GEO for Ecommerce Sites: The AEO Playbook to Get Your Products Recommended by AI

GEO for an ecommerce site means getting your products named, linked and recommended when shoppers ask ChatGPT, Gemini or Google's AI Mode what to buy. This is a 9-step playbook for a Shopify store, in the order we'd run it, starting with the step nobody else mentions: pick the products worth being recommended for, by contribution margin, before you touch a single page.

If the terms are new, read our guide to AEO for ecommerce first. AEO (answer engine optimisation) and GEO (generative engine optimisation) are two names for the same job, and the difference between AEO and SEO is smaller than you'd think: most of the work below also helps you rank in Google.

The channel is small but it converts. Shopify's Q2 2026 data shows AI-referred sessions to Shopify stores up 197% year on year, with half of them landing straight on a product page, and those shoppers converted about 80% better than organic visitors once they got there. Organic search still sent more sessions than every AI platform combined, so this is a second front, not a replacement.

StoreHero AI Search runs this playbook for you. It asks ChatGPT, Gemini and Google AI the questions your buyers ask, shows who gets named instead of you and which sources they cite, then writes the pages to close the gap. StoreHero already knows the contribution margin of every product, so the work goes where the money is. Book a demo to see it on your store, or follow the 9 steps below to do it yourself.

What is GEO for an ecommerce site?

GEO for an ecommerce site is the work of making your products the ones AI assistants name when a shopper asks for a recommendation. It covers your product pages, product feeds, structured data, reviews and what other sites say about you.

In short, an assistant answers from what it learned in training plus what it searches live. For a shopping question it runs several searches at once (Google calls this query fan-out), pulls listings from product feeds, reads pages, reviews and forums, then names the products most sources agree on. We break that down in how ChatGPT recommends products. You can't buy your way into the answer, but you can make your products easy to find, understand and agree with.

Step 1: Rank your products by contribution margin

Being recommended for a low-margin bestseller is worth less than being recommended for a mid-volume product that makes real money, so every hour of copy, comparisons and outreach should go where the dollars are.

Contribution margin is what an order leaves after product cost, shipping, payment fees, returns and the marketing that won it (here's the contribution margin formula). AI-referred orders often arrive with no ad cost attached, so the gap between a thin-margin product and a fat one gets wider, not narrower.

Here's an illustrative example with made-up numbers for a clothing store. Assume AI answers could send each product some extra orders a month, based on how often shoppers ask about its category:

The hoodie wins the most orders and earns the least. The merino jumper gets 40% of the hoodie's orders and more than twice the margin dollars, so it goes first in everything that follows. The score is expected AI orders × contribution margin per order. Shortlist your top 10 to 20 products and categories this way, and leave out anything about to sell out.

You need contribution margin per product for this, not gross margin. StoreHero already calculates it for every product from your Shopify orders, costs and ad spend, and the free product profitability report prompt ranks your catalogue in Claude or ChatGPT in a few minutes. A spreadsheet and the contribution margin calculator work too, just more slowly.

Step 2: Rewrite product and collection pages answer-first

On each shortlisted product page:

• Open with one or two sentences that say what it is, who it's for and why it beats the alternatives: "A 100% merino crew-neck jumper for people who run cold, machine washable at 30°C".
• Put the specs in text, not just in images: material, dimensions, weight, fit, care, ingredients, compatibility and what's in the box.
• Answer the 3 to 5 questions your support team gets most, on the page itself.
• Say who it's not for. That honest line helps an assistant match a product to a buyer.

Collection pages are your answer to "best X for Y" prompts. Replace keyword filler with a short intro that answers the question and compares the top 3 products.

Step 3: Check your structured data in the Rich Results Test

Structured data is the machine-readable label on each product: name, price, currency, availability, brand, GTIN and ratings. Google says plainly that there's no special schema needed for AI features, but the same markup feeds Google Shopping and lets every crawler read your product facts.

Most Shopify themes output product markup through the structured_data Liquid filter, which gives you a Product (or a ProductGroup for products with variants) with an Offer carrying price and availability. What's often missing is ratings, GTINs, shipping and returns. To check yours:

1. Open Google's Rich Results Test and paste the URL of one shortlisted product page.
2. Look for "Merchant listings" and "Product snippets" in the detected items, and fix anything marked as a critical error.
3. Check that price, availability and rating match what the page shows. Google's merchant listing documentation lists GTIN, shipping details and return policy as recommended properties, so fill them where you can.
4. Repeat for a product with variants and for a collection page.

Don't expect FAQ markup to win anything visible in Google: since 2023 Google only shows FAQ rich results for well-known government and health sites. Also check robots.txt doesn't block OAI-SearchBot. OpenAI says sites that block it won't appear in ChatGPT search answers.

Step 4: Send clean product feeds to every AI shopping surface

Feeds give assistants price and stock without reading your page. Set up three:

• Google Merchant Center. Google Merchant Center feeds the Shopping Graph, which Google says holds more than 50 billion product listings and powers shopping in AI Mode. Shopify's Google & YouTube app syncs your catalogue, but check that titles, GTINs, colour, size and material are filled, because those attributes are what AI answers filter on.
• Shopify Agentic Storefronts. Shopify's Agentic Storefronts syndicate your catalogue to ChatGPT, Google AI Mode and Microsoft Copilot, and it's switched on automatically for eligible stores. Check it's on in your admin.
• ChatGPT's product feed. OpenAI publishes a product feed spec with 9 required fields (ID, title, description, URL, brand, seller, image, availability and price) plus optional ones such as GTIN, return policy, review count and star rating. It sits on the open Agentic Commerce Protocol, built with Stripe. OpenAI scaled back Instant Checkout in March 2026 to focus on product discovery, so treat this as a visibility feed first.

If you're also weighing up ChatGPT ads, the same clean feed serves both.

Step 5: Build comparison and FAQ pages for real shopper questions

Shoppers ask assistants comparison questions: "X vs Y", "best X under $100", "is X worth it". If only a competitor or affiliate page answers them, they write the answer. For each shortlisted category, publish:

• One honest comparison page against the 2 or 3 alternatives shoppers actually consider, with specs, prices and a clear "choose X if, choose Y if".
• One buying guide that answers the category question in its first paragraph.
• An FAQ built from support tickets, reviews and on-site search terms.

This is the content StoreHero AI Search writes for you: it finds the prompts where a competitor gets named instead of you, then drafts the researched comparison or buying guide that answers them, ready to publish.

Step 6: Get reviews that say something specific

Assistants read reviews on your site and across the web, so what the reviews say matters as much as the stars. "Great product!" helps nobody, while "fits true to size, still soft after 20 washes" answers a buyer's question. Ask for reviews with prompts about fit, use case and durability, reply to bad ones, and check your review app adds ratings to your structured data and review counts to your feeds.

Step 7: Earn mentions on Reddit, in press and in roundups

Agreement across sources is what tips an assistant toward naming you. Your site is one vote; Reddit threads, gift guides, "best X" roundups, YouTube reviews and trade press are the rest. In margin order:

• Get your shortlisted products into the roundups that already rank for your category by sending samples to the writers.
• Answer real questions on Reddit openly as the brand, where your product is a fair answer. Don't astroturf, because moderators and models both spot it.
• Pitch one data-led or founder story a quarter to trade press.

Step 8: Should you add llms.txt?

llms.txt is a proposed file that gives AI models a curated map of your site. Some SEO tools now nag you for it, but there's little evidence anyone reads it: Google's AI features guide says you don't need new AI text files, and Search Engine Journal found that no AI platform has adopted it. Add one if your theme or an app makes it a 2-minute job, then spend the time on steps 1 to 7.

Step 9: Measure AI visibility in dollars, not mentions

Share of voice tells you how often you're named. It doesn't tell you whether the products being named make money, which is the same trap as judging ads on ROAS instead of profit. Track four things monthly:

1. Prompt checks. Ask ChatGPT, Gemini, Perplexity and Google AI Mode 10 to 20 real buyer prompts for your shortlisted categories, and log which products appear.
2. Platform reports. Merchant Center's AI performance report shows your share of voice in AI Mode and AI Overviews (English-language queries, US and a few other markets), Shopify's admin reports Agentic Storefronts sales, and Bing Webmaster Tools' AI Performance report shows Copilot citations. Search Console counts AI Overviews traffic under "Web".
3. AI-referred orders. In GA4 or Shopify, filter referrals from chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com.
4. Contribution margin from those orders. Total the margin of AI-referred orders by product and compare it with your step 1 shortlist.

That last number is the one that matters: if assistants keep recommending the $9-margin hoodie and ignoring the $48 jumper, steps 2 to 7 for the jumper are your next sprint.

StoreHero AI Search covers the first check for you, tracking your visibility, placement and share of voice across ChatGPT, Gemini, Copilot, AI Overviews and AI Mode. Because StoreHero already has the contribution margin of every product, you can see whether the products AI names are the ones that make you money.

Key takeaways

• Start AEO with the products that make the most contribution margin: priority = expected AI orders × contribution margin per order.
• Rewrite shortlisted product and collection pages answer-first, with specs in text and the questions shoppers actually ask.
• Most Shopify themes output Product and Offer markup already; add ratings, GTINs, shipping and returns, and check it in Google's Rich Results Test.
• Clean feeds matter as much as pages: Google Merchant Center, Shopify Agentic Storefronts and OpenAI's product feed spec.
• Reviews, Reddit, roundups and press decide whether assistants agree on you; llms.txt is optional at best.
• Measure AI visibility in dollars: the contribution margin of AI-referred orders, by product.

Frequently asked questions

Sources

• Shopify: AI and organic search are doing different jobs (Q2 2026 data)
• Google Search Central: AI features and your website
• Google: Shop with AI Mode (Shopping Graph, query fan-out)
• Shopify Liquid: structured_data filter
• Google Rich Results Test
• Google Search Central: Merchant listing structured data
• Google Search Central: Changes to HowTo and FAQ rich results (2023)
• OpenAI: Overview of OpenAI crawlers
• Shopify: Agentic Storefronts
• OpenAI: Product feed spec
• Agentic Commerce Protocol
• Modern Retail: Shopify says purchases are coming inside ChatGPT as OpenAI retreats on Instant Checkout
• llmstxt.org: the llms.txt proposal
• Search Engine Journal: LLMs.txt for AI SEO
• Google Merchant Center Help: About AI performance insights
• Practical Ecommerce: Bing's improved AI Performance section