GEO and AIO for online shops: how to get your products into ChatGPT, Perplexity and Google's AI answers

Short answer: GEO (generative engine optimisation) and AIO (AI optimisation, sometimes “AI Overviews optimisation”) are the AI side of SEO: making sure assistants such as ChatGPT, Perplexity, Claude and Google’s AI Overviews and AI Mode can find your shop, read your products and recommend them. For an online store that mostly comes down to three things: let AI crawlers in, publish product data they can read (prices, stock, GTINs, variants), and keep it that way as your store changes.

You can see where your store stands in about 20 seconds with the free check: it reads your store the way an AI assistant does and lists the fixes in order.

GEO, AIO, AI SEO: one idea, several names

The terms are new and used loosely, but they describe the same job:

• GEO (generative engine optimisation): getting your pages into answers written by generative AI, such as ChatGPT search, Perplexity and Copilot.
• AIO (AI optimisation, or AI Overviews optimisation): the same, with an eye on Google’s AI Overviews and AI Mode.
• AI SEO / LLM SEO: catch-all names for both.

Classic SEO decides where you appear in a list of links. GEO and AIO decide whether you’re mentioned at all when someone asks “where can I buy a waterproof dry bag in the UK?”, because an AI answer names a handful of shops, not ten pages of results.

How AI assistants choose which shops to recommend

When a shopper asks an assistant for a product, it typically:

1. Searches its own index, or a search engine’s, for pages that match.
2. Fetches some of those pages live to check details such as price and stock.
3. Reads what it fetched, preferring clear, machine-readable facts over guesswork.
4. Answers with the shops whose products it could understand and trust.

Each step can drop your store: an index that never crawled you, a firewall that blocks the live fetch, a page whose price only appears after JavaScript runs, or product data too thin to match the question. GEO for a shop is mostly removing those drop-off points.

The GEO checklist for online shops

1. Let AI assistants and AI search crawlers in

Check your robots.txt allows the assistant and search user agents (ChatGPT-User, OAI-SearchBot, Claude-User, Claude-SearchBot, Perplexity-User, PerplexityBot and others). Then check your bot protection: CDN and security-plugin settings often block these agents without anything looking wrong to you. Our guide to robots.txt for AI crawlers explains which to allow and gives a ready-made file.

2. Publish structured product data on every product page

AI assistants and shopping engines read schema.org Product markup (JSON-LD) to understand a page: name, price and currency, availability, brand, images and a GTIN (barcode). Without it they have to guess from the page text, and often skip you. This is usually one change to your product page template, and it fixes every product at once. See structured product data for AI shopping agents for the fields and a copy-ready example.

3. Describe sizes and colours properly

If a product comes in variants, mark them up (ProductGroup with each variant’s own price, stock and GTIN). Otherwise an assistant can’t answer “is it in stock in medium?” and moves on to a shop where it can.

4. Make your catalogue easy to find

Keep an XML sitemap that lists your product pages, and if you can, a product feed (the Google Merchant Center format works well). Search engines’ shopping results and AI answers draw on these to discover and refresh products.

5. Put the facts in the page, not behind JavaScript

Many AI agents read the page as it’s delivered, without running JavaScript. If your price, stock or description only appear after scripts run, an agent may see an empty product. Server-rendered content (the default on WooCommerce and Magento) is safest.

6. Write product descriptions that answer questions

Shoppers ask assistants specific things: materials, sizes, who it’s for, how it compares. A plain description that answers those questions gives an assistant something to quote. Keyword stuffing doesn’t help; specific facts do.

7. Add an llms.txt (cheap, optional)

An llms.txt file at the root of your site is a short, plain-text guide for AI tools: what you sell and where the important pages are. It’s a young convention and no guarantee of anything, but it takes ten minutes and does no harm.

8. Keep it working

GEO isn’t a one-off fix. A theme update can hide prices, a new security rule can block AI crawlers, and every new product needs complete data. Re-check regularly, or let Proxa Monitor check every day and email you when something breaks.

What doesn’t work

• Hidden text “for AI”: assistants and search engines treat it like any other hidden text, and it can count against you.
• Paying for “AI rankings”: nobody can guarantee your shop appears in ChatGPT or Google’s AI answers. Be wary of anyone who says they can.
• Blocking everything to “protect” content: blocking AI assistants and AI search crawlers stops them recommending you. Blocking training crawlers is a separate choice with little effect on recommendations today.

How to tell if it’s working

• Ask the assistants. Search for your products the way a shopper would, in ChatGPT, Perplexity and Google’s AI Mode, and note which shops come up.
• Look for AI visits. Your server logs (or Proxa Insights, for WooCommerce and Magento) show which AI agents fetch your pages and what they ask for.
• Watch referrals. Visits from chatgpt.com, perplexity.ai and similar sites in your analytics are shoppers an assistant sent your way.

Check your store now

The free check covers points 1 to 5 above in about 20 seconds: AI crawler access, structured data, variants, feeds and sitemaps, plus protocol readiness and the basics. No sign-up, and it reads only your public pages.