isn’t just about bots. Descriptive alt text also improves accessibility for users with screen readers, so it’s a win on both fronts.
Clear, high-quality product images. This one seems obvious, but unique product photography matters even more in an AI context. When AI platforms do show visual results, the quality of your images becomes a trust signal in itself.
Fit guides and comparison content. This is where you can really pull users in from AI discovery. If someone asks ChatGPT, “What size should I get in [brand]?” and you have a detailed fit guide on your site, that content becomes part of the AI’s answer. Comparison guides work the same way. This is exactly the type of content that tends to show up in AI answers, even when a product page doesn’t. Content that helps users evaluate and decide is what AI platforms are looking for to build their recommendations.
What Should Marketers Do Now?
The early data points suggest a few practical takeaways:
- Expect AI chat to grow as a discovery layer, especially for non-specific and complex shopping needs, but don’t expect it to replace search or strong ecommerce sites any time soon.
- Treat experiments like Walmart’s as signals that in-chat checkout user experience still needs work, not as proof that AI in online shopping doesn’t convert.
- If you run on Shopify, start planning now for how your product data, images, and naming conventions will look when found in an AI conversation, not just on a product page.
- Invest in your product feed quality, structured data, and on-page content. These aren’t just SEO best practices anymore. They’re the foundation of whether AI can find and recommend your products at all.
- Watch Google’s AI Mode and Direct Offers closely. They will likely reshape how product feeds, promotions, and SEO interact in AI-first experiences.
AI shopping is still developing, but the direction is clear. The assistants that help people think through what to buy will increasingly handle how to buy as well, whether that happens inside the chat window, on a retailer’s site, or somewhere in between. The brands that prepare their product data and content now will be the ones AI actually recommends.
The next wave of this shift goes beyond discovery into personalization, sizing, virtual try-on, and the way prices are being set in real time. But it all starts with the fundamentals covered here: clean product data, strong on-page content, and a structure AI engines can actually understand.
Want to learn more about shopping in AI platforms? Stay tuned for part two.
About the Author
Kavan Sohal