The Magento AEO Checklist: Make Your Store the One AI Recommends

When a shopper asks ChatGPT “where should I buy X in the UK?”, the engines recommend a handful of stores and ignore the rest. For Magento 2 merchants, getting into that handful comes down to crawler access, product schema, answer-shaped content and page speed. Work through this checklist top to bottom.

1. Crawler access

• robots.txt allows GPTBot, OAI-SearchBot, PerplexityBot, ClaudeBot, Google-Extended, Bingbot — run the free check to confirm all six.
• No bot-challenge rules (Cloudflare or similar) firing on AI crawler user agents — a challenge page is an empty page as far as the engine is concerned.
• Product and category content is server-rendered. Stock Luma is fine here; if you’ve gone headless, verify the rendered HTML actually contains your content, because most AI crawlers don’t execute JavaScript.
• Sitemap enabled (Stores → Configuration → Catalog → XML Sitemap) and referenced in robots.txt.

2. Product schema that engines can quote

Magento’s default structured data is thin. Every product page should carry complete JSON-LD Product markup:

• name, description, image, sku, brand
• offers with price, priceCurrency, availability — engines will not recommend a store when they can’t confirm the price
• aggregateRating and review where you have genuine reviews
• Category pages: ItemList markup; store-wide: Organization + WebSite with sameAs profile links

Test with Google’s Rich Results tool and by pasting rendered HTML into an LLM and asking “what does this store sell and at what price?” — if the model can’t answer, neither can the answer engine.

3. Answer-shaped commerce content

AI answers to shopping questions are assembled from comparison and guidance content, not from product grids:

• Category descriptions that answer the buying question in the first paragraph (“The best trade-quality X for Y starts at £Z…”) — not a keyword paragraph pasted under the products. Blank category descriptions are the most common wasted surface on Magento stores.
• A buying guide per major category, one H2 per real buyer question.
• FAQ blocks with FAQPage schema on delivery, returns, warranties, compatibility — the questions shoppers actually ask a chatbot before purchasing.
• Concrete facts everywhere: lead times, stock, dimensions, compatibility tables.

4. Speed and stability

Engines fetch with tight budgets; slow or flaky pages get sampled less and cited less:

• Sub-2.5s LCP on product pages (mobile).
• Full-page caching in front of Magento; no crawl-killing redirect chains.
• Consider a headless storefront — done right you get static-speed pages and perfect server-rendered HTML for AI crawlers, which is the strongest possible AEO foundation for Magento. Done wrong (client-side rendering) it makes AI visibility dramatically worse, so verify with the checker before and after any replatforming.

5. Entity trust

• Identical store name, address and one-liner on site, Google Business Profile, Trustpilot and social profiles.
• Get listed in “best UK X suppliers” roundups in your niche — Perplexity leans heavily on roundup articles when recommending merchants.
• Publish llms.txt indexing your top categories, buying guides and policies.

Monthly routine

1. Ask ChatGPT and Perplexity your five money questions (“best place to buy X in the UK…”). Log which stores get named.
2. Run the AI visibility check — confirm no regression after deploys or robots.txt changes.
3. Ship one new buying guide or upgraded category description.

The overlap with classic SEO is deliberate — everything here also strengthens your Google rankings. Start with AEO vs SEO if you want the full picture of what’s shared and what’s new.